RF.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "657b489e-5f3f-485b-996c-d6599daaabae",
   "metadata": {},
   "outputs": [],
   "source": [
    "import h2o"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "e397025b-553c-430b-b5e6-a6a57a427fb9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Checking whether there is an H2O instance running at http://localhost:54321..... not found.\n",
      "Attempting to start a local H2O server...\n",
      "; OpenJDK 64-Bit Server VM (build 19.0.2+7-44, mixed mode, sharing)\n",
      "  Starting server from C:\\Users\\Farzad\\anaconda3\\Lib\\site-packages\\h2o\\backend\\bin\\h2o.jar\n",
      "  Ice root: C:\\Users\\Farzad\\AppData\\Local\\Temp\\tmp66trxefi\n",
      "  JVM stdout: C:\\Users\\Farzad\\AppData\\Local\\Temp\\tmp66trxefi\\h2o_Farzad_started_from_python.out\n",
      "  JVM stderr: C:\\Users\\Farzad\\AppData\\Local\\Temp\\tmp66trxefi\\h2o_Farzad_started_from_python.err\n",
      "  Server is running at http://127.0.0.1:54321\n",
      "Connecting to H2O server at http://127.0.0.1:54321 ... successful.\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "\n",
       "<style>\n",
       "\n",
       "#h2o-table-1.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-1 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-1 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-1 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-1 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-1 .h2o-table th,\n",
       "#h2o-table-1 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-1 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-1\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption></caption>\n",
       "    <thead></thead>\n",
       "    <tbody><tr><td>H2O_cluster_uptime:</td>\n",
       "<td>01 secs</td></tr>\n",
       "<tr><td>H2O_cluster_timezone:</td>\n",
       "<td>Europe/Paris</td></tr>\n",
       "<tr><td>H2O_data_parsing_timezone:</td>\n",
       "<td>UTC</td></tr>\n",
       "<tr><td>H2O_cluster_version:</td>\n",
       "<td>3.40.0.2</td></tr>\n",
       "<tr><td>H2O_cluster_version_age:</td>\n",
       "<td>21 days, 20 hours and 43 minutes</td></tr>\n",
       "<tr><td>H2O_cluster_name:</td>\n",
       "<td>H2O_from_python_Farzad_eowqth</td></tr>\n",
       "<tr><td>H2O_cluster_total_nodes:</td>\n",
       "<td>1</td></tr>\n",
       "<tr><td>H2O_cluster_free_memory:</td>\n",
       "<td>3.844 Gb</td></tr>\n",
       "<tr><td>H2O_cluster_total_cores:</td>\n",
       "<td>16</td></tr>\n",
       "<tr><td>H2O_cluster_allowed_cores:</td>\n",
       "<td>16</td></tr>\n",
       "<tr><td>H2O_cluster_status:</td>\n",
       "<td>locked, healthy</td></tr>\n",
       "<tr><td>H2O_connection_url:</td>\n",
       "<td>http://127.0.0.1:54321</td></tr>\n",
       "<tr><td>H2O_connection_proxy:</td>\n",
       "<td>{\"http\": null, \"https\": null}</td></tr>\n",
       "<tr><td>H2O_internal_security:</td>\n",
       "<td>False</td></tr>\n",
       "<tr><td>Python_version:</td>\n",
       "<td>3.9.13 final</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n"
      ],
      "text/plain": [
       "--------------------------  --------------------------------\n",
       "H2O_cluster_uptime:         01 secs\n",
       "H2O_cluster_timezone:       Europe/Paris\n",
       "H2O_data_parsing_timezone:  UTC\n",
       "H2O_cluster_version:        3.40.0.2\n",
       "H2O_cluster_version_age:    21 days, 20 hours and 43 minutes\n",
       "H2O_cluster_name:           H2O_from_python_Farzad_eowqth\n",
       "H2O_cluster_total_nodes:    1\n",
       "H2O_cluster_free_memory:    3.844 Gb\n",
       "H2O_cluster_total_cores:    16\n",
       "H2O_cluster_allowed_cores:  16\n",
       "H2O_cluster_status:         locked, healthy\n",
       "H2O_connection_url:         http://127.0.0.1:54321\n",
       "H2O_connection_proxy:       {\"http\": null, \"https\": null}\n",
       "H2O_internal_security:      False\n",
       "Python_version:             3.9.13 final\n",
       "--------------------------  --------------------------------"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "h2o.init()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "20bfbf64-12c5-47ee-85cd-1293a137e545",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parse progress: |████████████████████████████████████████████████████████████████| (done) 100%\n"
     ]
    }
   ],
   "source": [
    "train_h2o = h2o.import_file('train_v1.csv')\n",
    "train_h2o['Anomaly'] = train_h2o['Anomaly'].asfactor()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "7dd3d7f9-9882-40d8-a838-3b4420af1dfd",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parse progress: |████████████████████████████████████████████████████████████████| (done) 100%\n"
     ]
    }
   ],
   "source": [
    "test_h2o = h2o.import_file('test_v1.csv')\n",
    "test_h2o['Anomaly'] = test_h2o['Anomaly'].asfactor()\n",
    "test_h2o = test_h2o.na_omit()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "49abc0f4-dd78-4060-9a63-d96117afb58c",
   "metadata": {},
   "outputs": [],
   "source": [
    "predictors = train_h2o.drop(['Anomaly','Index'], axis=1).columns\n",
    "response = 'Anomaly'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "95c016fb-43d9-4048-99b4-f4c3434e209c",
   "metadata": {},
   "outputs": [],
   "source": [
    "from h2o.estimators.random_forest import H2ORandomForestEstimator\n",
    "from h2o.grid.grid_search import H2OGridSearch\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "1b3e65a0-473f-4f43-a1fc-1452a898c3af",
   "metadata": {},
   "outputs": [],
   "source": [
    "hyper_parameters = {\n",
    "    'ntrees': [250, 500,1000,750]\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "a2b42ffc-761a-408a-ba51-12ba51f78f30",
   "metadata": {},
   "outputs": [],
   "source": [
    "rf_base_model = H2ORandomForestEstimator(seed=1,mtries=5,max_depth=None)\n",
    "grid_search = H2OGridSearch(model=rf_base_model,\n",
    "                            hyper_params=hyper_parameters,\n",
    "                            search_criteria={'strategy': 'RandomDiscrete', 'seed': 1})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "9207910d-e02d-43c9-97a7-1e5d8711d233",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "drf Grid Build progress: |███████████████████████████████████████████████████████| (done) 100%\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "\n",
       "<style>\n",
       "\n",
       "#h2o-table-2.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-2 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-2 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-2 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-2 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-2 .h2o-table th,\n",
       "#h2o-table-2 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-2 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-2\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Hyper-Parameter Search Summary: ordered by increasing logloss</caption>\n",
       "    <thead><tr><th></th>\n",
       "<th>ntrees</th>\n",
       "<th>model_ids</th>\n",
       "<th>logloss</th></tr></thead>\n",
       "    <tbody><tr><td></td>\n",
       "<td>750.0</td>\n",
       "<td>Grid_DRF_py_1_sid_8ee2_model_python_1680256840616_1_model_4</td>\n",
       "<td>0.0386293</td></tr>\n",
       "<tr><td></td>\n",
       "<td>500.0</td>\n",
       "<td>Grid_DRF_py_1_sid_8ee2_model_python_1680256840616_1_model_3</td>\n",
       "<td>0.0387643</td></tr>\n",
       "<tr><td></td>\n",
       "<td>1000.0</td>\n",
       "<td>Grid_DRF_py_1_sid_8ee2_model_python_1680256840616_1_model_1</td>\n",
       "<td>0.0387722</td></tr>\n",
       "<tr><td></td>\n",
       "<td>250.0</td>\n",
       "<td>Grid_DRF_py_1_sid_8ee2_model_python_1680256840616_1_model_2</td>\n",
       "<td>0.0388666</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n"
      ],
      "text/plain": [
       "Hyper-Parameter Search Summary: ordered by increasing logloss\n",
       "    ntrees    model_ids                                                    logloss\n",
       "--  --------  -----------------------------------------------------------  ---------\n",
       "    750       Grid_DRF_py_1_sid_8ee2_model_python_1680256840616_1_model_4  0.0386293\n",
       "    500       Grid_DRF_py_1_sid_8ee2_model_python_1680256840616_1_model_3  0.0387643\n",
       "    1000      Grid_DRF_py_1_sid_8ee2_model_python_1680256840616_1_model_1  0.0387722\n",
       "    250       Grid_DRF_py_1_sid_8ee2_model_python_1680256840616_1_model_2  0.0388666"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grid_search.train(x=predictors, y=response, training_frame=train_h2o)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "f133f17f-1870-48d3-b3da-b8d65496d2ee",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Model Details\n",
       "=============\n",
       "H2ORandomForestEstimator : Distributed Random Forest\n",
       "Model Key: Grid_DRF_py_1_sid_8ee2_model_python_1680256840616_1_model_4\n",
       "\n",
       "\n",
       "Model Summary: \n",
       "    number_of_trees    number_of_internal_trees    model_size_in_bytes    min_depth    max_depth    mean_depth    min_leaves    max_leaves    mean_leaves\n",
       "--  -----------------  --------------------------  ---------------------  -----------  -----------  ------------  ------------  ------------  -------------\n",
       "    750                750                         1.18245e+07            20           20           20            529           2104          1247.83\n",
       "\n",
       "ModelMetricsBinomial: drf\n",
       "** Reported on train data. **\n",
       "\n",
       "MSE: 0.00876326551021477\n",
       "RMSE: 0.09361231494955549\n",
       "LogLoss: 0.03862934399936804\n",
       "Mean Per-Class Error: 0.020403774742989475\n",
       "AUC: 0.9986596702389586\n",
       "AUCPR: 0.9940993331815087\n",
       "Gini: 0.9973193404779173\n",
       "\n",
       "Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.14396881333685163\n",
       "       0       1      Error    Rate\n",
       "-----  ------  -----  -------  -----------------\n",
       "0      194197  442    0.0023   (442.0/194639.0)\n",
       "1      1006    25099  0.0385   (1006.0/26105.0)\n",
       "Total  195203  25541  0.0066   (1448.0/220744.0)\n",
       "\n",
       "Maximum Metrics: Maximum metrics at their respective thresholds\n",
       "metric                       threshold    value     idx\n",
       "---------------------------  -----------  --------  -----\n",
       "max f1                       0.143969     0.971963  243\n",
       "max f2                       0.0859689    0.971238  268\n",
       "max f0point5                 0.258711     0.984314  212\n",
       "max accuracy                 0.147949     0.99344   242\n",
       "max precision                0.999988     1         0\n",
       "max recall                   0.0105712    1         363\n",
       "max specificity              0.999988     1         0\n",
       "max absolute_mcc             0.143969     0.968323  243\n",
       "max min_per_class_accuracy   0.0664525    0.985137  280\n",
       "max mean_per_class_accuracy  0.074615     0.985993  275\n",
       "max tns                      0.999988     194639    0\n",
       "max fns                      0.999988     17608     0\n",
       "max fps                      5.31039e-05  194639    399\n",
       "max tps                      0.0105712    26105     363\n",
       "max tnr                      0.999988     1         0\n",
       "max fnr                      0.999988     0.674507  0\n",
       "max fpr                      5.31039e-05  1         399\n",
       "max tpr                      0.0105712    1         363\n",
       "\n",
       "Gains/Lift Table: Avg response rate: 11.83 %, avg score: 11.79 %\n",
       "group    cumulative_data_fraction    lower_threshold    lift         cumulative_lift    response_rate    score       cumulative_response_rate    cumulative_score    capture_rate    cumulative_capture_rate    gain      cumulative_gain    kolmogorov_smirnov\n",
       "-------  --------------------------  -----------------  -----------  -----------------  ---------------  ----------  --------------------------  ------------------  --------------  -------------------------  --------  -----------------  --------------------\n",
       "1        0.0345196                   1                  8.456        8.456              1                1           1                           1                   0.291898        0.291898                   745.6     745.6              0.291898\n",
       "2        0.0400011                   0.999268           8.456        8.456              1                0.999771    1                           0.999969            0.0463513       0.338249                   745.6     745.6              0.338249\n",
       "3        0.0500036                   0.996949           8.456        8.456              1                0.998234    1                           0.999622            0.0845815       0.422831                   745.6     745.6              0.422831\n",
       "4        0.100003                    0.792424           8.45524      8.45562            0.999909         0.966908    0.999955                    0.983266            0.422754        0.845585                   745.524   745.562            0.84558\n",
       "5        0.150002                    0.0415816          2.9244       6.61194            0.345837         0.184885    0.781922                    0.717147            0.146217        0.991802                   192.44    561.194            0.954703\n",
       "6        0.200001                    0.0271066          0.0750828    4.97776            0.00887922       0.0327029   0.588666                    0.54604             0.00375407      0.995556                   -92.4917  397.776            0.902255\n",
       "7        0.299999                    0.0184421          0.0340937    3.3299             0.00403189       0.0219836   0.393791                    0.371357            0.00340931      0.998966                   -96.5906  232.99             0.792712\n",
       "8        0.400002                    0.0147032          0.00804422   2.49941            0.000951302      0.016408    0.295579                    0.282618            0.000804444     0.99977                    -99.1956  149.941            0.680209\n",
       "9        0.5                         0.0122225          0.0015323    1.99985            0.000181209      0.0134082   0.2365                      0.228777            0.000153227     0.999923                   -99.8468  99.9847            0.566973\n",
       "10       0.599998                    0.0101948          0.000766151  1.66667            9.06043e-05      0.0111871   0.197099                    0.192512            7.66137e-05     1                          -99.9234  66.6672            0.45365\n",
       "11       0.700001                    0.00835631         0            1.42857            0                0.00926296  0.168941                    0.166333            0               1                          -100      42.857             0.340235\n",
       "12       0.799999                    0.00649644         0            1.25               0                0.00747578  0.147824                    0.146476            0               1                          -100      25.0001            0.226825\n",
       "13       0.899997                    0.00330627         0            1.11111            0                0.0050131   0.131399                    0.130758            0               1                          -100      11.1114            0.113415\n",
       "14       1                           0                  0            1                  0                0.00175365  0.118259                    0.117858            0               1                          -100      0                  0\n",
       "\n",
       "Scoring History: \n",
       "     timestamp            duration          number_of_trees    training_rmse        training_logloss      training_auc        training_pr_auc     training_lift      training_classification_error\n",
       "---  -------------------  ----------------  -----------------  -------------------  --------------------  ------------------  ------------------  -----------------  -------------------------------\n",
       "     2023-03-31 12:03:47  2 min 38.452 sec  0.0                nan                  nan                   nan                 nan                 nan                nan\n",
       "     2023-03-31 12:03:47  2 min 38.545 sec  1.0                0.12356614338361153  0.18005137078994138   0.9681383397709885  0.9256120770981746  8.309322848071183  0.01611416350160404\n",
       "     2023-03-31 12:03:48  2 min 38.642 sec  2.0                0.11933823108044307  0.1683478121611553    0.9706427199134072  0.9334364698167581  8.317252443433322  0.015205926104007332\n",
       "     2023-03-31 12:03:48  2 min 38.750 sec  3.0                0.11950421657030527  0.1678200919736248    0.972167951428894   0.9376626101186627  8.322026471056695  0.01569905213270142\n",
       "     2023-03-31 12:03:48  2 min 38.858 sec  4.0                0.11765335599062994  0.14155635518622003   0.9744813890494384  0.9439172900498839  8.35396796895118   0.015342377260981912\n",
       "     2023-03-31 12:03:48  2 min 38.975 sec  5.0                0.11560025696853495  0.12269485875279688   0.9779051304823273  0.949996083425019   8.372256652134135  0.014861226441206814\n",
       "     2023-03-31 12:03:48  2 min 39.093 sec  6.0                0.11307670720428194  0.10458573518914849   0.9805319172701419  0.9562189939585805  8.395101476990419  0.01421122386789076\n",
       "     2023-03-31 12:03:48  2 min 39.211 sec  7.0                0.11123754900432054  0.09055795655104677   0.9825024001121536  0.9605477846049894  8.412705408990815  0.013694103652233022\n",
       "     2023-03-31 12:03:48  2 min 39.329 sec  8.0                0.10841484980313712  0.0787858007415495    0.9847009858288808  0.9652716213109077  8.42463173597659   0.012850065528363092\n",
       "     2023-03-31 12:03:48  2 min 39.453 sec  9.0                0.10677279338990388  0.07145057752240268   0.9864669526655273  0.9686248026178392  8.43158561237472   0.012308499910207353\n",
       "---  ---                  ---               ---                ---                  ---                   ---                 ---                 ---                ---\n",
       "     2023-03-31 12:04:20  3 min 10.685 sec  364.0              0.0938784286317644   0.03881330229697626   0.9985304880111622  0.9937435430070333  8.456004596820533  0.006559634689957598\n",
       "     2023-03-31 12:04:24  3 min 14.686 sec  409.0              0.09376456948333575  0.03873510117511102   0.9985578891329767  0.993807577505312   8.456004596820533  0.006555104555503208\n",
       "     2023-03-31 12:04:28  3 min 18.739 sec  455.0              0.09379874261322878  0.038765526346407696  0.9985461329040227  0.9938041887789107  8.456004596820533  0.006546044286594426\n",
       "     2023-03-31 12:04:32  3 min 22.763 sec  501.0              0.09377808992895037  0.03877091563216025   0.9985433966591513  0.9937763767752046  8.456004596820533  0.006505273076504911\n",
       "     2023-03-31 12:04:36  3 min 26.830 sec  548.0              0.09374411776183089  0.03873657597786283   0.9985502645294694  0.9938197383737123  8.456004596820533  0.006509803210959301\n",
       "     2023-03-31 12:04:40  3 min 30.854 sec  592.0              0.09381748476341661  0.03878838078912516   0.9985920576537718  0.9939073884515482  8.456004596820533  0.006509803210959301\n",
       "     2023-03-31 12:04:44  3 min 34.912 sec  639.0              0.09372651430871111  0.03871814208605994   0.9986173591115993  0.9939627651984904  8.456004596820533  0.006536984017685645\n",
       "     2023-03-31 12:04:48  3 min 38.947 sec  686.0              0.0937668422131925   0.038751821832396054  0.9986294698931776  0.9940149367084696  8.456004596820533  0.006514333345413692\n",
       "     2023-03-31 12:04:52  3 min 42.976 sec  731.0              0.0936181268883314   0.03863583355906762   0.9986687459201786  0.9941061028821648  8.456004596820533  0.006509803210959301\n",
       "     2023-03-31 12:04:54  3 min 44.715 sec  750.0              0.09361231494955549  0.03862934399936804   0.9986596702389586  0.9940993331815087  8.456004596820533  0.006559634689957598\n",
       "[51 rows x 10 columns]\n",
       "\n",
       "\n",
       "Variable Importances: \n",
       "variable         relative_importance    scaled_importance     percentage\n",
       "---------------  ---------------------  --------------------  ---------------------\n",
       "RollingSD_800    1701792.625            1.0                   0.16212885602302993\n",
       "RollingMean_800  1622702.75             0.9535255507409429    0.15459400673035867\n",
       "RollingMean_400  1610296.375            0.9462353704817589    0.15341205814473546\n",
       "RollingMean_200  914986.5625            0.5376604346842789    0.08717027120420714\n",
       "RollingSD_400    875501.5625            0.5144584302684941    0.08340855677083465\n",
       "RollingSD_200    469688.5625            0.275996355607664     0.044746973401195926\n",
       "RollingMean_100  342698.0625            0.20137474887693793   0.03264865766734288\n",
       "RollingSD_100    290041.875             0.1704331483984425    0.027632131378242778\n",
       "meandiffsd       218502.921875          0.12839573909600177   0.020816654297866187\n",
       "meanaddsd        214554.671875          0.1260756855583388    0.02044050667189271\n",
       "---              ---                    ---                   ---\n",
       "RollingMean_3    79953.078125           0.04698168093483188   0.007617086184003288\n",
       "RollingSD_5      62198.39453125         0.03654875078051886   0.005925607153116346\n",
       "diff50           54506.05859375         0.03202861370594434   0.0051927625001482935\n",
       "RollingSD_4      45940.96875            0.026995632766947737  0.004376771058123058\n",
       "diff30           39989.73046875         0.02349859194433282   0.0038097998300866665\n",
       "RollingSD_3      33845.82421875         0.019888336405706305  0.0032244732296583425\n",
       "diff5            29490.6171875          0.01732914854270214   0.0028095550290814543\n",
       "diff2            20835.2890625          0.012243142176327155  0.0019849666351752305\n",
       "RollingSD_2      17170.025390625        0.010089375837214596  0.0016357789624740022\n",
       "diff1            14637.75390625         0.008601373452449884  0.001394530838072559\n",
       "[36 rows x 4 columns]\n",
       "\n",
       "\n",
       "[tips]\n",
       "Use `model.explain()` to inspect the model.\n",
       "--\n",
       "Use `h2o.display.toggle_user_tips()` to switch on/off this section.,\n",
       " Model Details\n",
       "=============\n",
       "H2ORandomForestEstimator : Distributed Random Forest\n",
       "Model Key: Grid_DRF_py_1_sid_8ee2_model_python_1680256840616_1_model_3\n",
       "\n",
       "\n",
       "Model Summary: \n",
       "    number_of_trees    number_of_internal_trees    model_size_in_bytes    min_depth    max_depth    mean_depth    min_leaves    max_leaves    mean_leaves\n",
       "--  -----------------  --------------------------  ---------------------  -----------  -----------  ------------  ------------  ------------  -------------\n",
       "    500                500                         7.84675e+06            20           20           20            529           2104          1242.08\n",
       "\n",
       "ModelMetricsBinomial: drf\n",
       "** Reported on train data. **\n",
       "\n",
       "MSE: 0.008793003144326005\n",
       "RMSE: 0.09377101441450873\n",
       "LogLoss: 0.038764275879604494\n",
       "Mean Per-Class Error: 0.02048552613502108\n",
       "AUC: 0.998555980767991\n",
       "AUCPR: 0.993792544045919\n",
       "Gini: 0.997111961535982\n",
       "\n",
       "Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.1442250360302624\n",
       "       0       1      Error    Rate\n",
       "-----  ------  -----  -------  -----------------\n",
       "0      194195  444    0.0023   (444.0/194639.0)\n",
       "1      1010    25095  0.0387   (1010.0/26105.0)\n",
       "Total  195205  25539  0.0066   (1454.0/220744.0)\n",
       "\n",
       "Maximum Metrics: Maximum metrics at their respective thresholds\n",
       "metric                       threshold    value     idx\n",
       "---------------------------  -----------  --------  -----\n",
       "max f1                       0.144225     0.971846  241\n",
       "max f2                       0.0869308    0.970642  266\n",
       "max f0point5                 0.252136     0.984371  211\n",
       "max accuracy                 0.160052     0.993436  236\n",
       "max precision                0.999987     1         0\n",
       "max recall                   0.0102638    1         365\n",
       "max specificity              0.999987     1         0\n",
       "max absolute_mcc             0.160052     0.968255  236\n",
       "max min_per_class_accuracy   0.0654109    0.984541  281\n",
       "max mean_per_class_accuracy  0.0753907    0.985577  273\n",
       "max tns                      0.999987     194639    0\n",
       "max fns                      0.999987     16659     0\n",
       "max fps                      6.54995e-05  194639    399\n",
       "max tps                      0.0102638    26105     365\n",
       "max tnr                      0.999987     1         0\n",
       "max fnr                      0.999987     0.638154  0\n",
       "max fpr                      6.54995e-05  1         399\n",
       "max tpr                      0.0102638    1         365\n",
       "\n",
       "Gains/Lift Table: Avg response rate: 11.83 %, avg score: 11.79 %\n",
       "group    cumulative_data_fraction    lower_threshold    lift        cumulative_lift    response_rate    score       cumulative_response_rate    cumulative_score    capture_rate    cumulative_capture_rate    gain      cumulative_gain    kolmogorov_smirnov\n",
       "-------  --------------------------  -----------------  ----------  -----------------  ---------------  ----------  --------------------------  ------------------  --------------  -------------------------  --------  -----------------  --------------------\n",
       "1        0.0390543                   1                  8.456       8.456              1                1           1                           1                   0.330243        0.330243                   745.6     745.6              0.330243\n",
       "2        0.0400011                   0.999977           8.456       8.456              1                0.999987    1                           1                   0.00800613      0.338249                   745.6     745.6              0.338249\n",
       "3        0.0500036                   0.997941           8.456       8.456              1                0.999039    1                           0.999807            0.0845815       0.422831                   745.6     745.6              0.422831\n",
       "4        0.100003                    0.790767           8.456       8.456              1                0.966912    1                           0.98336             0.422793        0.845623                   745.6     745.6              0.845623\n",
       "5        0.150002                    0.0419894          2.90831     6.60683            0.343934         0.185061    0.781318                    0.717269            0.145413        0.991036                   190.831   560.683            0.953834\n",
       "6        0.200001                    0.0273737          0.0850427   4.97642            0.0100571        0.0330236   0.588507                    0.546211            0.00425206      0.995288                   -91.4957  397.642            0.901951\n",
       "7        0.299999                    0.0184825          0.0337106   3.32887            0.00398659       0.0220896   0.39367                     0.371507            0.003371        0.998659                   -96.6289  232.887            0.792365\n",
       "8        0.400002                    0.0146547          0.00995951  2.49913            0.0011778        0.0163915   0.295545                    0.282726            0.000995978     0.999655                   -99.004   149.913            0.680079\n",
       "9        0.5                         0.0121497          0.0015323   1.99962            0.000181209      0.0133494   0.236473                    0.228852            0.000153227     0.999808                   -99.8468  99.9617            0.566843\n",
       "10       0.599998                    0.0101563          0.00191538  1.66667            0.000226511      0.0111307   0.197099                    0.192565            0.000191534     1                          -99.8085  66.6672            0.45365\n",
       "11       0.700001                    0.00828759         0           1.42857            0                0.00920575  0.168941                    0.16637             0               1                          -100      42.857             0.340235\n",
       "12       0.799999                    0.00643481         0           1.25               0                0.00740752  0.147824                    0.1465              0               1                          -100      25.0001            0.226825\n",
       "13       0.899997                    0.00318098         0           1.11111            0                0.00495579  0.131399                    0.130773            0               1                          -100      11.1114            0.113415\n",
       "14       1                           0                  0           1                  0                0.00169476  0.118259                    0.117865            0               1                          -100      0                  0\n",
       "\n",
       "Scoring History: \n",
       "     timestamp            duration          number_of_trees    training_rmse        training_logloss      training_auc        training_pr_auc     training_lift      training_classification_error\n",
       "---  -------------------  ----------------  -----------------  -------------------  --------------------  ------------------  ------------------  -----------------  -------------------------------\n",
       "     2023-03-31 12:03:03  1 min 54.440 sec  0.0                nan                  nan                   nan                 nan                 nan                nan\n",
       "     2023-03-31 12:03:03  1 min 54.533 sec  1.0                0.12356614338361153  0.18005137078994138   0.9681383397709885  0.9256120770981746  8.309322848071183  0.01611416350160404\n",
       "     2023-03-31 12:03:04  1 min 54.628 sec  2.0                0.11933823108044307  0.1683478121611553    0.9706427199134072  0.9334364698167581  8.317252443433322  0.015205926104007332\n",
       "     2023-03-31 12:03:04  1 min 54.742 sec  3.0                0.11950421657030527  0.1678200919736248    0.972167951428894   0.9376626101186627  8.322026471056695  0.01569905213270142\n",
       "     2023-03-31 12:03:04  1 min 54.849 sec  4.0                0.11765335599062994  0.14155635518622003   0.9744813890494384  0.9439172900498839  8.35396796895118   0.015342377260981912\n",
       "     2023-03-31 12:03:04  1 min 54.959 sec  5.0                0.11560025696853495  0.12269485875279688   0.9779051304823273  0.949996083425019   8.372256652134135  0.014861226441206814\n",
       "     2023-03-31 12:03:04  1 min 55.074 sec  6.0                0.11307670720428194  0.10458573518914849   0.9805319172701419  0.9562189939585805  8.395101476990419  0.01421122386789076\n",
       "     2023-03-31 12:03:04  1 min 55.191 sec  7.0                0.11123754900432054  0.09055795655104677   0.9825024001121536  0.9605477846049894  8.412705408990815  0.013694103652233022\n",
       "     2023-03-31 12:03:04  1 min 55.303 sec  8.0                0.10841484980313712  0.0787858007415495    0.9847009858288808  0.9652716213109077  8.42463173597659   0.012850065528363092\n",
       "     2023-03-31 12:03:04  1 min 55.417 sec  9.0                0.10677279338990388  0.07145057752240268   0.9864669526655273  0.9686248026178392  8.43158561237472   0.012308499910207353\n",
       "---  ---                  ---               ---                ---                  ---                   ---                 ---                 ---                ---\n",
       "     2023-03-31 12:03:11  2 min  2.471 sec  83.0               0.09497295347284872  0.03961250493405792   0.9974742040062086  0.9912224734799048  8.456004596820533  0.007230094589207408\n",
       "     2023-03-31 12:03:15  2 min  6.548 sec  131.0              0.09453764280435786  0.03927366322351723   0.9978924808470166  0.9921925309794523  8.456004596820533  0.00694922625303519\n",
       "     2023-03-31 12:03:19  2 min 10.555 sec  177.0              0.09418431188492182  0.03901983339562515   0.9981547169424794  0.9927870348827145  8.456004596820533  0.006709129126952488\n",
       "     2023-03-31 12:03:24  2 min 14.618 sec  224.0              0.09398330339536647  0.038825498371301316  0.9983127642608365  0.9931938551622496  8.456004596820533  0.006659297647954191\n",
       "     2023-03-31 12:03:28  2 min 18.677 sec  271.0              0.09383158886802753  0.03872189013422673   0.9984257164806173  0.993472414416706   8.456004596820533  0.006586815496683942\n",
       "     2023-03-31 12:03:32  2 min 22.724 sec  319.0              0.09389669051870558  0.038781578346564705  0.9984694429647336  0.993564580967988   8.456004596820533  0.006604936034501504\n",
       "     2023-03-31 12:03:36  2 min 26.752 sec  365.0              0.09390050061305333  0.038831379751485824  0.9985302904142513  0.9937281617790645  8.456004596820533  0.006541514152140036\n",
       "     2023-03-31 12:03:40  2 min 30.763 sec  412.0              0.09376583826617686  0.03873479058254904   0.9985704912499016  0.9938288244087871  8.456004596820533  0.006541514152140036\n",
       "     2023-03-31 12:03:44  2 min 34.766 sec  459.0              0.09383708016273416  0.038790712728398506  0.9985642031808776  0.9938115626918305  8.456004596820533  0.006546044286594426\n",
       "     2023-03-31 12:03:47  2 min 38.386 sec  500.0              0.09377101441450873  0.038764275879604494  0.998555980767991   0.993792544045919   8.456004596820533  0.006586815496683942\n",
       "[46 rows x 10 columns]\n",
       "\n",
       "\n",
       "Variable Importances: \n",
       "variable         relative_importance    scaled_importance     percentage\n",
       "---------------  ---------------------  --------------------  ---------------------\n",
       "RollingSD_800    1132386.875            1.0                   0.16182596238136246\n",
       "RollingMean_800  1124733.5              0.9932413778639037    0.1607322418498167\n",
       "RollingMean_400  1067010.5              0.9422667496035751    0.15248322357455685\n",
       "RollingMean_200  595804.625             0.5261493559787153    0.08514462588758967\n",
       "RollingSD_400    572266.0               0.5053626217629906    0.08178079261836443\n",
       "RollingSD_200    316346.75              0.2793627840308552    0.045208151379329854\n",
       "RollingMean_100  226496.34375           0.20001675112138684   0.032367903242611885\n",
       "RollingSD_100    186271.796875          0.16449483916439778   0.0266195356545461\n",
       "meanaddsd        145359.125             0.12836525061278195   0.020772830216738217\n",
       "meandiffsd       142920.96875           0.12621213818819652   0.020424400726514406\n",
       "---              ---                    ---                   ---\n",
       "RollingSD_25     52245.1484375          0.04613719002836376   0.0074661951779117644\n",
       "RollingSD_5      42013.4453125          0.0371016710278455    0.006004013620037811\n",
       "diff50           37561.64453125         0.033170328410288226  0.005367820317500741\n",
       "RollingSD_4      30465.6171875          0.02690389464952073   0.004353748643465481\n",
       "diff30           27220.6484375          0.024038293836194453  0.0038900200340481407\n",
       "RollingSD_3      22604.453125           0.019961775983141805  0.00323033360931309\n",
       "diff5            20222.275390625        0.017858097649378884  0.0028899038384110846\n",
       "diff2            13092.845703125        0.01156216659887108   0.0018710587370759568\n",
       "RollingSD_2      11427.10546875         0.010091167357224977  0.001633012869134322\n",
       "diff1            9702.68359375          0.00856834692096727   0.0013865809865029124\n",
       "[36 rows x 4 columns]\n",
       "\n",
       "\n",
       "[tips]\n",
       "Use `model.explain()` to inspect the model.\n",
       "--\n",
       "Use `h2o.display.toggle_user_tips()` to switch on/off this section.,\n",
       " Model Details\n",
       "=============\n",
       "H2ORandomForestEstimator : Distributed Random Forest\n",
       "Model Key: Grid_DRF_py_1_sid_8ee2_model_python_1680256840616_1_model_1\n",
       "\n",
       "\n",
       "Model Summary: \n",
       "    number_of_trees    number_of_internal_trees    model_size_in_bytes    min_depth    max_depth    mean_depth    min_leaves    max_leaves    mean_leaves\n",
       "--  -----------------  --------------------------  ---------------------  -----------  -----------  ------------  ------------  ------------  -------------\n",
       "    1000               1000                        1.57199e+07            20           20           20            529           2104          1244.12\n",
       "\n",
       "ModelMetricsBinomial: drf\n",
       "** Reported on train data. **\n",
       "\n",
       "MSE: 0.008787596769409472\n",
       "RMSE: 0.09374218244424157\n",
       "LogLoss: 0.038772191821099936\n",
       "Mean Per-Class Error: 0.020643439130776937\n",
       "AUC: 0.9986719708434658\n",
       "AUCPR: 0.9940605676518495\n",
       "Gini: 0.9973439416869316\n",
       "\n",
       "Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.14668068235093254\n",
       "       0       1      Error    Rate\n",
       "-----  ------  -----  -------  -----------------\n",
       "0      194223  416    0.0021   (416.0/194639.0)\n",
       "1      1022    25083  0.0391   (1022.0/26105.0)\n",
       "Total  195245  25499  0.0065   (1438.0/220744.0)\n",
       "\n",
       "Maximum Metrics: Maximum metrics at their respective thresholds\n",
       "metric                       threshold    value     idx\n",
       "---------------------------  -----------  --------  -----\n",
       "max f1                       0.146681     0.972134  244\n",
       "max f2                       0.0867018    0.970911  269\n",
       "max f0point5                 0.265993     0.984262  211\n",
       "max accuracy                 0.146681     0.993486  244\n",
       "max precision                0.999985     1         0\n",
       "max recall                   0.0120333    1         362\n",
       "max specificity              0.999985     1         0\n",
       "max absolute_mcc             0.146681     0.968532  244\n",
       "max min_per_class_accuracy   0.0661424    0.985328  281\n",
       "max mean_per_class_accuracy  0.0749758    0.98567   275\n",
       "max tns                      0.999985     194639    0\n",
       "max fns                      0.999985     18209     0\n",
       "max fps                      6.62895e-05  194639    399\n",
       "max tps                      0.0120333    26105     362\n",
       "max tnr                      0.999985     1         0\n",
       "max fnr                      0.999985     0.697529  0\n",
       "max fpr                      6.62895e-05  1         399\n",
       "max tpr                      0.0120333    1         362\n",
       "\n",
       "Gains/Lift Table: Avg response rate: 11.83 %, avg score: 11.79 %\n",
       "group    cumulative_data_fraction    lower_threshold    lift        cumulative_lift    response_rate    score       cumulative_response_rate    cumulative_score    capture_rate    cumulative_capture_rate    gain      cumulative_gain    kolmogorov_smirnov\n",
       "-------  --------------------------  -----------------  ----------  -----------------  ---------------  ----------  --------------------------  ------------------  --------------  -------------------------  --------  -----------------  --------------------\n",
       "1        0.0314663                   1                  8.456       8.456              1                1           1                           1                   0.266079        0.266079                   745.6     745.6              0.266079\n",
       "2        0.0400011                   0.998674           8.456       8.456              1                0.999516    1                           0.999897            0.0721701       0.338249                   745.6     745.6              0.338249\n",
       "3        0.0500852                   0.997283           8.456       8.456              1                0.997829    1                           0.99948             0.085271        0.42352                    745.6     745.6              0.42352\n",
       "4        0.100003                    0.790391           8.45524     8.45562            0.999909         0.966574    0.999955                    0.983055            0.422065        0.845585                   745.524   745.562            0.84558\n",
       "5        0.150002                    0.0414513          2.92516     6.61219            0.345927         0.184582    0.781952                    0.716905            0.146256        0.991841                   192.516   561.219            0.954746\n",
       "6        0.200001                    0.0270692          0.0727843   4.97738            0.00860741       0.0326276   0.58862                     0.54584             0.00363915      0.99548                    -92.7216  397.738            0.902169\n",
       "7        0.299999                    0.0185067          0.035626    3.33015            0.0042131        0.0219907   0.393821                    0.371226            0.00356254      0.999042                   -96.4374  233.015            0.792799\n",
       "8        0.400002                    0.0147807          0.00766116  2.49951            0.000906002      0.0164696   0.29559                     0.282535            0.000766137     0.999808                   -99.2339  149.951            0.680253\n",
       "9        0.5                         0.0122854          0.00191538  2                  0.000226511      0.0134845   0.236518                    0.228726            0.000191534     1                          -99.8085  100                0.56706\n",
       "10       0.599998                    0.0102761          0           1.66667            0                0.0112545   0.197099                    0.192481            0               1                          -100      66.6672            0.45365\n",
       "11       0.700001                    0.00842227         0           1.42857            0                0.00933448  0.168941                    0.166317            0               1                          -100      42.857             0.340235\n",
       "12       0.799999                    0.00654586         0           1.25               0                0.00753547  0.147824                    0.146469            0               1                          -100      25.0001            0.226825\n",
       "13       0.899997                    0.00339627         0           1.11111            0                0.00508004  0.131399                    0.13076             0               1                          -100      11.1114            0.113415\n",
       "14       1                           0                  0           1                  0                0.00181055  0.118259                    0.117864            0               1                          -100      0                  0\n",
       "\n",
       "Scoring History: \n",
       "     timestamp            duration          number_of_trees    training_rmse        training_logloss      training_auc        training_pr_auc     training_lift      training_classification_error\n",
       "---  -------------------  ----------------  -----------------  -------------------  --------------------  ------------------  ------------------  -----------------  -------------------------------\n",
       "     2023-03-31 12:01:09  0.044 sec         0.0                nan                  nan                   nan                 nan                 nan                nan\n",
       "     2023-03-31 12:01:10  1.553 sec         1.0                0.12356614338361153  0.18005137078994138   0.9681383397709885  0.9256120770981746  8.309322848071183  0.01611416350160404\n",
       "     2023-03-31 12:01:12  2.602 sec         2.0                0.11933823108044307  0.1683478121611553    0.9706427199134072  0.9334364698167581  8.317252443433322  0.015205926104007332\n",
       "     2023-03-31 12:01:12  3.094 sec         3.0                0.11950421657030527  0.1678200919736248    0.972167951428894   0.9376626101186627  8.322026471056695  0.01569905213270142\n",
       "     2023-03-31 12:01:12  3.319 sec         4.0                0.11765335599062994  0.14155635518622003   0.9744813890494384  0.9439172900498839  8.35396796895118   0.015342377260981912\n",
       "     2023-03-31 12:01:12  3.519 sec         5.0                0.11560025696853495  0.12269485875279688   0.9779051304823273  0.949996083425019   8.372256652134135  0.014861226441206814\n",
       "     2023-03-31 12:01:13  3.724 sec         6.0                0.11307670720428194  0.10458573518914849   0.9805319172701419  0.9562189939585805  8.395101476990419  0.01421122386789076\n",
       "     2023-03-31 12:01:13  3.972 sec         7.0                0.11123754900432054  0.09055795655104677   0.9825024001121536  0.9605477846049894  8.412705408990815  0.013694103652233022\n",
       "     2023-03-31 12:01:17  8.027 sec         47.0               0.09562066410265267  0.04000072141041679   0.9967345905030699  0.9888777915610875  8.456004596820533  0.007936795564092342\n",
       "     2023-03-31 12:01:21  12.030 sec        92.0               0.09485854827881213  0.03953043094159284   0.9976284956055929  0.9915588523267173  8.456004596820533  0.00709419055557569\n",
       "---  ---                  ---               ---                ---                  ---                   ---                 ---                 ---                ---\n",
       "     2023-03-31 12:02:05  56.412 sec        594.0              0.09379959763785162  0.03877584760015194   0.9985930121806815  0.9939213099124533  8.456004596820533  0.006496212807596129\n",
       "     2023-03-31 12:02:09  1 min  0.475 sec  640.0              0.09372987018087134  0.03872097564891996   0.998614374886541   0.9939715971168206  8.456004596820533  0.006536984017685645\n",
       "     2023-03-31 12:02:13  1 min  4.539 sec  687.0              0.09377652534208619  0.03875744952220617   0.9986303830900563  0.9940178976256349  8.456004596820533  0.006527923748776864\n",
       "     2023-03-31 12:02:17  1 min  8.549 sec  733.0              0.0936156534829607   0.038633995590187466  0.9986546173474368  0.9940905475530059  8.456004596820533  0.006505273076504911\n",
       "     2023-03-31 12:02:21  1 min 12.560 sec  780.0              0.09356214159127706  0.038595105693731466  0.9986762172089514  0.9941237680547789  8.456004596820533  0.006505273076504911\n",
       "     2023-03-31 12:02:25  1 min 16.597 sec  828.0              0.0935733392995092   0.03859977056723037   0.9986788769932651  0.9941080178226754  8.456004596820533  0.006509803210959301\n",
       "     2023-03-31 12:02:30  1 min 20.648 sec  876.0              0.093659419879776    0.038679598536223055  0.998662865050366   0.9940627364802973  8.456004596820533  0.006514333345413692\n",
       "     2023-03-31 12:02:34  1 min 24.730 sec  924.0              0.09375182294581745  0.038770718152232826  0.9986509649535418  0.9940249300324462  8.456004596820533  0.006559634689957598\n",
       "     2023-03-31 12:02:38  1 min 28.731 sec  971.0              0.09372310533706434  0.038755437659419946  0.9986691907100376  0.9940486020090279  8.456004596820533  0.006559634689957598\n",
       "     2023-03-31 12:02:40  1 min 31.195 sec  1000.0             0.09374218244424157  0.038772191821099936  0.9986719708434658  0.9940605676518495  8.456004596820533  0.006514333345413692\n",
       "[30 rows x 10 columns]\n",
       "\n",
       "\n",
       "Variable Importances: \n",
       "variable         relative_importance    scaled_importance     percentage\n",
       "---------------  ---------------------  --------------------  ---------------------\n",
       "RollingMean_800  2267263.75             1.0                   0.1616998840599283\n",
       "RollingSD_800    2242460.75             0.9890603817045988    0.15993094904990207\n",
       "RollingMean_400  2130553.25             0.9397024276509515    0.15194977360199202\n",
       "RollingMean_200  1188431.75             0.5241700485883038    0.08475823608441573\n",
       "RollingSD_400    1170099.5              0.5160844211442096    0.08345079106415391\n",
       "RollingSD_200    629952.8125            0.2778471681999944    0.04492785488431849\n",
       "RollingMean_100  457740.28125           0.20189105976311755   0.032645760956432165\n",
       "RollingSD_100    380024.40625           0.16761367364074867   0.027103111594567724\n",
       "meandiffsd       284386.90625           0.12543177045458428   0.02028230273993782\n",
       "meanaddsd        283921.3125            0.12522641554164132   0.02024909687432381\n",
       "---              ---                    ---                   ---\n",
       "RollingMean_3    105856.9140625         0.04668928088428177   0.0075496513058297895\n",
       "RollingSD_5      84553.4140625          0.03729315306280533   0.006030298526484782\n",
       "diff50           73058.9140625          0.03222338559530183   0.005210517714778669\n",
       "RollingSD_4      61549.79296875         0.027147169343994495  0.00438969413547915\n",
       "diff30           51977.16796875         0.022925064615332027  0.003706980290365554\n",
       "RollingSD_3      45336.94140625         0.01999632438274991   0.003233403334315378\n",
       "diff5            39115.7109375          0.01725238668747736   0.00278970892712214\n",
       "diff2            28626.916015625        0.01262619579024496   0.0020416543954005647\n",
       "RollingSD_2      22211.353515625        0.009796545953519965  0.001584100344871938\n",
       "diff1            19616.64453125         0.008652122864510139  0.0013990472640635443\n",
       "[36 rows x 4 columns]\n",
       "\n",
       "\n",
       "[tips]\n",
       "Use `model.explain()` to inspect the model.\n",
       "--\n",
       "Use `h2o.display.toggle_user_tips()` to switch on/off this section.,\n",
       " Model Details\n",
       "=============\n",
       "H2ORandomForestEstimator : Distributed Random Forest\n",
       "Model Key: Grid_DRF_py_1_sid_8ee2_model_python_1680256840616_1_model_2\n",
       "\n",
       "\n",
       "Model Summary: \n",
       "    number_of_trees    number_of_internal_trees    model_size_in_bytes    min_depth    max_depth    mean_depth    min_leaves    max_leaves    mean_leaves\n",
       "--  -----------------  --------------------------  ---------------------  -----------  -----------  ------------  ------------  ------------  -------------\n",
       "    250                250                         3.93607e+06            20           20           20            634           2023          1246.13\n",
       "\n",
       "ModelMetricsBinomial: drf\n",
       "** Reported on train data. **\n",
       "\n",
       "MSE: 0.008832658384371717\n",
       "RMSE: 0.09398222376796432\n",
       "LogLoss: 0.038866566796179294\n",
       "Mean Per-Class Error: 0.020174611824091488\n",
       "AUC: 0.9983650968382951\n",
       "AUCPR: 0.9933066714645468\n",
       "Gini: 0.9967301936765902\n",
       "\n",
       "Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.14269225393334023\n",
       "       0       1      Error    Rate\n",
       "-----  ------  -----  -------  -----------------\n",
       "0      194152  487    0.0025   (487.0/194639.0)\n",
       "1      988     25117  0.0378   (988.0/26105.0)\n",
       "Total  195140  25604  0.0067   (1475.0/220744.0)\n",
       "\n",
       "Maximum Metrics: Maximum metrics at their respective thresholds\n",
       "metric                       threshold    value     idx\n",
       "---------------------------  -----------  --------  -----\n",
       "max f1                       0.142692     0.971475  240\n",
       "max f2                       0.0989797    0.969783  257\n",
       "max f0point5                 0.259594     0.984253  208\n",
       "max accuracy                 0.155422     0.993323  236\n",
       "max precision                0.999994     1         0\n",
       "max recall                   0.00881617   1         371\n",
       "max specificity              0.999994     1         0\n",
       "max absolute_mcc             0.142692     0.967749  240\n",
       "max min_per_class_accuracy   0.0668946    0.983878  277\n",
       "max mean_per_class_accuracy  0.0728594    0.984526  272\n",
       "max tns                      0.999994     194639    0\n",
       "max fns                      0.999994     15147     0\n",
       "max fps                      5.35447e-05  194639    399\n",
       "max tps                      0.00881617   26105     371\n",
       "max tnr                      0.999994     1         0\n",
       "max fnr                      0.999994     0.580234  0\n",
       "max fpr                      5.35447e-05  1         399\n",
       "max tpr                      0.00881617   1         371\n",
       "\n",
       "Gains/Lift Table: Avg response rate: 11.83 %, avg score: 11.79 %\n",
       "group    cumulative_data_fraction    lower_threshold    lift         cumulative_lift    response_rate    score       cumulative_response_rate    cumulative_score    capture_rate    cumulative_capture_rate    gain      cumulative_gain    kolmogorov_smirnov\n",
       "-------  --------------------------  -----------------  -----------  -----------------  ---------------  ----------  --------------------------  ------------------  --------------  -------------------------  --------  -----------------  --------------------\n",
       "1        0.0468325                   1                  8.456        8.456              1                1           1                           1                   0.396016        0.396016                   745.6     745.6              0.396016\n",
       "2        0.0500082                   0.999736           8.456        8.456              1                0.999904    1                           0.999994            0.0268531       0.422869                   745.6     745.6              0.422869\n",
       "3        0.100003                    0.788949           8.456        8.456              1                0.967387    1                           0.983693            0.422754        0.845623                   745.6     745.6              0.845623\n",
       "4        0.150002                    0.0437943          2.88992      6.6007             0.34176          0.186356    0.780593                    0.717922            0.144493        0.990117                   188.992   560.07             0.952791\n",
       "5        0.200001                    0.0278652          0.101132     4.97584            0.0119598        0.0340984   0.588439                    0.54697             0.0050565       0.995173                   -89.8868  397.584            0.901821\n",
       "6        0.299999                    0.0183856          0.0291137    3.32696            0.00344296       0.0222708   0.393443                    0.372073            0.00291132      0.998085                   -97.0886  232.696            0.791713\n",
       "7        0.400002                    0.0143696          0.0126409    2.49836            0.0014949        0.016165    0.295454                    0.283094            0.00126413      0.999349                   -98.7359  149.836            0.679731\n",
       "8        0.5                         0.0118646          0.00536305   1.99977            0.00063423       0.0130529   0.236491                    0.229087            0.000536296     0.999885                   -99.4637  99.977             0.56693\n",
       "9        0.599998                    0.00985419         0.000766151  1.66661            9.06043e-05      0.0108328   0.197092                    0.192712            7.66137e-05     0.999962                   -99.9234  66.6608            0.453607\n",
       "10       0.700001                    0.00805033         0.000383058  1.42857            4.53001e-05      0.00893957  0.168941                    0.166458            3.83068e-05     1                          -99.9617  42.857             0.340235\n",
       "11       0.799999                    0.00620931         0            1.25               0                0.00718596  0.147824                    0.146549            0               1                          -100      25.0001            0.226825\n",
       "12       0.899997                    0.0030333          0            1.11111            0                0.00473529  0.131399                    0.130792            0               1                          -100      11.1114            0.113415\n",
       "13       1                           0                  0            1                  0                0.00159943  0.118259                    0.117873            0               1                          -100      0                  0\n",
       "\n",
       "Scoring History: \n",
       "     timestamp            duration          number_of_trees    training_rmse        training_logloss      training_auc        training_pr_auc     training_lift      training_classification_error\n",
       "---  -------------------  ----------------  -----------------  -------------------  --------------------  ------------------  ------------------  -----------------  -------------------------------\n",
       "     2023-03-31 12:02:40  1 min 31.254 sec  0.0                nan                  nan                   nan                 nan                 nan                nan\n",
       "     2023-03-31 12:02:40  1 min 31.385 sec  1.0                0.12356614338361153  0.18005137078994138   0.9681383397709885  0.9256120770981746  8.309322848071183  0.01611416350160404\n",
       "     2023-03-31 12:02:40  1 min 31.513 sec  2.0                0.11933823108044307  0.1683478121611553    0.9706427199134072  0.9334364698167581  8.317252443433322  0.015205926104007332\n",
       "     2023-03-31 12:02:41  1 min 31.654 sec  3.0                0.11950421657030527  0.1678200919736248    0.972167951428894   0.9376626101186627  8.322026471056695  0.01569905213270142\n",
       "     2023-03-31 12:02:41  1 min 31.758 sec  4.0                0.11765335599062994  0.14155635518622003   0.9744813890494384  0.9439172900498839  8.35396796895118   0.015342377260981912\n",
       "     2023-03-31 12:02:41  1 min 31.867 sec  5.0                0.11560025696853495  0.12269485875279688   0.9779051304823273  0.949996083425019   8.372256652134135  0.014861226441206814\n",
       "     2023-03-31 12:02:41  1 min 31.982 sec  6.0                0.11307670720428194  0.10458573518914849   0.9805319172701419  0.9562189939585805  8.395101476990419  0.01421122386789076\n",
       "     2023-03-31 12:02:41  1 min 32.100 sec  7.0                0.11123754900432054  0.09055795655104677   0.9825024001121536  0.9605477846049894  8.412705408990815  0.013694103652233022\n",
       "     2023-03-31 12:02:41  1 min 32.216 sec  8.0                0.10841484980313712  0.0787858007415495    0.9847009858288808  0.9652716213109077  8.42463173597659   0.012850065528363092\n",
       "     2023-03-31 12:02:41  1 min 32.331 sec  9.0                0.10677279338990388  0.07145057752240268   0.9864669526655273  0.9686248026178392  8.43158561237472   0.012308499910207353\n",
       "---  ---                  ---               ---                ---                  ---                   ---                 ---                 ---                ---\n",
       "     2023-03-31 12:02:44  1 min 34.750 sec  30.0               0.09643047450611354  0.04059673529417404   0.9958963918862147  0.9867999932825667  8.456004596820533  0.008525713043163121\n",
       "     2023-03-31 12:02:44  1 min 34.861 sec  31.0               0.09666028221918359  0.0406561962210137    0.995932139411009   0.9868022824416514  8.456004596820533  0.00852118290870873\n",
       "     2023-03-31 12:02:44  1 min 34.980 sec  32.0               0.09700292222595112  0.040922414023576915  0.9959519139612195  0.9868505905932174  8.456004596820533  0.008548363715435074\n",
       "     2023-03-31 12:02:44  1 min 35.100 sec  33.0               0.09692909935555034  0.04090229117173476   0.9960232858079534  0.9869866268348438  8.456004596820533  0.008516652774254339\n",
       "     2023-03-31 12:02:44  1 min 35.221 sec  34.0               0.09659904317994042  0.04061157999911285   0.996127176418406   0.9873055202659694  8.456004596820533  0.008453230891892871\n",
       "     2023-03-31 12:02:48  1 min 39.306 sec  81.0               0.09492087980751868  0.03956233480673826   0.9974720335891446  0.9911812758836616  8.456004596820533  0.007284456202660095\n",
       "     2023-03-31 12:02:52  1 min 43.336 sec  126.0              0.0944816311357562   0.03925755170996604   0.9978970541133675  0.9921878720594828  8.456004596820533  0.00687674410176494\n",
       "     2023-03-31 12:02:56  1 min 47.411 sec  172.0              0.09411131900363887  0.03895933349979217   0.9981362513733981  0.9927626025794368  8.456004596820533  0.0067363099336788315\n",
       "     2023-03-31 12:03:00  1 min 51.420 sec  217.0              0.09410995463443168  0.038920713266311115  0.9982813281471242  0.9931254393324402  8.456004596820533  0.006704598992498098\n",
       "     2023-03-31 12:03:03  1 min 54.362 sec  250.0              0.09398222376796432  0.038866566796179294  0.9983650968382951  0.9933066714645468  8.456004596820533  0.006681948320226144\n",
       "[40 rows x 10 columns]\n",
       "\n",
       "\n",
       "Variable Importances: \n",
       "variable         relative_importance    scaled_importance     percentage\n",
       "---------------  ---------------------  --------------------  ---------------------\n",
       "RollingMean_800  578920.25              1.0                   0.16497924194226976\n",
       "RollingMean_400  554233.4375            0.957357144615342     0.15794405598665506\n",
       "RollingSD_800    537411.1875            0.9282991698770253    0.15315009334194993\n",
       "RollingMean_200  302899.5625            0.523214661259474     0.08631955818766947\n",
       "RollingSD_400    299645.34375           0.517593474662529     0.08539217908408946\n",
       "RollingSD_200    159688.796875          0.27583902424384016   0.045507713117844115\n",
       "RollingMean_100  107877.703125          0.18634294296148735   0.030742717471077794\n",
       "RollingSD_100    88614.2421875          0.15306813362894112   0.025253064651620753\n",
       "meanaddsd        73032.6796875          0.12615326495747903   0.020812670021227192\n",
       "meandiffsd       69459.5546875          0.11998121448938778   0.01979440981377207\n",
       "---              ---                    ---                   ---\n",
       "RollingMean_3    25506.146484375        0.04405813492337675   0.007268677701048937\n",
       "RollingSD_5      21549.966796875        0.03722441354724593   0.00614125552877019\n",
       "diff50           18917.818359375        0.03267776236774409   0.005391152463800049\n",
       "RollingSD_4      16055.107421875        0.02773284821506071   0.004575344275420745\n",
       "diff30           14769.77734375         0.025512628628468256  0.004209054131079342\n",
       "RollingSD_3      11568.3349609375       0.019982605481389706  0.0032967151043511184\n",
       "diff5            9414.802734375         0.016262693755098392  0.002683006887655417\n",
       "diff2            6405.67529296875       0.011064866521716506  0.001825473290945188\n",
       "RollingSD_2      6173.0947265625        0.01066311763418623   0.0017591930640292928\n",
       "diff1            5127.4921875           0.008856992284343137  0.001461219872959463\n",
       "[36 rows x 4 columns]\n",
       "\n",
       "\n",
       "[tips]\n",
       "Use `model.explain()` to inspect the model.\n",
       "--\n",
       "Use `h2o.display.toggle_user_tips()` to switch on/off this section.]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grid_search.models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "b985c0aa",
   "metadata": {},
   "outputs": [],
   "source": [
    "validation_h2o = test_h2o[:100000, :]\n",
    "test = test_h2o[100000:, :]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "dd0bbf8d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<pre style='margin: 1em 0 1em 0;'>ModelMetricsBinomial: drf\n",
       "** Reported on test data. **\n",
       "\n",
       "MSE: 0.013074625276969178\n",
       "RMSE: 0.11434432769914378\n",
       "LogLoss: 0.06041470000271687\n",
       "Mean Per-Class Error: 0.12998313353680432\n",
       "AUC: 0.9915474959363119\n",
       "AUCPR: 0.9025074939549784\n",
       "Gini: 0.9830949918726237</pre>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-18.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-18 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-18 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-18 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-18 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-18 .h2o-table th,\n",
       "#h2o-table-18 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-18 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-18\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.8134880277749358</caption>\n",
       "    <thead><tr><th></th>\n",
       "<th>0</th>\n",
       "<th>1</th>\n",
       "<th>Error</th>\n",
       "<th>Rate</th></tr></thead>\n",
       "    <tbody><tr><td>0</td>\n",
       "<td>95493.0</td>\n",
       "<td>305.0</td>\n",
       "<td>0.0032</td>\n",
       "<td> (305.0/95798.0)</td></tr>\n",
       "<tr><td>1</td>\n",
       "<td>1079.0</td>\n",
       "<td>3123.0</td>\n",
       "<td>0.2568</td>\n",
       "<td> (1079.0/4202.0)</td></tr>\n",
       "<tr><td>Total</td>\n",
       "<td>96572.0</td>\n",
       "<td>3428.0</td>\n",
       "<td>0.0138</td>\n",
       "<td> (1384.0/100000.0)</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-19.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-19 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-19 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-19 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-19 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-19 .h2o-table th,\n",
       "#h2o-table-19 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-19 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-19\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Maximum Metrics: Maximum metrics at their respective thresholds</caption>\n",
       "    <thead><tr><th>metric</th>\n",
       "<th>threshold</th>\n",
       "<th>value</th>\n",
       "<th>idx</th></tr></thead>\n",
       "    <tbody><tr><td>max f1</td>\n",
       "<td>0.8134880</td>\n",
       "<td>0.8186107</td>\n",
       "<td>57.0</td></tr>\n",
       "<tr><td>max f2</td>\n",
       "<td>0.2216464</td>\n",
       "<td>0.8505727</td>\n",
       "<td>192.0</td></tr>\n",
       "<tr><td>max f0point5</td>\n",
       "<td>0.8568616</td>\n",
       "<td>0.8748983</td>\n",
       "<td>48.0</td></tr>\n",
       "<tr><td>max accuracy</td>\n",
       "<td>0.8195778</td>\n",
       "<td>0.98616</td>\n",
       "<td>56.0</td></tr>\n",
       "<tr><td>max precision</td>\n",
       "<td>0.9991262</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max recall</td>\n",
       "<td>0.0200631</td>\n",
       "<td>1.0</td>\n",
       "<td>348.0</td></tr>\n",
       "<tr><td>max specificity</td>\n",
       "<td>0.9991262</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max absolute_mcc</td>\n",
       "<td>0.8134880</td>\n",
       "<td>0.8160406</td>\n",
       "<td>57.0</td></tr>\n",
       "<tr><td>max min_per_class_accuracy</td>\n",
       "<td>0.1347039</td>\n",
       "<td>0.9563978</td>\n",
       "<td>231.0</td></tr>\n",
       "<tr><td>max mean_per_class_accuracy</td>\n",
       "<td>0.1700958</td>\n",
       "<td>0.9592078</td>\n",
       "<td>213.0</td></tr>\n",
       "<tr><td>max tns</td>\n",
       "<td>0.9991262</td>\n",
       "<td>95798.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fns</td>\n",
       "<td>0.9991262</td>\n",
       "<td>3932.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fps</td>\n",
       "<td>0.0001844</td>\n",
       "<td>95798.0</td>\n",
       "<td>399.0</td></tr>\n",
       "<tr><td>max tps</td>\n",
       "<td>0.0200631</td>\n",
       "<td>4202.0</td>\n",
       "<td>348.0</td></tr>\n",
       "<tr><td>max tnr</td>\n",
       "<td>0.9991262</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fnr</td>\n",
       "<td>0.9991262</td>\n",
       "<td>0.9357449</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fpr</td>\n",
       "<td>0.0001844</td>\n",
       "<td>1.0</td>\n",
       "<td>399.0</td></tr>\n",
       "<tr><td>max tpr</td>\n",
       "<td>0.0200631</td>\n",
       "<td>1.0</td>\n",
       "<td>348.0</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-20.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-20 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-20 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-20 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-20 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-20 .h2o-table th,\n",
       "#h2o-table-20 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-20 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-20\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Gains/Lift Table: Avg response rate:  4.20 %, avg score:  6.98 %</caption>\n",
       "    <thead><tr><th>group</th>\n",
       "<th>cumulative_data_fraction</th>\n",
       "<th>lower_threshold</th>\n",
       "<th>lift</th>\n",
       "<th>cumulative_lift</th>\n",
       "<th>response_rate</th>\n",
       "<th>score</th>\n",
       "<th>cumulative_response_rate</th>\n",
       "<th>cumulative_score</th>\n",
       "<th>capture_rate</th>\n",
       "<th>cumulative_capture_rate</th>\n",
       "<th>gain</th>\n",
       "<th>cumulative_gain</th>\n",
       "<th>kolmogorov_smirnov</th></tr></thead>\n",
       "    <tbody><tr><td>1</td>\n",
       "<td>0.01</td>\n",
       "<td>0.9966851</td>\n",
       "<td>23.7981913</td>\n",
       "<td>23.7981913</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9983258</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9983258</td>\n",
       "<td>0.2379819</td>\n",
       "<td>0.2379819</td>\n",
       "<td>2279.8191337</td>\n",
       "<td>2279.8191337</td>\n",
       "<td>0.2379819</td></tr>\n",
       "<tr><td>2</td>\n",
       "<td>0.02</td>\n",
       "<td>0.9816592</td>\n",
       "<td>23.7029986</td>\n",
       "<td>23.7505950</td>\n",
       "<td>0.996</td>\n",
       "<td>0.9914786</td>\n",
       "<td>0.998</td>\n",
       "<td>0.9949022</td>\n",
       "<td>0.2370300</td>\n",
       "<td>0.4750119</td>\n",
       "<td>2270.2998572</td>\n",
       "<td>2275.0594955</td>\n",
       "<td>0.4749701</td></tr>\n",
       "<tr><td>3</td>\n",
       "<td>0.03</td>\n",
       "<td>0.8764554</td>\n",
       "<td>19.1337458</td>\n",
       "<td>22.2116452</td>\n",
       "<td>0.804</td>\n",
       "<td>0.9472527</td>\n",
       "<td>0.9333333</td>\n",
       "<td>0.9790190</td>\n",
       "<td>0.1913375</td>\n",
       "<td>0.6663494</td>\n",
       "<td>1813.3745835</td>\n",
       "<td>2121.1645248</td>\n",
       "<td>0.6642616</td></tr>\n",
       "<tr><td>4</td>\n",
       "<td>0.04</td>\n",
       "<td>0.5443167</td>\n",
       "<td>10.8995716</td>\n",
       "<td>19.3836268</td>\n",
       "<td>0.458</td>\n",
       "<td>0.7545140</td>\n",
       "<td>0.8145</td>\n",
       "<td>0.9228928</td>\n",
       "<td>0.1089957</td>\n",
       "<td>0.7753451</td>\n",
       "<td>989.9571633</td>\n",
       "<td>1838.3626844</td>\n",
       "<td>0.7675996</td></tr>\n",
       "<tr><td>5</td>\n",
       "<td>0.05</td>\n",
       "<td>0.3552505</td>\n",
       "<td>10.6615897</td>\n",
       "<td>17.6392194</td>\n",
       "<td>0.448</td>\n",
       "<td>0.4339174</td>\n",
       "<td>0.7412</td>\n",
       "<td>0.8250977</td>\n",
       "<td>0.1066159</td>\n",
       "<td>0.8819610</td>\n",
       "<td>966.1589719</td>\n",
       "<td>1663.9219419</td>\n",
       "<td>0.8684534</td></tr>\n",
       "<tr><td>6</td>\n",
       "<td>0.1</td>\n",
       "<td>0.0998281</td>\n",
       "<td>1.7658258</td>\n",
       "<td>9.7025226</td>\n",
       "<td>0.0742</td>\n",
       "<td>0.1748601</td>\n",
       "<td>0.4077</td>\n",
       "<td>0.4999789</td>\n",
       "<td>0.0882913</td>\n",
       "<td>0.9702523</td>\n",
       "<td>76.5825797</td>\n",
       "<td>870.2522608</td>\n",
       "<td>0.9084242</td></tr>\n",
       "<tr><td>7</td>\n",
       "<td>0.15</td>\n",
       "<td>0.0632485</td>\n",
       "<td>0.3284150</td>\n",
       "<td>6.5778201</td>\n",
       "<td>0.0138</td>\n",
       "<td>0.0786937</td>\n",
       "<td>0.2764</td>\n",
       "<td>0.3595505</td>\n",
       "<td>0.0164208</td>\n",
       "<td>0.9866730</td>\n",
       "<td>-67.1584960</td>\n",
       "<td>557.7820086</td>\n",
       "<td>0.8733721</td></tr>\n",
       "<tr><td>8</td>\n",
       "<td>0.2</td>\n",
       "<td>0.0467124</td>\n",
       "<td>0.1047120</td>\n",
       "<td>4.9595431</td>\n",
       "<td>0.0044</td>\n",
       "<td>0.0540732</td>\n",
       "<td>0.2084</td>\n",
       "<td>0.2831812</td>\n",
       "<td>0.0052356</td>\n",
       "<td>0.9919086</td>\n",
       "<td>-89.5287958</td>\n",
       "<td>395.9543075</td>\n",
       "<td>0.8266442</td></tr>\n",
       "<tr><td>9</td>\n",
       "<td>0.3</td>\n",
       "<td>0.0301801</td>\n",
       "<td>0.0499762</td>\n",
       "<td>3.3230208</td>\n",
       "<td>0.0021</td>\n",
       "<td>0.0373094</td>\n",
       "<td>0.1396333</td>\n",
       "<td>0.2012239</td>\n",
       "<td>0.0049976</td>\n",
       "<td>0.9969062</td>\n",
       "<td>-95.0023798</td>\n",
       "<td>232.3020784</td>\n",
       "<td>0.7274747</td></tr>\n",
       "<tr><td>10</td>\n",
       "<td>0.4</td>\n",
       "<td>0.0222973</td>\n",
       "<td>0.0190386</td>\n",
       "<td>2.4970252</td>\n",
       "<td>0.0008</td>\n",
       "<td>0.0258261</td>\n",
       "<td>0.104925</td>\n",
       "<td>0.1573745</td>\n",
       "<td>0.0019039</td>\n",
       "<td>0.9988101</td>\n",
       "<td>-98.0961447</td>\n",
       "<td>149.7025226</td>\n",
       "<td>0.6250758</td></tr>\n",
       "<tr><td>11</td>\n",
       "<td>0.5</td>\n",
       "<td>0.0174501</td>\n",
       "<td>0.0118991</td>\n",
       "<td>2.0</td>\n",
       "<td>0.0005</td>\n",
       "<td>0.0196974</td>\n",
       "<td>0.08404</td>\n",
       "<td>0.1298390</td>\n",
       "<td>0.0011899</td>\n",
       "<td>1.0</td>\n",
       "<td>-98.8100904</td>\n",
       "<td>100.0</td>\n",
       "<td>0.5219316</td></tr>\n",
       "<tr><td>12</td>\n",
       "<td>0.6</td>\n",
       "<td>0.0139650</td>\n",
       "<td>0.0</td>\n",
       "<td>1.6666667</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0156240</td>\n",
       "<td>0.0700333</td>\n",
       "<td>0.1108032</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>66.6666667</td>\n",
       "<td>0.4175453</td></tr>\n",
       "<tr><td>13</td>\n",
       "<td>0.7</td>\n",
       "<td>0.0112073</td>\n",
       "<td>0.0</td>\n",
       "<td>1.4285714</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0125606</td>\n",
       "<td>0.0600286</td>\n",
       "<td>0.0967685</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>42.8571429</td>\n",
       "<td>0.3131589</td></tr>\n",
       "<tr><td>14</td>\n",
       "<td>0.8</td>\n",
       "<td>0.0085433</td>\n",
       "<td>0.0</td>\n",
       "<td>1.25</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0098787</td>\n",
       "<td>0.052525</td>\n",
       "<td>0.0859073</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>25.0</td>\n",
       "<td>0.2087726</td></tr>\n",
       "<tr><td>15</td>\n",
       "<td>0.9</td>\n",
       "<td>0.0058104</td>\n",
       "<td>0.0</td>\n",
       "<td>1.1111111</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0072949</td>\n",
       "<td>0.0466889</td>\n",
       "<td>0.0771726</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>11.1111111</td>\n",
       "<td>0.1043863</td></tr>\n",
       "<tr><td>16</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0000081</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0034672</td>\n",
       "<td>0.04202</td>\n",
       "<td>0.0698021</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>"
      ],
      "text/plain": [
       "ModelMetricsBinomial: drf\n",
       "** Reported on test data. **\n",
       "\n",
       "MSE: 0.013074625276969178\n",
       "RMSE: 0.11434432769914378\n",
       "LogLoss: 0.06041470000271687\n",
       "Mean Per-Class Error: 0.12998313353680432\n",
       "AUC: 0.9915474959363119\n",
       "AUCPR: 0.9025074939549784\n",
       "Gini: 0.9830949918726237\n",
       "\n",
       "Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.8134880277749358\n",
       "       0      1     Error    Rate\n",
       "-----  -----  ----  -------  -----------------\n",
       "0      95493  305   0.0032   (305.0/95798.0)\n",
       "1      1079   3123  0.2568   (1079.0/4202.0)\n",
       "Total  96572  3428  0.0138   (1384.0/100000.0)\n",
       "\n",
       "Maximum Metrics: Maximum metrics at their respective thresholds\n",
       "metric                       threshold    value     idx\n",
       "---------------------------  -----------  --------  -----\n",
       "max f1                       0.813488     0.818611  57\n",
       "max f2                       0.221646     0.850573  192\n",
       "max f0point5                 0.856862     0.874898  48\n",
       "max accuracy                 0.819578     0.98616   56\n",
       "max precision                0.999126     1         0\n",
       "max recall                   0.0200631    1         348\n",
       "max specificity              0.999126     1         0\n",
       "max absolute_mcc             0.813488     0.816041  57\n",
       "max min_per_class_accuracy   0.134704     0.956398  231\n",
       "max mean_per_class_accuracy  0.170096     0.959208  213\n",
       "max tns                      0.999126     95798     0\n",
       "max fns                      0.999126     3932      0\n",
       "max fps                      0.000184415  95798     399\n",
       "max tps                      0.0200631    4202      348\n",
       "max tnr                      0.999126     1         0\n",
       "max fnr                      0.999126     0.935745  0\n",
       "max fpr                      0.000184415  1         399\n",
       "max tpr                      0.0200631    1         348\n",
       "\n",
       "Gains/Lift Table: Avg response rate:  4.20 %, avg score:  6.98 %\n",
       "group    cumulative_data_fraction    lower_threshold    lift       cumulative_lift    response_rate    score       cumulative_response_rate    cumulative_score    capture_rate    cumulative_capture_rate    gain      cumulative_gain    kolmogorov_smirnov\n",
       "-------  --------------------------  -----------------  ---------  -----------------  ---------------  ----------  --------------------------  ------------------  --------------  -------------------------  --------  -----------------  --------------------\n",
       "1        0.01                        0.996685           23.7982    23.7982            1                0.998326    1                           0.998326            0.237982        0.237982                   2279.82   2279.82            0.237982\n",
       "2        0.02                        0.981659           23.703     23.7506            0.996            0.991479    0.998                       0.994902            0.23703         0.475012                   2270.3    2275.06            0.47497\n",
       "3        0.03                        0.876455           19.1337    22.2116            0.804            0.947253    0.933333                    0.979019            0.191337        0.666349                   1813.37   2121.16            0.664262\n",
       "4        0.04                        0.544317           10.8996    19.3836            0.458            0.754514    0.8145                      0.922893            0.108996        0.775345                   989.957   1838.36            0.7676\n",
       "5        0.05                        0.35525            10.6616    17.6392            0.448            0.433917    0.7412                      0.825098            0.106616        0.881961                   966.159   1663.92            0.868453\n",
       "6        0.1                         0.0998281          1.76583    9.70252            0.0742           0.17486     0.4077                      0.499979            0.0882913       0.970252                   76.5826   870.252            0.908424\n",
       "7        0.15                        0.0632485          0.328415   6.57782            0.0138           0.0786937   0.2764                      0.359551            0.0164208       0.986673                   -67.1585  557.782            0.873372\n",
       "8        0.2                         0.0467124          0.104712   4.95954            0.0044           0.0540732   0.2084                      0.283181            0.0052356       0.991909                   -89.5288  395.954            0.826644\n",
       "9        0.3                         0.0301801          0.0499762  3.32302            0.0021           0.0373094   0.139633                    0.201224            0.00499762      0.996906                   -95.0024  232.302            0.727475\n",
       "10       0.4                         0.0222973          0.0190386  2.49703            0.0008           0.0258261   0.104925                    0.157374            0.00190386      0.99881                    -98.0961  149.703            0.625076\n",
       "11       0.5                         0.0174501          0.0118991  2                  0.0005           0.0196974   0.08404                     0.129839            0.00118991      1                          -98.8101  100                0.521932\n",
       "12       0.6                         0.013965           0          1.66667            0                0.015624    0.0700333                   0.110803            0               1                          -100      66.6667            0.417545\n",
       "13       0.7                         0.0112073          0          1.42857            0                0.0125606   0.0600286                   0.0967685           0               1                          -100      42.8571            0.313159\n",
       "14       0.8                         0.00854333         0          1.25               0                0.00987867  0.052525                    0.0859073           0               1                          -100      25                 0.208773\n",
       "15       0.9                         0.00581039         0          1.11111            0                0.00729486  0.0466889                   0.0771726           0               1                          -100      11.1111            0.104386\n",
       "16       1                           8.14605e-06        0          1                  0                0.00346723  0.04202                     0.0698021           0               1                          -100      0                  0"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grid_search.models[0].model_performance(validation_h2o)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "f4c52ee3-e6d4-47fd-8bea-bb2a4d545993",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "drf Model Build progress: |██████████████████████████████████████████████████████| (done) 100%\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<pre style='margin: 1em 0 1em 0;'>Model Details\n",
       "=============\n",
       "H2ORandomForestEstimator : Distributed Random Forest\n",
       "Model Key: DRF_model_python_1680256840616_167\n",
       "</pre>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-21.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-21 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-21 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-21 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-21 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-21 .h2o-table th,\n",
       "#h2o-table-21 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-21 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-21\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Model Summary: </caption>\n",
       "    <thead><tr><th></th>\n",
       "<th>number_of_trees</th>\n",
       "<th>number_of_internal_trees</th>\n",
       "<th>model_size_in_bytes</th>\n",
       "<th>min_depth</th>\n",
       "<th>max_depth</th>\n",
       "<th>mean_depth</th>\n",
       "<th>min_leaves</th>\n",
       "<th>max_leaves</th>\n",
       "<th>mean_leaves</th></tr></thead>\n",
       "    <tbody><tr><td></td>\n",
       "<td>180.0</td>\n",
       "<td>180.0</td>\n",
       "<td>2819223.0</td>\n",
       "<td>20.0</td>\n",
       "<td>20.0</td>\n",
       "<td>20.0</td>\n",
       "<td>634.0</td>\n",
       "<td>2023.0</td>\n",
       "<td>1239.6833</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>\n",
       "<div style='margin: 1em 0 1em 0;'><pre style='margin: 1em 0 1em 0;'>ModelMetricsBinomial: drf\n",
       "** Reported on train data. **\n",
       "\n",
       "MSE: 0.008868154883546902\n",
       "RMSE: 0.09417088129324745\n",
       "LogLoss: 0.039011034569354765\n",
       "Mean Per-Class Error: 0.021732389113083696\n",
       "AUC: 0.9981431577200072\n",
       "AUCPR: 0.9927655561977535\n",
       "Gini: 0.9962863154400143</pre>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-22.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-22 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-22 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-22 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-22 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-22 .h2o-table th,\n",
       "#h2o-table-22 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-22 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-22\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.15699625656024563</caption>\n",
       "    <thead><tr><th></th>\n",
       "<th>0</th>\n",
       "<th>1</th>\n",
       "<th>Error</th>\n",
       "<th>Rate</th></tr></thead>\n",
       "    <tbody><tr><td>0</td>\n",
       "<td>194239.0</td>\n",
       "<td>400.0</td>\n",
       "<td>0.0021</td>\n",
       "<td> (400.0/194639.0)</td></tr>\n",
       "<tr><td>1</td>\n",
       "<td>1081.0</td>\n",
       "<td>25024.0</td>\n",
       "<td>0.0414</td>\n",
       "<td> (1081.0/26105.0)</td></tr>\n",
       "<tr><td>Total</td>\n",
       "<td>195320.0</td>\n",
       "<td>25424.0</td>\n",
       "<td>0.0067</td>\n",
       "<td> (1481.0/220744.0)</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-23.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-23 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-23 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-23 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-23 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-23 .h2o-table th,\n",
       "#h2o-table-23 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-23 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-23\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Maximum Metrics: Maximum metrics at their respective thresholds</caption>\n",
       "    <thead><tr><th>metric</th>\n",
       "<th>threshold</th>\n",
       "<th>value</th>\n",
       "<th>idx</th></tr></thead>\n",
       "    <tbody><tr><td>max f1</td>\n",
       "<td>0.1569963</td>\n",
       "<td>0.9712589</td>\n",
       "<td>238.0</td></tr>\n",
       "<tr><td>max f2</td>\n",
       "<td>0.1102345</td>\n",
       "<td>0.9684029</td>\n",
       "<td>255.0</td></tr>\n",
       "<tr><td>max f0point5</td>\n",
       "<td>0.2557006</td>\n",
       "<td>0.9838297</td>\n",
       "<td>211.0</td></tr>\n",
       "<tr><td>max accuracy</td>\n",
       "<td>0.1569963</td>\n",
       "<td>0.9932909</td>\n",
       "<td>238.0</td></tr>\n",
       "<tr><td>max precision</td>\n",
       "<td>0.9999928</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max recall</td>\n",
       "<td>0.0082992</td>\n",
       "<td>1.0</td>\n",
       "<td>372.0</td></tr>\n",
       "<tr><td>max specificity</td>\n",
       "<td>0.9999928</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max absolute_mcc</td>\n",
       "<td>0.1569963</td>\n",
       "<td>0.9675701</td>\n",
       "<td>238.0</td></tr>\n",
       "<tr><td>max min_per_class_accuracy</td>\n",
       "<td>0.0661280</td>\n",
       "<td>0.9830096</td>\n",
       "<td>280.0</td></tr>\n",
       "<tr><td>max mean_per_class_accuracy</td>\n",
       "<td>0.0799662</td>\n",
       "<td>0.9839688</td>\n",
       "<td>270.0</td></tr>\n",
       "<tr><td>max tns</td>\n",
       "<td>0.9999928</td>\n",
       "<td>194639.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fns</td>\n",
       "<td>0.9999928</td>\n",
       "<td>14318.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fps</td>\n",
       "<td>0.0000629</td>\n",
       "<td>194639.0</td>\n",
       "<td>399.0</td></tr>\n",
       "<tr><td>max tps</td>\n",
       "<td>0.0082992</td>\n",
       "<td>26105.0</td>\n",
       "<td>372.0</td></tr>\n",
       "<tr><td>max tnr</td>\n",
       "<td>0.9999928</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fnr</td>\n",
       "<td>0.9999928</td>\n",
       "<td>0.5484773</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fpr</td>\n",
       "<td>0.0000629</td>\n",
       "<td>1.0</td>\n",
       "<td>399.0</td></tr>\n",
       "<tr><td>max tpr</td>\n",
       "<td>0.0082992</td>\n",
       "<td>1.0</td>\n",
       "<td>372.0</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-24.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-24 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-24 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-24 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-24 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-24 .h2o-table th,\n",
       "#h2o-table-24 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-24 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-24\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Gains/Lift Table: Avg response rate: 11.83 %, avg score: 11.79 %</caption>\n",
       "    <thead><tr><th>group</th>\n",
       "<th>cumulative_data_fraction</th>\n",
       "<th>lower_threshold</th>\n",
       "<th>lift</th>\n",
       "<th>cumulative_lift</th>\n",
       "<th>response_rate</th>\n",
       "<th>score</th>\n",
       "<th>cumulative_response_rate</th>\n",
       "<th>cumulative_score</th>\n",
       "<th>capture_rate</th>\n",
       "<th>cumulative_capture_rate</th>\n",
       "<th>gain</th>\n",
       "<th>cumulative_gain</th>\n",
       "<th>kolmogorov_smirnov</th></tr></thead>\n",
       "    <tbody><tr><td>1</td>\n",
       "<td>0.0503343</td>\n",
       "<td>1.0</td>\n",
       "<td>8.4560046</td>\n",
       "<td>8.4560046</td>\n",
       "<td>1.0</td>\n",
       "<td>1.0</td>\n",
       "<td>1.0</td>\n",
       "<td>1.0</td>\n",
       "<td>0.4256273</td>\n",
       "<td>0.4256273</td>\n",
       "<td>745.6004597</td>\n",
       "<td>745.6004597</td>\n",
       "<td>0.4256273</td></tr>\n",
       "<tr><td>2</td>\n",
       "<td>0.1000027</td>\n",
       "<td>0.7891256</td>\n",
       "<td>8.4552333</td>\n",
       "<td>8.4556215</td>\n",
       "<td>0.9999088</td>\n",
       "<td>0.9676415</td>\n",
       "<td>0.9999547</td>\n",
       "<td>0.9839285</td>\n",
       "<td>0.4199579</td>\n",
       "<td>0.8455851</td>\n",
       "<td>745.5233345</td>\n",
       "<td>745.5621539</td>\n",
       "<td>0.8455800</td></tr>\n",
       "<tr><td>3</td>\n",
       "<td>0.1500018</td>\n",
       "<td>0.0451140</td>\n",
       "<td>2.8791941</td>\n",
       "<td>6.5968685</td>\n",
       "<td>0.3404911</td>\n",
       "<td>0.1874717</td>\n",
       "<td>0.7801401</td>\n",
       "<td>0.7184509</td>\n",
       "<td>0.1439571</td>\n",
       "<td>0.9895422</td>\n",
       "<td>187.9194099</td>\n",
       "<td>559.6868529</td>\n",
       "<td>0.9521397</td></tr>\n",
       "<tr><td>4</td>\n",
       "<td>0.2000009</td>\n",
       "<td>0.0284768</td>\n",
       "<td>0.0942365</td>\n",
       "<td>4.9712474</td>\n",
       "<td>0.0111443</td>\n",
       "<td>0.0348379</td>\n",
       "<td>0.5878955</td>\n",
       "<td>0.5475515</td>\n",
       "<td>0.0047117</td>\n",
       "<td>0.9942540</td>\n",
       "<td>-90.5763471</td>\n",
       "<td>397.1247351</td>\n",
       "<td>0.9007784</td></tr>\n",
       "<tr><td>5</td>\n",
       "<td>0.2999991</td>\n",
       "<td>0.0180366</td>\n",
       "<td>0.0340937</td>\n",
       "<td>3.3255543</td>\n",
       "<td>0.0040319</td>\n",
       "<td>0.0222708</td>\n",
       "<td>0.3932773</td>\n",
       "<td>0.3724606</td>\n",
       "<td>0.0034093</td>\n",
       "<td>0.9976633</td>\n",
       "<td>-96.5906297</td>\n",
       "<td>232.5554320</td>\n",
       "<td>0.7912350</td></tr>\n",
       "<tr><td>6</td>\n",
       "<td>0.4000018</td>\n",
       "<td>0.0141939</td>\n",
       "<td>0.0134070</td>\n",
       "<td>2.4974987</td>\n",
       "<td>0.0015855</td>\n",
       "<td>0.0159008</td>\n",
       "<td>0.2953521</td>\n",
       "<td>0.2833186</td>\n",
       "<td>0.0013407</td>\n",
       "<td>0.9990040</td>\n",
       "<td>-98.6592971</td>\n",
       "<td>149.7498742</td>\n",
       "<td>0.6793404</td></tr>\n",
       "<tr><td>7</td>\n",
       "<td>0.5</td>\n",
       "<td>0.0117178</td>\n",
       "<td>0.0080446</td>\n",
       "<td>1.9996169</td>\n",
       "<td>0.0009513</td>\n",
       "<td>0.0128952</td>\n",
       "<td>0.2364730</td>\n",
       "<td>0.2292349</td>\n",
       "<td>0.0008044</td>\n",
       "<td>0.9998085</td>\n",
       "<td>-99.1955418</td>\n",
       "<td>99.9616932</td>\n",
       "<td>0.5668428</td></tr>\n",
       "<tr><td>8</td>\n",
       "<td>0.5999982</td>\n",
       "<td>0.0097511</td>\n",
       "<td>0.0015323</td>\n",
       "<td>1.6666079</td>\n",
       "<td>0.0001812</td>\n",
       "<td>0.0107103</td>\n",
       "<td>0.1970916</td>\n",
       "<td>0.1928147</td>\n",
       "<td>0.0001532</td>\n",
       "<td>0.9999617</td>\n",
       "<td>-99.8467699</td>\n",
       "<td>66.6607855</td>\n",
       "<td>0.4536066</td></tr>\n",
       "<tr><td>9</td>\n",
       "<td>0.7000009</td>\n",
       "<td>0.0079907</td>\n",
       "<td>0.0003831</td>\n",
       "<td>1.4285696</td>\n",
       "<td>0.0000453</td>\n",
       "<td>0.0088535</td>\n",
       "<td>0.1689414</td>\n",
       "<td>0.1665338</td>\n",
       "<td>0.0000383</td>\n",
       "<td>1.0</td>\n",
       "<td>-99.9616942</td>\n",
       "<td>42.8569580</td>\n",
       "<td>0.3402350</td></tr>\n",
       "<tr><td>10</td>\n",
       "<td>0.7999991</td>\n",
       "<td>0.0060763</td>\n",
       "<td>0.0</td>\n",
       "<td>1.2500014</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0070966</td>\n",
       "<td>0.1478241</td>\n",
       "<td>0.1466045</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>25.0001416</td>\n",
       "<td>0.2268250</td></tr>\n",
       "<tr><td>11</td>\n",
       "<td>0.8999973</td>\n",
       "<td>0.0029329</td>\n",
       "<td>0.0</td>\n",
       "<td>1.1111145</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0046032</td>\n",
       "<td>0.1313995</td>\n",
       "<td>0.1308268</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>11.1114467</td>\n",
       "<td>0.1134151</td></tr>\n",
       "<tr><td>12</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0015302</td>\n",
       "<td>0.1182592</td>\n",
       "<td>0.1178968</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div></div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-25.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-25 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-25 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-25 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-25 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-25 .h2o-table th,\n",
       "#h2o-table-25 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-25 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-25\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Scoring History: </caption>\n",
       "    <thead><tr><th></th>\n",
       "<th>timestamp</th>\n",
       "<th>duration</th>\n",
       "<th>number_of_trees</th>\n",
       "<th>training_rmse</th>\n",
       "<th>training_logloss</th>\n",
       "<th>training_auc</th>\n",
       "<th>training_pr_auc</th>\n",
       "<th>training_lift</th>\n",
       "<th>training_classification_error</th></tr></thead>\n",
       "    <tbody><tr><td></td>\n",
       "<td>2023-03-31 12:14:48</td>\n",
       "<td> 0.002 sec</td>\n",
       "<td>0.0</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:14:48</td>\n",
       "<td> 0.105 sec</td>\n",
       "<td>1.0</td>\n",
       "<td>0.1235661</td>\n",
       "<td>0.1800514</td>\n",
       "<td>0.9681383</td>\n",
       "<td>0.9256121</td>\n",
       "<td>8.3093228</td>\n",
       "<td>0.0161142</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:14:48</td>\n",
       "<td> 0.203 sec</td>\n",
       "<td>2.0</td>\n",
       "<td>0.1193382</td>\n",
       "<td>0.1683478</td>\n",
       "<td>0.9706427</td>\n",
       "<td>0.9334365</td>\n",
       "<td>8.3172524</td>\n",
       "<td>0.0152059</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:14:48</td>\n",
       "<td> 0.312 sec</td>\n",
       "<td>3.0</td>\n",
       "<td>0.1195042</td>\n",
       "<td>0.1678201</td>\n",
       "<td>0.9721680</td>\n",
       "<td>0.9376626</td>\n",
       "<td>8.3220265</td>\n",
       "<td>0.0156991</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:14:48</td>\n",
       "<td> 0.420 sec</td>\n",
       "<td>4.0</td>\n",
       "<td>0.1176534</td>\n",
       "<td>0.1415564</td>\n",
       "<td>0.9744814</td>\n",
       "<td>0.9439173</td>\n",
       "<td>8.3539680</td>\n",
       "<td>0.0153424</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:14:48</td>\n",
       "<td> 0.538 sec</td>\n",
       "<td>5.0</td>\n",
       "<td>0.1156003</td>\n",
       "<td>0.1226949</td>\n",
       "<td>0.9779051</td>\n",
       "<td>0.9499961</td>\n",
       "<td>8.3722567</td>\n",
       "<td>0.0148612</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:14:48</td>\n",
       "<td> 0.659 sec</td>\n",
       "<td>6.0</td>\n",
       "<td>0.1130767</td>\n",
       "<td>0.1045857</td>\n",
       "<td>0.9805319</td>\n",
       "<td>0.9562190</td>\n",
       "<td>8.3951015</td>\n",
       "<td>0.0142112</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:14:48</td>\n",
       "<td> 0.778 sec</td>\n",
       "<td>7.0</td>\n",
       "<td>0.1112375</td>\n",
       "<td>0.0905580</td>\n",
       "<td>0.9825024</td>\n",
       "<td>0.9605478</td>\n",
       "<td>8.4127054</td>\n",
       "<td>0.0136941</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:14:49</td>\n",
       "<td> 0.895 sec</td>\n",
       "<td>8.0</td>\n",
       "<td>0.1084148</td>\n",
       "<td>0.0787858</td>\n",
       "<td>0.9847010</td>\n",
       "<td>0.9652716</td>\n",
       "<td>8.4246317</td>\n",
       "<td>0.0128501</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:14:49</td>\n",
       "<td> 1.011 sec</td>\n",
       "<td>9.0</td>\n",
       "<td>0.1067728</td>\n",
       "<td>0.0714506</td>\n",
       "<td>0.9864670</td>\n",
       "<td>0.9686248</td>\n",
       "<td>8.4315856</td>\n",
       "<td>0.0123085</td></tr>\n",
       "<tr><td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:14:51</td>\n",
       "<td> 3.439 sec</td>\n",
       "<td>30.0</td>\n",
       "<td>0.0964305</td>\n",
       "<td>0.0405967</td>\n",
       "<td>0.9958964</td>\n",
       "<td>0.9868000</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0085257</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:14:51</td>\n",
       "<td> 3.540 sec</td>\n",
       "<td>31.0</td>\n",
       "<td>0.0966603</td>\n",
       "<td>0.0406562</td>\n",
       "<td>0.9959321</td>\n",
       "<td>0.9868023</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0085212</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:14:51</td>\n",
       "<td> 3.650 sec</td>\n",
       "<td>32.0</td>\n",
       "<td>0.0970029</td>\n",
       "<td>0.0409224</td>\n",
       "<td>0.9959519</td>\n",
       "<td>0.9868506</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0085484</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:14:51</td>\n",
       "<td> 3.759 sec</td>\n",
       "<td>33.0</td>\n",
       "<td>0.0969291</td>\n",
       "<td>0.0409023</td>\n",
       "<td>0.9960233</td>\n",
       "<td>0.9869866</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0085167</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:14:52</td>\n",
       "<td> 3.876 sec</td>\n",
       "<td>34.0</td>\n",
       "<td>0.0965990</td>\n",
       "<td>0.0406116</td>\n",
       "<td>0.9961272</td>\n",
       "<td>0.9873055</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0084532</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:14:52</td>\n",
       "<td> 3.997 sec</td>\n",
       "<td>35.0</td>\n",
       "<td>0.0961825</td>\n",
       "<td>0.0402895</td>\n",
       "<td>0.9961986</td>\n",
       "<td>0.9876011</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0083083</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:14:56</td>\n",
       "<td> 8.029 sec</td>\n",
       "<td>82.0</td>\n",
       "<td>0.0949094</td>\n",
       "<td>0.0395490</td>\n",
       "<td>0.9974679</td>\n",
       "<td>0.9912040</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0072754</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:15:00</td>\n",
       "<td>12.091 sec</td>\n",
       "<td>129.0</td>\n",
       "<td>0.0944533</td>\n",
       "<td>0.0392170</td>\n",
       "<td>0.9978965</td>\n",
       "<td>0.9922045</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0068949</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:15:04</td>\n",
       "<td>16.140 sec</td>\n",
       "<td>175.0</td>\n",
       "<td>0.0941779</td>\n",
       "<td>0.0390152</td>\n",
       "<td>0.9981490</td>\n",
       "<td>0.9927645</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0066955</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:15:04</td>\n",
       "<td>16.601 sec</td>\n",
       "<td>180.0</td>\n",
       "<td>0.0941709</td>\n",
       "<td>0.0390110</td>\n",
       "<td>0.9981432</td>\n",
       "<td>0.9927656</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0067091</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "<pre style='font-size: smaller; margin-bottom: 1em;'>[40 rows x 10 columns]</pre></div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-26.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-26 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-26 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-26 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-26 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-26 .h2o-table th,\n",
       "#h2o-table-26 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-26 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-26\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Variable Importances: </caption>\n",
       "    <thead><tr><th>variable</th>\n",
       "<th>relative_importance</th>\n",
       "<th>scaled_importance</th>\n",
       "<th>percentage</th></tr></thead>\n",
       "    <tbody><tr><td>RollingMean_400</td>\n",
       "<td>420454.5625000</td>\n",
       "<td>1.0</td>\n",
       "<td>0.1662408</td></tr>\n",
       "<tr><td>RollingMean_800</td>\n",
       "<td>413218.9062500</td>\n",
       "<td>0.9827909</td>\n",
       "<td>0.1633800</td></tr>\n",
       "<tr><td>RollingSD_800</td>\n",
       "<td>372931.0937500</td>\n",
       "<td>0.8869712</td>\n",
       "<td>0.1474508</td></tr>\n",
       "<tr><td>RollingMean_200</td>\n",
       "<td>232538.1875000</td>\n",
       "<td>0.5530638</td>\n",
       "<td>0.0919418</td></tr>\n",
       "<tr><td>RollingSD_400</td>\n",
       "<td>211802.5468750</td>\n",
       "<td>0.5037466</td>\n",
       "<td>0.0837432</td></tr>\n",
       "<tr><td>RollingSD_200</td>\n",
       "<td>114920.9218750</td>\n",
       "<td>0.2733254</td>\n",
       "<td>0.0454378</td></tr>\n",
       "<tr><td>RollingMean_100</td>\n",
       "<td>79479.3671875</td>\n",
       "<td>0.1890320</td>\n",
       "<td>0.0314248</td></tr>\n",
       "<tr><td>RollingSD_100</td>\n",
       "<td>64288.7734375</td>\n",
       "<td>0.1529030</td>\n",
       "<td>0.0254187</td></tr>\n",
       "<tr><td>meanaddsd</td>\n",
       "<td>54571.5</td>\n",
       "<td>0.1297917</td>\n",
       "<td>0.0215767</td></tr>\n",
       "<tr><td>meandiffsd</td>\n",
       "<td>46951.3515625</td>\n",
       "<td>0.1116681</td>\n",
       "<td>0.0185638</td></tr>\n",
       "<tr><td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td></tr>\n",
       "<tr><td>RollingMean_3</td>\n",
       "<td>17191.9765625</td>\n",
       "<td>0.0408890</td>\n",
       "<td>0.0067974</td></tr>\n",
       "<tr><td>RollingSD_5</td>\n",
       "<td>16108.4423828</td>\n",
       "<td>0.0383120</td>\n",
       "<td>0.0063690</td></tr>\n",
       "<tr><td>diff50</td>\n",
       "<td>12328.0527344</td>\n",
       "<td>0.0293208</td>\n",
       "<td>0.0048743</td></tr>\n",
       "<tr><td>RollingSD_4</td>\n",
       "<td>10791.5673828</td>\n",
       "<td>0.0256664</td>\n",
       "<td>0.0042668</td></tr>\n",
       "<tr><td>diff30</td>\n",
       "<td>10252.0791016</td>\n",
       "<td>0.0243833</td>\n",
       "<td>0.0040535</td></tr>\n",
       "<tr><td>RollingSD_3</td>\n",
       "<td>8212.7656250</td>\n",
       "<td>0.0195331</td>\n",
       "<td>0.0032472</td></tr>\n",
       "<tr><td>diff5</td>\n",
       "<td>6854.6352539</td>\n",
       "<td>0.0163029</td>\n",
       "<td>0.0027102</td></tr>\n",
       "<tr><td>diff2</td>\n",
       "<td>4912.2934570</td>\n",
       "<td>0.0116833</td>\n",
       "<td>0.0019422</td></tr>\n",
       "<tr><td>RollingSD_2</td>\n",
       "<td>4508.2905273</td>\n",
       "<td>0.0107224</td>\n",
       "<td>0.0017825</td></tr>\n",
       "<tr><td>diff1</td>\n",
       "<td>3924.4677734</td>\n",
       "<td>0.0093339</td>\n",
       "<td>0.0015517</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "<pre style='font-size: smaller; margin-bottom: 1em;'>[36 rows x 4 columns]</pre></div><pre style=\"font-size: smaller; margin: 1em 0 0 0;\">\n",
       "\n",
       "[tips]\n",
       "Use `model.explain()` to inspect the model.\n",
       "--\n",
       "Use `h2o.display.toggle_user_tips()` to switch on/off this section.</pre>"
      ],
      "text/plain": [
       "Model Details\n",
       "=============\n",
       "H2ORandomForestEstimator : Distributed Random Forest\n",
       "Model Key: DRF_model_python_1680256840616_167\n",
       "\n",
       "\n",
       "Model Summary: \n",
       "    number_of_trees    number_of_internal_trees    model_size_in_bytes    min_depth    max_depth    mean_depth    min_leaves    max_leaves    mean_leaves\n",
       "--  -----------------  --------------------------  ---------------------  -----------  -----------  ------------  ------------  ------------  -------------\n",
       "    180                180                         2.81922e+06            20           20           20            634           2023          1239.68\n",
       "\n",
       "ModelMetricsBinomial: drf\n",
       "** Reported on train data. **\n",
       "\n",
       "MSE: 0.008868154883546902\n",
       "RMSE: 0.09417088129324745\n",
       "LogLoss: 0.039011034569354765\n",
       "Mean Per-Class Error: 0.021732389113083696\n",
       "AUC: 0.9981431577200072\n",
       "AUCPR: 0.9927655561977535\n",
       "Gini: 0.9962863154400143\n",
       "\n",
       "Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.15699625656024563\n",
       "       0       1      Error    Rate\n",
       "-----  ------  -----  -------  -----------------\n",
       "0      194239  400    0.0021   (400.0/194639.0)\n",
       "1      1081    25024  0.0414   (1081.0/26105.0)\n",
       "Total  195320  25424  0.0067   (1481.0/220744.0)\n",
       "\n",
       "Maximum Metrics: Maximum metrics at their respective thresholds\n",
       "metric                       threshold    value     idx\n",
       "---------------------------  -----------  --------  -----\n",
       "max f1                       0.156996     0.971259  238\n",
       "max f2                       0.110235     0.968403  255\n",
       "max f0point5                 0.255701     0.98383   211\n",
       "max accuracy                 0.156996     0.993291  238\n",
       "max precision                0.999993     1         0\n",
       "max recall                   0.00829923   1         372\n",
       "max specificity              0.999993     1         0\n",
       "max absolute_mcc             0.156996     0.96757   238\n",
       "max min_per_class_accuracy   0.066128     0.98301   280\n",
       "max mean_per_class_accuracy  0.0799662    0.983969  270\n",
       "max tns                      0.999993     194639    0\n",
       "max fns                      0.999993     14318     0\n",
       "max fps                      6.28685e-05  194639    399\n",
       "max tps                      0.00829923   26105     372\n",
       "max tnr                      0.999993     1         0\n",
       "max fnr                      0.999993     0.548477  0\n",
       "max fpr                      6.28685e-05  1         399\n",
       "max tpr                      0.00829923   1         372\n",
       "\n",
       "Gains/Lift Table: Avg response rate: 11.83 %, avg score: 11.79 %\n",
       "group    cumulative_data_fraction    lower_threshold    lift         cumulative_lift    response_rate    score       cumulative_response_rate    cumulative_score    capture_rate    cumulative_capture_rate    gain      cumulative_gain    kolmogorov_smirnov\n",
       "-------  --------------------------  -----------------  -----------  -----------------  ---------------  ----------  --------------------------  ------------------  --------------  -------------------------  --------  -----------------  --------------------\n",
       "1        0.0503343                   1                  8.456        8.456              1                1           1                           1                   0.425627        0.425627                   745.6     745.6              0.425627\n",
       "2        0.100003                    0.789126           8.45523      8.45562            0.999909         0.967642    0.999955                    0.983929            0.419958        0.845585                   745.523   745.562            0.84558\n",
       "3        0.150002                    0.045114           2.87919      6.59687            0.340491         0.187472    0.78014                     0.718451            0.143957        0.989542                   187.919   559.687            0.95214\n",
       "4        0.200001                    0.0284768          0.0942365    4.97125            0.0111443        0.0348379   0.587896                    0.547552            0.00471174      0.994254                   -90.5763  397.125            0.900778\n",
       "5        0.299999                    0.0180366          0.0340937    3.32555            0.00403189       0.0222708   0.393277                    0.372461            0.00340931      0.997663                   -96.5906  232.555            0.791235\n",
       "6        0.400002                    0.0141939          0.013407     2.4975             0.0015855        0.0159008   0.295352                    0.283319            0.00134074      0.999004                   -98.6593  149.75             0.67934\n",
       "7        0.5                         0.0117178          0.00804458   1.99962            0.000951345      0.0128952   0.236473                    0.229235            0.000804444     0.999808                   -99.1955  99.9617            0.566843\n",
       "8        0.599998                    0.00975107         0.0015323    1.66661            0.000181209      0.0107103   0.197092                    0.192815            0.000153227     0.999962                   -99.8468  66.6608            0.453607\n",
       "9        0.700001                    0.00799073         0.000383058  1.42857            4.53001e-05      0.00885349  0.168941                    0.166534            3.83068e-05     1                          -99.9617  42.857             0.340235\n",
       "10       0.799999                    0.00607634         0            1.25               0                0.00709662  0.147824                    0.146605            0               1                          -100      25.0001            0.226825\n",
       "11       0.899997                    0.00293289         0            1.11111            0                0.00460324  0.131399                    0.130827            0               1                          -100      11.1114            0.113415\n",
       "12       1                           0                  0            1                  0                0.00153023  0.118259                    0.117897            0               1                          -100      0                  0\n",
       "\n",
       "Scoring History: \n",
       "     timestamp            duration    number_of_trees    training_rmse        training_logloss      training_auc        training_pr_auc     training_lift      training_classification_error\n",
       "---  -------------------  ----------  -----------------  -------------------  --------------------  ------------------  ------------------  -----------------  -------------------------------\n",
       "     2023-03-31 12:14:48  0.002 sec   0.0                nan                  nan                   nan                 nan                 nan                nan\n",
       "     2023-03-31 12:14:48  0.105 sec   1.0                0.12356614338361153  0.18005137078994138   0.9681383397709885  0.9256120770981746  8.309322848071183  0.01611416350160404\n",
       "     2023-03-31 12:14:48  0.203 sec   2.0                0.11933823108044307  0.1683478121611553    0.9706427199134072  0.9334364698167581  8.317252443433322  0.015205926104007332\n",
       "     2023-03-31 12:14:48  0.312 sec   3.0                0.11950421657030527  0.1678200919736248    0.972167951428894   0.9376626101186627  8.322026471056695  0.01569905213270142\n",
       "     2023-03-31 12:14:48  0.420 sec   4.0                0.11765335599062994  0.14155635518622003   0.9744813890494384  0.9439172900498839  8.35396796895118   0.015342377260981912\n",
       "     2023-03-31 12:14:48  0.538 sec   5.0                0.11560025696853495  0.12269485875279688   0.9779051304823273  0.949996083425019   8.372256652134135  0.014861226441206814\n",
       "     2023-03-31 12:14:48  0.659 sec   6.0                0.11307670720428194  0.10458573518914849   0.9805319172701419  0.9562189939585805  8.395101476990419  0.01421122386789076\n",
       "     2023-03-31 12:14:48  0.778 sec   7.0                0.11123754900432054  0.09055795655104677   0.9825024001121536  0.9605477846049894  8.412705408990815  0.013694103652233022\n",
       "     2023-03-31 12:14:49  0.895 sec   8.0                0.10841484980313712  0.0787858007415495    0.9847009858288808  0.9652716213109077  8.42463173597659   0.012850065528363092\n",
       "     2023-03-31 12:14:49  1.011 sec   9.0                0.10677279338990388  0.07145057752240268   0.9864669526655273  0.9686248026178392  8.43158561237472   0.012308499910207353\n",
       "---  ---                  ---         ---                ---                  ---                   ---                 ---                 ---                ---\n",
       "     2023-03-31 12:14:51  3.439 sec   30.0               0.09643047450611354  0.04059673529417404   0.9958963918862147  0.9867999932825667  8.456004596820533  0.008525713043163121\n",
       "     2023-03-31 12:14:51  3.540 sec   31.0               0.09666028221918359  0.0406561962210137    0.995932139411009   0.9868022824416514  8.456004596820533  0.00852118290870873\n",
       "     2023-03-31 12:14:51  3.650 sec   32.0               0.09700292222595112  0.040922414023576915  0.9959519139612195  0.9868505905932174  8.456004596820533  0.008548363715435074\n",
       "     2023-03-31 12:14:51  3.759 sec   33.0               0.09692909935555034  0.04090229117173476   0.9960232858079534  0.9869866268348438  8.456004596820533  0.008516652774254339\n",
       "     2023-03-31 12:14:52  3.876 sec   34.0               0.09659904317994042  0.04061157999911285   0.996127176418406   0.9873055202659694  8.456004596820533  0.008453230891892871\n",
       "     2023-03-31 12:14:52  3.997 sec   35.0               0.09618245919407442  0.04028945009893666   0.9961985643051283  0.987601085220849   8.456004596820533  0.008308266589352372\n",
       "     2023-03-31 12:14:56  8.029 sec   82.0               0.09490935064348244  0.0395489735796184    0.9974678521708509  0.9912039821357801  8.456004596820533  0.007275395933751314\n",
       "     2023-03-31 12:15:00  12.091 sec  129.0              0.09445330237215648  0.03921697966536843   0.9978964673253301  0.9922044623222084  8.456004596820533  0.006894864639582503\n",
       "     2023-03-31 12:15:04  16.140 sec  175.0              0.09417787325487154  0.039015164466589716  0.9981490478398742  0.9927644638112433  8.456004596820533  0.006695538723589316\n",
       "     2023-03-31 12:15:04  16.601 sec  180.0              0.09417088129324745  0.039011034569354765  0.9981431577200072  0.9927655561977535  8.456004596820533  0.006709129126952488\n",
       "[40 rows x 10 columns]\n",
       "\n",
       "\n",
       "Variable Importances: \n",
       "variable         relative_importance    scaled_importance     percentage\n",
       "---------------  ---------------------  --------------------  ---------------------\n",
       "RollingMean_400  420454.5625            1.0                   0.16624081359235682\n",
       "RollingMean_800  413218.90625           0.982790872319289     0.16337995420550067\n",
       "RollingSD_800    372931.09375           0.8869712140417075    0.14745081625529394\n",
       "RollingMean_200  232538.1875            0.553063774875793     0.09194177190381189\n",
       "RollingSD_400    211802.546875          0.5037465775508144    0.08374324089641266\n",
       "RollingSD_200    114920.921875          0.27332542473005034   0.04543784098260006\n",
       "RollingMean_100  79479.3671875          0.18903200078248644   0.03142483360507158\n",
       "RollingSD_100    64288.7734375          0.15290302251744503   0.02541872286403052\n",
       "meanaddsd        54571.5                0.1297916704138512    0.02157667288710965\n",
       "meandiffsd       46951.3515625          0.11166807486480777   0.018563791617817856\n",
       "---              ---                    ---                   ---\n",
       "RollingMean_3    17191.9765625          0.04088902368017472   0.0067974245635893885\n",
       "RollingSD_5      16108.4423828125       0.03831196951944718   0.006369012983238475\n",
       "diff50           12328.052734375        0.029320772882265964  0.004874309139104607\n",
       "RollingSD_4      10791.5673828125       0.02566642949156367   0.004266808120688406\n",
       "diff30           10252.0791015625       0.024383322280068016  0.004053503333923148\n",
       "RollingSD_3      8212.765625            0.01953306339730824   0.0032471923511196072\n",
       "diff5            6854.63525390625       0.0163029156186318    0.0027102099563688915\n",
       "diff2            4912.29345703125       0.011683292072805726  0.001942239979620357\n",
       "RollingSD_2      4508.29052734375       0.010722420278989718  0.0017825038708584363\n",
       "diff1            3924.4677734375        0.009333868920586396  0.0015516699633226956\n",
       "[36 rows x 4 columns]\n",
       "\n",
       "\n",
       "[tips]\n",
       "Use `model.explain()` to inspect the model.\n",
       "--\n",
       "Use `h2o.display.toggle_user_tips()` to switch on/off this section."
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf_next_model = H2ORandomForestEstimator(seed=1,mtries=5,max_depth=None,ntrees=180)\n",
    "rf_next_model.train(x=predictors, y=response, training_frame=train_h2o)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "6084e106-ab6e-4853-9cc0-810962ce101e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<pre style='margin: 1em 0 1em 0;'>ModelMetricsBinomial: drf\n",
       "** Reported on test data. **\n",
       "\n",
       "MSE: 0.013103536045933374\n",
       "RMSE: 0.11447067766870857\n",
       "LogLoss: 0.06021293925083577\n",
       "Mean Per-Class Error: 0.12964287191678184\n",
       "AUC: 0.9912719814049472\n",
       "AUCPR: 0.9029267705830357\n",
       "Gini: 0.9825439628098944</pre>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-27.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-27 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-27 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-27 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-27 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-27 .h2o-table th,\n",
       "#h2o-table-27 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-27 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-27\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.7959731126734471</caption>\n",
       "    <thead><tr><th></th>\n",
       "<th>0</th>\n",
       "<th>1</th>\n",
       "<th>Error</th>\n",
       "<th>Rate</th></tr></thead>\n",
       "    <tbody><tr><td>0</td>\n",
       "<td>95467.0</td>\n",
       "<td>331.0</td>\n",
       "<td>0.0035</td>\n",
       "<td> (331.0/95798.0)</td></tr>\n",
       "<tr><td>1</td>\n",
       "<td>1075.0</td>\n",
       "<td>3127.0</td>\n",
       "<td>0.2558</td>\n",
       "<td> (1075.0/4202.0)</td></tr>\n",
       "<tr><td>Total</td>\n",
       "<td>96542.0</td>\n",
       "<td>3458.0</td>\n",
       "<td>0.0141</td>\n",
       "<td> (1406.0/100000.0)</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-28.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-28 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-28 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-28 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-28 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-28 .h2o-table th,\n",
       "#h2o-table-28 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-28 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-28\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Maximum Metrics: Maximum metrics at their respective thresholds</caption>\n",
       "    <thead><tr><th>metric</th>\n",
       "<th>threshold</th>\n",
       "<th>value</th>\n",
       "<th>idx</th></tr></thead>\n",
       "    <tbody><tr><td>max f1</td>\n",
       "<td>0.7959731</td>\n",
       "<td>0.8164491</td>\n",
       "<td>58.0</td></tr>\n",
       "<tr><td>max f2</td>\n",
       "<td>0.3102251</td>\n",
       "<td>0.8476105</td>\n",
       "<td>160.0</td></tr>\n",
       "<tr><td>max f0point5</td>\n",
       "<td>0.8884340</td>\n",
       "<td>0.8788993</td>\n",
       "<td>40.0</td></tr>\n",
       "<tr><td>max accuracy</td>\n",
       "<td>0.7959731</td>\n",
       "<td>0.98594</td>\n",
       "<td>58.0</td></tr>\n",
       "<tr><td>max precision</td>\n",
       "<td>0.9997086</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max recall</td>\n",
       "<td>0.0190502</td>\n",
       "<td>1.0</td>\n",
       "<td>351.0</td></tr>\n",
       "<tr><td>max specificity</td>\n",
       "<td>0.9997086</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max absolute_mcc</td>\n",
       "<td>0.7959731</td>\n",
       "<td>0.8133666</td>\n",
       "<td>58.0</td></tr>\n",
       "<tr><td>max min_per_class_accuracy</td>\n",
       "<td>0.1426000</td>\n",
       "<td>0.9569253</td>\n",
       "<td>226.0</td></tr>\n",
       "<tr><td>max mean_per_class_accuracy</td>\n",
       "<td>0.1426000</td>\n",
       "<td>0.9579142</td>\n",
       "<td>226.0</td></tr>\n",
       "<tr><td>max tns</td>\n",
       "<td>0.9997086</td>\n",
       "<td>95798.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fns</td>\n",
       "<td>0.9997086</td>\n",
       "<td>4015.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fps</td>\n",
       "<td>0.0001983</td>\n",
       "<td>95798.0</td>\n",
       "<td>399.0</td></tr>\n",
       "<tr><td>max tps</td>\n",
       "<td>0.0190502</td>\n",
       "<td>4202.0</td>\n",
       "<td>351.0</td></tr>\n",
       "<tr><td>max tnr</td>\n",
       "<td>0.9997086</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fnr</td>\n",
       "<td>0.9997086</td>\n",
       "<td>0.9554974</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fpr</td>\n",
       "<td>0.0001983</td>\n",
       "<td>1.0</td>\n",
       "<td>399.0</td></tr>\n",
       "<tr><td>max tpr</td>\n",
       "<td>0.0190502</td>\n",
       "<td>1.0</td>\n",
       "<td>351.0</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-29.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-29 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-29 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-29 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-29 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-29 .h2o-table th,\n",
       "#h2o-table-29 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-29 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-29\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Gains/Lift Table: Avg response rate:  4.20 %, avg score:  7.00 %</caption>\n",
       "    <thead><tr><th>group</th>\n",
       "<th>cumulative_data_fraction</th>\n",
       "<th>lower_threshold</th>\n",
       "<th>lift</th>\n",
       "<th>cumulative_lift</th>\n",
       "<th>response_rate</th>\n",
       "<th>score</th>\n",
       "<th>cumulative_response_rate</th>\n",
       "<th>cumulative_score</th>\n",
       "<th>capture_rate</th>\n",
       "<th>cumulative_capture_rate</th>\n",
       "<th>gain</th>\n",
       "<th>cumulative_gain</th>\n",
       "<th>kolmogorov_smirnov</th></tr></thead>\n",
       "    <tbody><tr><td>1</td>\n",
       "<td>0.01032</td>\n",
       "<td>0.9969996</td>\n",
       "<td>23.7981913</td>\n",
       "<td>23.7981913</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9983429</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9983429</td>\n",
       "<td>0.2455973</td>\n",
       "<td>0.2455973</td>\n",
       "<td>2279.8191337</td>\n",
       "<td>2279.8191337</td>\n",
       "<td>0.2455973</td></tr>\n",
       "<tr><td>2</td>\n",
       "<td>0.02</td>\n",
       "<td>0.9800136</td>\n",
       "<td>23.6752668</td>\n",
       "<td>23.7386959</td>\n",
       "<td>0.9948347</td>\n",
       "<td>0.9906148</td>\n",
       "<td>0.9975</td>\n",
       "<td>0.9946025</td>\n",
       "<td>0.2291766</td>\n",
       "<td>0.4747739</td>\n",
       "<td>2267.5266795</td>\n",
       "<td>2273.8695859</td>\n",
       "<td>0.4747217</td></tr>\n",
       "<tr><td>3</td>\n",
       "<td>0.03</td>\n",
       "<td>0.9167327</td>\n",
       "<td>19.3241314</td>\n",
       "<td>22.2671744</td>\n",
       "<td>0.812</td>\n",
       "<td>0.9530255</td>\n",
       "<td>0.9356667</td>\n",
       "<td>0.9807435</td>\n",
       "<td>0.1932413</td>\n",
       "<td>0.6680152</td>\n",
       "<td>1832.4131366</td>\n",
       "<td>2126.7174361</td>\n",
       "<td>0.6660006</td></tr>\n",
       "<tr><td>4</td>\n",
       "<td>0.04</td>\n",
       "<td>0.5510434</td>\n",
       "<td>10.8281771</td>\n",
       "<td>19.4074250</td>\n",
       "<td>0.455</td>\n",
       "<td>0.7673460</td>\n",
       "<td>0.8155</td>\n",
       "<td>0.9273941</td>\n",
       "<td>0.1082818</td>\n",
       "<td>0.7762970</td>\n",
       "<td>982.8177059</td>\n",
       "<td>1840.7425036</td>\n",
       "<td>0.7685933</td></tr>\n",
       "<tr><td>5</td>\n",
       "<td>0.05</td>\n",
       "<td>0.3545707</td>\n",
       "<td>10.2332223</td>\n",
       "<td>17.5725845</td>\n",
       "<td>0.43</td>\n",
       "<td>0.4277765</td>\n",
       "<td>0.7384</td>\n",
       "<td>0.8274706</td>\n",
       "<td>0.1023322</td>\n",
       "<td>0.8786292</td>\n",
       "<td>923.3222275</td>\n",
       "<td>1657.2584484</td>\n",
       "<td>0.8649755</td></tr>\n",
       "<tr><td>6</td>\n",
       "<td>0.1</td>\n",
       "<td>0.1012878</td>\n",
       "<td>1.8086625</td>\n",
       "<td>9.6906235</td>\n",
       "<td>0.076</td>\n",
       "<td>0.1802490</td>\n",
       "<td>0.4072</td>\n",
       "<td>0.5038598</td>\n",
       "<td>0.0904331</td>\n",
       "<td>0.9690624</td>\n",
       "<td>80.8662542</td>\n",
       "<td>869.0623513</td>\n",
       "<td>0.9071821</td></tr>\n",
       "<tr><td>7</td>\n",
       "<td>0.15</td>\n",
       "<td>0.0634049</td>\n",
       "<td>0.3188958</td>\n",
       "<td>6.5667143</td>\n",
       "<td>0.0134</td>\n",
       "<td>0.0790632</td>\n",
       "<td>0.2759333</td>\n",
       "<td>0.3622609</td>\n",
       "<td>0.0159448</td>\n",
       "<td>0.9850071</td>\n",
       "<td>-68.1104236</td>\n",
       "<td>556.6714263</td>\n",
       "<td>0.8716332</td></tr>\n",
       "<tr><td>8</td>\n",
       "<td>0.2</td>\n",
       "<td>0.0465454</td>\n",
       "<td>0.1285102</td>\n",
       "<td>4.9571633</td>\n",
       "<td>0.0054</td>\n",
       "<td>0.0541206</td>\n",
       "<td>0.2083</td>\n",
       "<td>0.2852258</td>\n",
       "<td>0.0064255</td>\n",
       "<td>0.9914327</td>\n",
       "<td>-87.1489767</td>\n",
       "<td>395.7163256</td>\n",
       "<td>0.8261474</td></tr>\n",
       "<tr><td>9</td>\n",
       "<td>0.3</td>\n",
       "<td>0.0303223</td>\n",
       "<td>0.0404569</td>\n",
       "<td>3.3182611</td>\n",
       "<td>0.0017</td>\n",
       "<td>0.0375452</td>\n",
       "<td>0.1394333</td>\n",
       "<td>0.2026656</td>\n",
       "<td>0.0040457</td>\n",
       "<td>0.9954783</td>\n",
       "<td>-95.9543075</td>\n",
       "<td>231.8261145</td>\n",
       "<td>0.7259842</td></tr>\n",
       "<tr><td>10</td>\n",
       "<td>0.4</td>\n",
       "<td>0.0219472</td>\n",
       "<td>0.0285578</td>\n",
       "<td>2.4958353</td>\n",
       "<td>0.0012</td>\n",
       "<td>0.0256665</td>\n",
       "<td>0.104875</td>\n",
       "<td>0.1584158</td>\n",
       "<td>0.0028558</td>\n",
       "<td>0.9983341</td>\n",
       "<td>-97.1442170</td>\n",
       "<td>149.5835317</td>\n",
       "<td>0.6245789</td></tr>\n",
       "<tr><td>11</td>\n",
       "<td>0.5</td>\n",
       "<td>0.0169230</td>\n",
       "<td>0.0166587</td>\n",
       "<td>2.0</td>\n",
       "<td>0.0007</td>\n",
       "<td>0.0191906</td>\n",
       "<td>0.08404</td>\n",
       "<td>0.1305708</td>\n",
       "<td>0.0016659</td>\n",
       "<td>1.0</td>\n",
       "<td>-98.3341266</td>\n",
       "<td>100.0</td>\n",
       "<td>0.5219316</td></tr>\n",
       "<tr><td>12</td>\n",
       "<td>0.6</td>\n",
       "<td>0.0135262</td>\n",
       "<td>0.0</td>\n",
       "<td>1.6666667</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0151481</td>\n",
       "<td>0.0700333</td>\n",
       "<td>0.1113337</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>66.6666667</td>\n",
       "<td>0.4175453</td></tr>\n",
       "<tr><td>13</td>\n",
       "<td>0.7</td>\n",
       "<td>0.0107965</td>\n",
       "<td>0.0</td>\n",
       "<td>1.4285714</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0121393</td>\n",
       "<td>0.0600286</td>\n",
       "<td>0.0971630</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>42.8571429</td>\n",
       "<td>0.3131589</td></tr>\n",
       "<tr><td>14</td>\n",
       "<td>0.8</td>\n",
       "<td>0.0082955</td>\n",
       "<td>0.0</td>\n",
       "<td>1.25</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0095504</td>\n",
       "<td>0.052525</td>\n",
       "<td>0.0862115</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>25.0</td>\n",
       "<td>0.2087726</td></tr>\n",
       "<tr><td>15</td>\n",
       "<td>0.9</td>\n",
       "<td>0.0055877</td>\n",
       "<td>0.0</td>\n",
       "<td>1.1111111</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0070731</td>\n",
       "<td>0.0466889</td>\n",
       "<td>0.0774183</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>11.1111111</td>\n",
       "<td>0.1043863</td></tr>\n",
       "<tr><td>16</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0031996</td>\n",
       "<td>0.04202</td>\n",
       "<td>0.0699964</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>"
      ],
      "text/plain": [
       "ModelMetricsBinomial: drf\n",
       "** Reported on test data. **\n",
       "\n",
       "MSE: 0.013103536045933374\n",
       "RMSE: 0.11447067766870857\n",
       "LogLoss: 0.06021293925083577\n",
       "Mean Per-Class Error: 0.12964287191678184\n",
       "AUC: 0.9912719814049472\n",
       "AUCPR: 0.9029267705830357\n",
       "Gini: 0.9825439628098944\n",
       "\n",
       "Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.7959731126734471\n",
       "       0      1     Error    Rate\n",
       "-----  -----  ----  -------  -----------------\n",
       "0      95467  331   0.0035   (331.0/95798.0)\n",
       "1      1075   3127  0.2558   (1075.0/4202.0)\n",
       "Total  96542  3458  0.0141   (1406.0/100000.0)\n",
       "\n",
       "Maximum Metrics: Maximum metrics at their respective thresholds\n",
       "metric                       threshold    value     idx\n",
       "---------------------------  -----------  --------  -----\n",
       "max f1                       0.795973     0.816449  58\n",
       "max f2                       0.310225     0.84761   160\n",
       "max f0point5                 0.888434     0.878899  40\n",
       "max accuracy                 0.795973     0.98594   58\n",
       "max precision                0.999709     1         0\n",
       "max recall                   0.0190502    1         351\n",
       "max specificity              0.999709     1         0\n",
       "max absolute_mcc             0.795973     0.813367  58\n",
       "max min_per_class_accuracy   0.1426       0.956925  226\n",
       "max mean_per_class_accuracy  0.1426       0.957914  226\n",
       "max tns                      0.999709     95798     0\n",
       "max fns                      0.999709     4015      0\n",
       "max fps                      0.000198328  95798     399\n",
       "max tps                      0.0190502    4202      351\n",
       "max tnr                      0.999709     1         0\n",
       "max fnr                      0.999709     0.955497  0\n",
       "max fpr                      0.000198328  1         399\n",
       "max tpr                      0.0190502    1         351\n",
       "\n",
       "Gains/Lift Table: Avg response rate:  4.20 %, avg score:  7.00 %\n",
       "group    cumulative_data_fraction    lower_threshold    lift       cumulative_lift    response_rate    score       cumulative_response_rate    cumulative_score    capture_rate    cumulative_capture_rate    gain      cumulative_gain    kolmogorov_smirnov\n",
       "-------  --------------------------  -----------------  ---------  -----------------  ---------------  ----------  --------------------------  ------------------  --------------  -------------------------  --------  -----------------  --------------------\n",
       "1        0.01032                     0.997              23.7982    23.7982            1                0.998343    1                           0.998343            0.245597        0.245597                   2279.82   2279.82            0.245597\n",
       "2        0.02                        0.980014           23.6753    23.7387            0.994835         0.990615    0.9975                      0.994603            0.229177        0.474774                   2267.53   2273.87            0.474722\n",
       "3        0.03                        0.916733           19.3241    22.2672            0.812            0.953025    0.935667                    0.980743            0.193241        0.668015                   1832.41   2126.72            0.666001\n",
       "4        0.04                        0.551043           10.8282    19.4074            0.455            0.767346    0.8155                      0.927394            0.108282        0.776297                   982.818   1840.74            0.768593\n",
       "5        0.05                        0.354571           10.2332    17.5726            0.43             0.427777    0.7384                      0.827471            0.102332        0.878629                   923.322   1657.26            0.864975\n",
       "6        0.1                         0.101288           1.80866    9.69062            0.076            0.180249    0.4072                      0.50386             0.0904331       0.969062                   80.8663   869.062            0.907182\n",
       "7        0.15                        0.0634049          0.318896   6.56671            0.0134           0.0790632   0.275933                    0.362261            0.0159448       0.985007                   -68.1104  556.671            0.871633\n",
       "8        0.2                         0.0465454          0.12851    4.95716            0.0054           0.0541206   0.2083                      0.285226            0.00642551      0.991433                   -87.149   395.716            0.826147\n",
       "9        0.3                         0.0303223          0.0404569  3.31826            0.0017           0.0375452   0.139433                    0.202666            0.00404569      0.995478                   -95.9543  231.826            0.725984\n",
       "10       0.4                         0.0219472          0.0285578  2.49584            0.0012           0.0256665   0.104875                    0.158416            0.00285578      0.998334                   -97.1442  149.584            0.624579\n",
       "11       0.5                         0.016923           0.0166587  2                  0.0007           0.0191906   0.08404                     0.130571            0.00166587      1                          -98.3341  100                0.521932\n",
       "12       0.6                         0.0135262          0          1.66667            0                0.0151481   0.0700333                   0.111334            0               1                          -100      66.6667            0.417545\n",
       "13       0.7                         0.0107965          0          1.42857            0                0.0121393   0.0600286                   0.097163            0               1                          -100      42.8571            0.313159\n",
       "14       0.8                         0.00829551         0          1.25               0                0.0095504   0.052525                    0.0862115           0               1                          -100      25                 0.208773\n",
       "15       0.9                         0.00558768         0          1.11111            0                0.00707311  0.0466889                   0.0774183           0               1                          -100      11.1111            0.104386\n",
       "16       1                           0                  0          1                  0                0.00319961  0.04202                     0.0699964           0               1                          -100      0                  0"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf_next_model.model_performance(validation_h2o)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "d7decc86-f03c-4010-a3b8-65c53136a7cb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<pre style='margin: 1em 0 1em 0;'>ModelMetricsBinomial: drf\n",
       "** Reported on test data. **\n",
       "\n",
       "MSE: 0.013103536045933374\n",
       "RMSE: 0.11447067766870857\n",
       "LogLoss: 0.06021293925083577\n",
       "Mean Per-Class Error: 0.12964287191678184\n",
       "AUC: 0.9912719814049472\n",
       "AUCPR: 0.9029267705830357\n",
       "Gini: 0.9825439628098944</pre>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-36.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-36 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-36 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-36 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-36 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-36 .h2o-table th,\n",
       "#h2o-table-36 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-36 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-36\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.7959731126734471</caption>\n",
       "    <thead><tr><th></th>\n",
       "<th>0</th>\n",
       "<th>1</th>\n",
       "<th>Error</th>\n",
       "<th>Rate</th></tr></thead>\n",
       "    <tbody><tr><td>0</td>\n",
       "<td>95467.0</td>\n",
       "<td>331.0</td>\n",
       "<td>0.0035</td>\n",
       "<td> (331.0/95798.0)</td></tr>\n",
       "<tr><td>1</td>\n",
       "<td>1075.0</td>\n",
       "<td>3127.0</td>\n",
       "<td>0.2558</td>\n",
       "<td> (1075.0/4202.0)</td></tr>\n",
       "<tr><td>Total</td>\n",
       "<td>96542.0</td>\n",
       "<td>3458.0</td>\n",
       "<td>0.0141</td>\n",
       "<td> (1406.0/100000.0)</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-37.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-37 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-37 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-37 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-37 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-37 .h2o-table th,\n",
       "#h2o-table-37 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-37 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-37\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Maximum Metrics: Maximum metrics at their respective thresholds</caption>\n",
       "    <thead><tr><th>metric</th>\n",
       "<th>threshold</th>\n",
       "<th>value</th>\n",
       "<th>idx</th></tr></thead>\n",
       "    <tbody><tr><td>max f1</td>\n",
       "<td>0.7959731</td>\n",
       "<td>0.8164491</td>\n",
       "<td>58.0</td></tr>\n",
       "<tr><td>max f2</td>\n",
       "<td>0.3102251</td>\n",
       "<td>0.8476105</td>\n",
       "<td>160.0</td></tr>\n",
       "<tr><td>max f0point5</td>\n",
       "<td>0.8884340</td>\n",
       "<td>0.8788993</td>\n",
       "<td>40.0</td></tr>\n",
       "<tr><td>max accuracy</td>\n",
       "<td>0.7959731</td>\n",
       "<td>0.98594</td>\n",
       "<td>58.0</td></tr>\n",
       "<tr><td>max precision</td>\n",
       "<td>0.9997086</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max recall</td>\n",
       "<td>0.0190502</td>\n",
       "<td>1.0</td>\n",
       "<td>351.0</td></tr>\n",
       "<tr><td>max specificity</td>\n",
       "<td>0.9997086</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max absolute_mcc</td>\n",
       "<td>0.7959731</td>\n",
       "<td>0.8133666</td>\n",
       "<td>58.0</td></tr>\n",
       "<tr><td>max min_per_class_accuracy</td>\n",
       "<td>0.1426000</td>\n",
       "<td>0.9569253</td>\n",
       "<td>226.0</td></tr>\n",
       "<tr><td>max mean_per_class_accuracy</td>\n",
       "<td>0.1426000</td>\n",
       "<td>0.9579142</td>\n",
       "<td>226.0</td></tr>\n",
       "<tr><td>max tns</td>\n",
       "<td>0.9997086</td>\n",
       "<td>95798.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fns</td>\n",
       "<td>0.9997086</td>\n",
       "<td>4015.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fps</td>\n",
       "<td>0.0001983</td>\n",
       "<td>95798.0</td>\n",
       "<td>399.0</td></tr>\n",
       "<tr><td>max tps</td>\n",
       "<td>0.0190502</td>\n",
       "<td>4202.0</td>\n",
       "<td>351.0</td></tr>\n",
       "<tr><td>max tnr</td>\n",
       "<td>0.9997086</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fnr</td>\n",
       "<td>0.9997086</td>\n",
       "<td>0.9554974</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fpr</td>\n",
       "<td>0.0001983</td>\n",
       "<td>1.0</td>\n",
       "<td>399.0</td></tr>\n",
       "<tr><td>max tpr</td>\n",
       "<td>0.0190502</td>\n",
       "<td>1.0</td>\n",
       "<td>351.0</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-38.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-38 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-38 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-38 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-38 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-38 .h2o-table th,\n",
       "#h2o-table-38 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-38 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-38\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Gains/Lift Table: Avg response rate:  4.20 %, avg score:  7.00 %</caption>\n",
       "    <thead><tr><th>group</th>\n",
       "<th>cumulative_data_fraction</th>\n",
       "<th>lower_threshold</th>\n",
       "<th>lift</th>\n",
       "<th>cumulative_lift</th>\n",
       "<th>response_rate</th>\n",
       "<th>score</th>\n",
       "<th>cumulative_response_rate</th>\n",
       "<th>cumulative_score</th>\n",
       "<th>capture_rate</th>\n",
       "<th>cumulative_capture_rate</th>\n",
       "<th>gain</th>\n",
       "<th>cumulative_gain</th>\n",
       "<th>kolmogorov_smirnov</th></tr></thead>\n",
       "    <tbody><tr><td>1</td>\n",
       "<td>0.01032</td>\n",
       "<td>0.9969996</td>\n",
       "<td>23.7981913</td>\n",
       "<td>23.7981913</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9983429</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9983429</td>\n",
       "<td>0.2455973</td>\n",
       "<td>0.2455973</td>\n",
       "<td>2279.8191337</td>\n",
       "<td>2279.8191337</td>\n",
       "<td>0.2455973</td></tr>\n",
       "<tr><td>2</td>\n",
       "<td>0.02</td>\n",
       "<td>0.9800136</td>\n",
       "<td>23.6752668</td>\n",
       "<td>23.7386959</td>\n",
       "<td>0.9948347</td>\n",
       "<td>0.9906148</td>\n",
       "<td>0.9975</td>\n",
       "<td>0.9946025</td>\n",
       "<td>0.2291766</td>\n",
       "<td>0.4747739</td>\n",
       "<td>2267.5266795</td>\n",
       "<td>2273.8695859</td>\n",
       "<td>0.4747217</td></tr>\n",
       "<tr><td>3</td>\n",
       "<td>0.03</td>\n",
       "<td>0.9167327</td>\n",
       "<td>19.3241314</td>\n",
       "<td>22.2671744</td>\n",
       "<td>0.812</td>\n",
       "<td>0.9530255</td>\n",
       "<td>0.9356667</td>\n",
       "<td>0.9807435</td>\n",
       "<td>0.1932413</td>\n",
       "<td>0.6680152</td>\n",
       "<td>1832.4131366</td>\n",
       "<td>2126.7174361</td>\n",
       "<td>0.6660006</td></tr>\n",
       "<tr><td>4</td>\n",
       "<td>0.04</td>\n",
       "<td>0.5510434</td>\n",
       "<td>10.8281771</td>\n",
       "<td>19.4074250</td>\n",
       "<td>0.455</td>\n",
       "<td>0.7673460</td>\n",
       "<td>0.8155</td>\n",
       "<td>0.9273941</td>\n",
       "<td>0.1082818</td>\n",
       "<td>0.7762970</td>\n",
       "<td>982.8177059</td>\n",
       "<td>1840.7425036</td>\n",
       "<td>0.7685933</td></tr>\n",
       "<tr><td>5</td>\n",
       "<td>0.05</td>\n",
       "<td>0.3545707</td>\n",
       "<td>10.2332223</td>\n",
       "<td>17.5725845</td>\n",
       "<td>0.43</td>\n",
       "<td>0.4277765</td>\n",
       "<td>0.7384</td>\n",
       "<td>0.8274706</td>\n",
       "<td>0.1023322</td>\n",
       "<td>0.8786292</td>\n",
       "<td>923.3222275</td>\n",
       "<td>1657.2584484</td>\n",
       "<td>0.8649755</td></tr>\n",
       "<tr><td>6</td>\n",
       "<td>0.1</td>\n",
       "<td>0.1012878</td>\n",
       "<td>1.8086625</td>\n",
       "<td>9.6906235</td>\n",
       "<td>0.076</td>\n",
       "<td>0.1802490</td>\n",
       "<td>0.4072</td>\n",
       "<td>0.5038598</td>\n",
       "<td>0.0904331</td>\n",
       "<td>0.9690624</td>\n",
       "<td>80.8662542</td>\n",
       "<td>869.0623513</td>\n",
       "<td>0.9071821</td></tr>\n",
       "<tr><td>7</td>\n",
       "<td>0.15</td>\n",
       "<td>0.0634049</td>\n",
       "<td>0.3188958</td>\n",
       "<td>6.5667143</td>\n",
       "<td>0.0134</td>\n",
       "<td>0.0790632</td>\n",
       "<td>0.2759333</td>\n",
       "<td>0.3622609</td>\n",
       "<td>0.0159448</td>\n",
       "<td>0.9850071</td>\n",
       "<td>-68.1104236</td>\n",
       "<td>556.6714263</td>\n",
       "<td>0.8716332</td></tr>\n",
       "<tr><td>8</td>\n",
       "<td>0.2</td>\n",
       "<td>0.0465454</td>\n",
       "<td>0.1285102</td>\n",
       "<td>4.9571633</td>\n",
       "<td>0.0054</td>\n",
       "<td>0.0541206</td>\n",
       "<td>0.2083</td>\n",
       "<td>0.2852258</td>\n",
       "<td>0.0064255</td>\n",
       "<td>0.9914327</td>\n",
       "<td>-87.1489767</td>\n",
       "<td>395.7163256</td>\n",
       "<td>0.8261474</td></tr>\n",
       "<tr><td>9</td>\n",
       "<td>0.3</td>\n",
       "<td>0.0303223</td>\n",
       "<td>0.0404569</td>\n",
       "<td>3.3182611</td>\n",
       "<td>0.0017</td>\n",
       "<td>0.0375452</td>\n",
       "<td>0.1394333</td>\n",
       "<td>0.2026656</td>\n",
       "<td>0.0040457</td>\n",
       "<td>0.9954783</td>\n",
       "<td>-95.9543075</td>\n",
       "<td>231.8261145</td>\n",
       "<td>0.7259842</td></tr>\n",
       "<tr><td>10</td>\n",
       "<td>0.4</td>\n",
       "<td>0.0219472</td>\n",
       "<td>0.0285578</td>\n",
       "<td>2.4958353</td>\n",
       "<td>0.0012</td>\n",
       "<td>0.0256665</td>\n",
       "<td>0.104875</td>\n",
       "<td>0.1584158</td>\n",
       "<td>0.0028558</td>\n",
       "<td>0.9983341</td>\n",
       "<td>-97.1442170</td>\n",
       "<td>149.5835317</td>\n",
       "<td>0.6245789</td></tr>\n",
       "<tr><td>11</td>\n",
       "<td>0.5</td>\n",
       "<td>0.0169230</td>\n",
       "<td>0.0166587</td>\n",
       "<td>2.0</td>\n",
       "<td>0.0007</td>\n",
       "<td>0.0191906</td>\n",
       "<td>0.08404</td>\n",
       "<td>0.1305708</td>\n",
       "<td>0.0016659</td>\n",
       "<td>1.0</td>\n",
       "<td>-98.3341266</td>\n",
       "<td>100.0</td>\n",
       "<td>0.5219316</td></tr>\n",
       "<tr><td>12</td>\n",
       "<td>0.6</td>\n",
       "<td>0.0135262</td>\n",
       "<td>0.0</td>\n",
       "<td>1.6666667</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0151481</td>\n",
       "<td>0.0700333</td>\n",
       "<td>0.1113337</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>66.6666667</td>\n",
       "<td>0.4175453</td></tr>\n",
       "<tr><td>13</td>\n",
       "<td>0.7</td>\n",
       "<td>0.0107965</td>\n",
       "<td>0.0</td>\n",
       "<td>1.4285714</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0121393</td>\n",
       "<td>0.0600286</td>\n",
       "<td>0.0971630</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>42.8571429</td>\n",
       "<td>0.3131589</td></tr>\n",
       "<tr><td>14</td>\n",
       "<td>0.8</td>\n",
       "<td>0.0082955</td>\n",
       "<td>0.0</td>\n",
       "<td>1.25</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0095504</td>\n",
       "<td>0.052525</td>\n",
       "<td>0.0862115</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>25.0</td>\n",
       "<td>0.2087726</td></tr>\n",
       "<tr><td>15</td>\n",
       "<td>0.9</td>\n",
       "<td>0.0055877</td>\n",
       "<td>0.0</td>\n",
       "<td>1.1111111</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0070731</td>\n",
       "<td>0.0466889</td>\n",
       "<td>0.0774183</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>11.1111111</td>\n",
       "<td>0.1043863</td></tr>\n",
       "<tr><td>16</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0031996</td>\n",
       "<td>0.04202</td>\n",
       "<td>0.0699964</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>"
      ],
      "text/plain": [
       "ModelMetricsBinomial: drf\n",
       "** Reported on test data. **\n",
       "\n",
       "MSE: 0.013103536045933374\n",
       "RMSE: 0.11447067766870857\n",
       "LogLoss: 0.06021293925083577\n",
       "Mean Per-Class Error: 0.12964287191678184\n",
       "AUC: 0.9912719814049472\n",
       "AUCPR: 0.9029267705830357\n",
       "Gini: 0.9825439628098944\n",
       "\n",
       "Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.7959731126734471\n",
       "       0      1     Error    Rate\n",
       "-----  -----  ----  -------  -----------------\n",
       "0      95467  331   0.0035   (331.0/95798.0)\n",
       "1      1075   3127  0.2558   (1075.0/4202.0)\n",
       "Total  96542  3458  0.0141   (1406.0/100000.0)\n",
       "\n",
       "Maximum Metrics: Maximum metrics at their respective thresholds\n",
       "metric                       threshold    value     idx\n",
       "---------------------------  -----------  --------  -----\n",
       "max f1                       0.795973     0.816449  58\n",
       "max f2                       0.310225     0.84761   160\n",
       "max f0point5                 0.888434     0.878899  40\n",
       "max accuracy                 0.795973     0.98594   58\n",
       "max precision                0.999709     1         0\n",
       "max recall                   0.0190502    1         351\n",
       "max specificity              0.999709     1         0\n",
       "max absolute_mcc             0.795973     0.813367  58\n",
       "max min_per_class_accuracy   0.1426       0.956925  226\n",
       "max mean_per_class_accuracy  0.1426       0.957914  226\n",
       "max tns                      0.999709     95798     0\n",
       "max fns                      0.999709     4015      0\n",
       "max fps                      0.000198328  95798     399\n",
       "max tps                      0.0190502    4202      351\n",
       "max tnr                      0.999709     1         0\n",
       "max fnr                      0.999709     0.955497  0\n",
       "max fpr                      0.000198328  1         399\n",
       "max tpr                      0.0190502    1         351\n",
       "\n",
       "Gains/Lift Table: Avg response rate:  4.20 %, avg score:  7.00 %\n",
       "group    cumulative_data_fraction    lower_threshold    lift       cumulative_lift    response_rate    score       cumulative_response_rate    cumulative_score    capture_rate    cumulative_capture_rate    gain      cumulative_gain    kolmogorov_smirnov\n",
       "-------  --------------------------  -----------------  ---------  -----------------  ---------------  ----------  --------------------------  ------------------  --------------  -------------------------  --------  -----------------  --------------------\n",
       "1        0.01032                     0.997              23.7982    23.7982            1                0.998343    1                           0.998343            0.245597        0.245597                   2279.82   2279.82            0.245597\n",
       "2        0.02                        0.980014           23.6753    23.7387            0.994835         0.990615    0.9975                      0.994603            0.229177        0.474774                   2267.53   2273.87            0.474722\n",
       "3        0.03                        0.916733           19.3241    22.2672            0.812            0.953025    0.935667                    0.980743            0.193241        0.668015                   1832.41   2126.72            0.666001\n",
       "4        0.04                        0.551043           10.8282    19.4074            0.455            0.767346    0.8155                      0.927394            0.108282        0.776297                   982.818   1840.74            0.768593\n",
       "5        0.05                        0.354571           10.2332    17.5726            0.43             0.427777    0.7384                      0.827471            0.102332        0.878629                   923.322   1657.26            0.864975\n",
       "6        0.1                         0.101288           1.80866    9.69062            0.076            0.180249    0.4072                      0.50386             0.0904331       0.969062                   80.8663   869.062            0.907182\n",
       "7        0.15                        0.0634049          0.318896   6.56671            0.0134           0.0790632   0.275933                    0.362261            0.0159448       0.985007                   -68.1104  556.671            0.871633\n",
       "8        0.2                         0.0465454          0.12851    4.95716            0.0054           0.0541206   0.2083                      0.285226            0.00642551      0.991433                   -87.149   395.716            0.826147\n",
       "9        0.3                         0.0303223          0.0404569  3.31826            0.0017           0.0375452   0.139433                    0.202666            0.00404569      0.995478                   -95.9543  231.826            0.725984\n",
       "10       0.4                         0.0219472          0.0285578  2.49584            0.0012           0.0256665   0.104875                    0.158416            0.00285578      0.998334                   -97.1442  149.584            0.624579\n",
       "11       0.5                         0.016923           0.0166587  2                  0.0007           0.0191906   0.08404                     0.130571            0.00166587      1                          -98.3341  100                0.521932\n",
       "12       0.6                         0.0135262          0          1.66667            0                0.0151481   0.0700333                   0.111334            0               1                          -100      66.6667            0.417545\n",
       "13       0.7                         0.0107965          0          1.42857            0                0.0121393   0.0600286                   0.097163            0               1                          -100      42.8571            0.313159\n",
       "14       0.8                         0.00829551         0          1.25               0                0.0095504   0.052525                    0.0862115           0               1                          -100      25                 0.208773\n",
       "15       0.9                         0.00558768         0          1.11111            0                0.00707311  0.0466889                   0.0774183           0               1                          -100      11.1111            0.104386\n",
       "16       1                           0                  0          1                  0                0.00319961  0.04202                     0.0699964           0               1                          -100      0                  0"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf_next_model.model_performance(validation_h2o) #looks overfitted"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "b314a0dd-d778-4030-ac69-c6dd890eb825",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Max Depth: 20\n"
     ]
    }
   ],
   "source": [
    "max_depth = rf_next_model.params['max_depth']['actual']\n",
    "print(\"Max Depth:\", max_depth)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "6dfba082-1ff3-45fc-ade6-3987d4eb333b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "drf Model Build progress: |██████████████████████████████████████████████████████| (done) 100%\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<pre style='margin: 1em 0 1em 0;'>Model Details\n",
       "=============\n",
       "H2ORandomForestEstimator : Distributed Random Forest\n",
       "Model Key: DRF_model_python_1680256840616_211\n",
       "</pre>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-39.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-39 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-39 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-39 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-39 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-39 .h2o-table th,\n",
       "#h2o-table-39 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-39 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-39\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Model Summary: </caption>\n",
       "    <thead><tr><th></th>\n",
       "<th>number_of_trees</th>\n",
       "<th>number_of_internal_trees</th>\n",
       "<th>model_size_in_bytes</th>\n",
       "<th>min_depth</th>\n",
       "<th>max_depth</th>\n",
       "<th>mean_depth</th>\n",
       "<th>min_leaves</th>\n",
       "<th>max_leaves</th>\n",
       "<th>mean_leaves</th></tr></thead>\n",
       "    <tbody><tr><td></td>\n",
       "<td>150.0</td>\n",
       "<td>150.0</td>\n",
       "<td>1028619.0</td>\n",
       "<td>15.0</td>\n",
       "<td>15.0</td>\n",
       "<td>15.0</td>\n",
       "<td>315.0</td>\n",
       "<td>912.0</td>\n",
       "<td>539.87335</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>\n",
       "<div style='margin: 1em 0 1em 0;'><pre style='margin: 1em 0 1em 0;'>ModelMetricsBinomial: drf\n",
       "** Reported on train data. **\n",
       "\n",
       "MSE: 0.01496256045483796\n",
       "RMSE: 0.1223215453419305\n",
       "LogLoss: 0.06856855758096712\n",
       "Mean Per-Class Error: 0.040119712474570184\n",
       "AUC: 0.9895241336871461\n",
       "AUCPR: 0.9714032767310343\n",
       "Gini: 0.9790482673742922</pre>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-40.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-40 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-40 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-40 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-40 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-40 .h2o-table th,\n",
       "#h2o-table-40 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-40 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-40\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.155163909279127</caption>\n",
       "    <thead><tr><th></th>\n",
       "<th>0</th>\n",
       "<th>1</th>\n",
       "<th>Error</th>\n",
       "<th>Rate</th></tr></thead>\n",
       "    <tbody><tr><td>0</td>\n",
       "<td>194321.0</td>\n",
       "<td>318.0</td>\n",
       "<td>0.0016</td>\n",
       "<td> (318.0/194639.0)</td></tr>\n",
       "<tr><td>1</td>\n",
       "<td>2052.0</td>\n",
       "<td>24053.0</td>\n",
       "<td>0.0786</td>\n",
       "<td> (2052.0/26105.0)</td></tr>\n",
       "<tr><td>Total</td>\n",
       "<td>196373.0</td>\n",
       "<td>24371.0</td>\n",
       "<td>0.0107</td>\n",
       "<td> (2370.0/220744.0)</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-41.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-41 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-41 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-41 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-41 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-41 .h2o-table th,\n",
       "#h2o-table-41 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-41 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-41\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Maximum Metrics: Maximum metrics at their respective thresholds</caption>\n",
       "    <thead><tr><th>metric</th>\n",
       "<th>threshold</th>\n",
       "<th>value</th>\n",
       "<th>idx</th></tr></thead>\n",
       "    <tbody><tr><td>max f1</td>\n",
       "<td>0.1551639</td>\n",
       "<td>0.9530470</td>\n",
       "<td>244.0</td></tr>\n",
       "<tr><td>max f2</td>\n",
       "<td>0.1085940</td>\n",
       "<td>0.9378563</td>\n",
       "<td>258.0</td></tr>\n",
       "<tr><td>max f0point5</td>\n",
       "<td>0.2180372</td>\n",
       "<td>0.9763182</td>\n",
       "<td>227.0</td></tr>\n",
       "<tr><td>max accuracy</td>\n",
       "<td>0.1551639</td>\n",
       "<td>0.9892636</td>\n",
       "<td>244.0</td></tr>\n",
       "<tr><td>max precision</td>\n",
       "<td>0.9999772</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max recall</td>\n",
       "<td>0.0136117</td>\n",
       "<td>1.0</td>\n",
       "<td>374.0</td></tr>\n",
       "<tr><td>max specificity</td>\n",
       "<td>0.9999772</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max absolute_mcc</td>\n",
       "<td>0.1551639</td>\n",
       "<td>0.9477068</td>\n",
       "<td>244.0</td></tr>\n",
       "<tr><td>max min_per_class_accuracy</td>\n",
       "<td>0.0527100</td>\n",
       "<td>0.9543382</td>\n",
       "<td>295.0</td></tr>\n",
       "<tr><td>max mean_per_class_accuracy</td>\n",
       "<td>0.0867589</td>\n",
       "<td>0.9642208</td>\n",
       "<td>269.0</td></tr>\n",
       "<tr><td>max tns</td>\n",
       "<td>0.9999772</td>\n",
       "<td>194639.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fns</td>\n",
       "<td>0.9999772</td>\n",
       "<td>19082.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fps</td>\n",
       "<td>0.0007168</td>\n",
       "<td>194639.0</td>\n",
       "<td>399.0</td></tr>\n",
       "<tr><td>max tps</td>\n",
       "<td>0.0136117</td>\n",
       "<td>26105.0</td>\n",
       "<td>374.0</td></tr>\n",
       "<tr><td>max tnr</td>\n",
       "<td>0.9999772</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fnr</td>\n",
       "<td>0.9999772</td>\n",
       "<td>0.7309711</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fpr</td>\n",
       "<td>0.0007168</td>\n",
       "<td>1.0</td>\n",
       "<td>399.0</td></tr>\n",
       "<tr><td>max tpr</td>\n",
       "<td>0.0136117</td>\n",
       "<td>1.0</td>\n",
       "<td>374.0</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-42.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-42 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-42 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-42 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-42 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-42 .h2o-table th,\n",
       "#h2o-table-42 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-42 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-42\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Gains/Lift Table: Avg response rate: 11.83 %, avg score: 11.79 %</caption>\n",
       "    <thead><tr><th>group</th>\n",
       "<th>cumulative_data_fraction</th>\n",
       "<th>lower_threshold</th>\n",
       "<th>lift</th>\n",
       "<th>cumulative_lift</th>\n",
       "<th>response_rate</th>\n",
       "<th>score</th>\n",
       "<th>cumulative_response_rate</th>\n",
       "<th>cumulative_score</th>\n",
       "<th>capture_rate</th>\n",
       "<th>cumulative_capture_rate</th>\n",
       "<th>gain</th>\n",
       "<th>cumulative_gain</th>\n",
       "<th>kolmogorov_smirnov</th></tr></thead>\n",
       "    <tbody><tr><td>1</td>\n",
       "<td>0.0278377</td>\n",
       "<td>1.0</td>\n",
       "<td>8.4560046</td>\n",
       "<td>8.4560046</td>\n",
       "<td>1.0</td>\n",
       "<td>1.0</td>\n",
       "<td>1.0</td>\n",
       "<td>1.0</td>\n",
       "<td>0.2353955</td>\n",
       "<td>0.2353955</td>\n",
       "<td>745.6004597</td>\n",
       "<td>745.6004597</td>\n",
       "<td>0.2353955</td></tr>\n",
       "<tr><td>2</td>\n",
       "<td>0.0300031</td>\n",
       "<td>0.9998490</td>\n",
       "<td>8.4560046</td>\n",
       "<td>8.4560046</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9999085</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9999934</td>\n",
       "<td>0.0183107</td>\n",
       "<td>0.2537062</td>\n",
       "<td>745.6004597</td>\n",
       "<td>745.6004597</td>\n",
       "<td>0.2537062</td></tr>\n",
       "<tr><td>3</td>\n",
       "<td>0.0400011</td>\n",
       "<td>0.9969279</td>\n",
       "<td>8.4560046</td>\n",
       "<td>8.4560046</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9987728</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9996883</td>\n",
       "<td>0.0845432</td>\n",
       "<td>0.3382494</td>\n",
       "<td>745.6004597</td>\n",
       "<td>745.6004597</td>\n",
       "<td>0.3382494</td></tr>\n",
       "<tr><td>4</td>\n",
       "<td>0.0500036</td>\n",
       "<td>0.9867200</td>\n",
       "<td>8.4560046</td>\n",
       "<td>8.4560046</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9933516</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9984207</td>\n",
       "<td>0.0845815</td>\n",
       "<td>0.4228309</td>\n",
       "<td>745.6004597</td>\n",
       "<td>745.6004597</td>\n",
       "<td>0.4228309</td></tr>\n",
       "<tr><td>5</td>\n",
       "<td>0.1000027</td>\n",
       "<td>0.4987511</td>\n",
       "<td>8.4498754</td>\n",
       "<td>8.4529401</td>\n",
       "<td>0.9992752</td>\n",
       "<td>0.8650691</td>\n",
       "<td>0.9996376</td>\n",
       "<td>0.9317479</td>\n",
       "<td>0.4224861</td>\n",
       "<td>0.8453170</td>\n",
       "<td>744.9875392</td>\n",
       "<td>745.2940133</td>\n",
       "<td>0.8452759</td></tr>\n",
       "<tr><td>6</td>\n",
       "<td>0.1500018</td>\n",
       "<td>0.0529423</td>\n",
       "<td>2.1697386</td>\n",
       "<td>6.3586029</td>\n",
       "<td>0.2565915</td>\n",
       "<td>0.1227044</td>\n",
       "<td>0.7519630</td>\n",
       "<td>0.6620749</td>\n",
       "<td>0.1084850</td>\n",
       "<td>0.9538020</td>\n",
       "<td>116.9738608</td>\n",
       "<td>535.8602877</td>\n",
       "<td>0.9116059</td></tr>\n",
       "<tr><td>7</td>\n",
       "<td>0.2000009</td>\n",
       "<td>0.0395958</td>\n",
       "<td>0.2689189</td>\n",
       "<td>4.8362164</td>\n",
       "<td>0.0318021</td>\n",
       "<td>0.0450081</td>\n",
       "<td>0.5719269</td>\n",
       "<td>0.5078117</td>\n",
       "<td>0.0134457</td>\n",
       "<td>0.9672477</td>\n",
       "<td>-73.1081126</td>\n",
       "<td>383.6216360</td>\n",
       "<td>0.8701499</td></tr>\n",
       "<tr><td>8</td>\n",
       "<td>0.2999991</td>\n",
       "<td>0.0297699</td>\n",
       "<td>0.1428871</td>\n",
       "<td>3.2717969</td>\n",
       "<td>0.0168977</td>\n",
       "<td>0.0338603</td>\n",
       "<td>0.3869200</td>\n",
       "<td>0.3498303</td>\n",
       "<td>0.0142885</td>\n",
       "<td>0.9815361</td>\n",
       "<td>-85.7112906</td>\n",
       "<td>227.1796895</td>\n",
       "<td>0.7729448</td></tr>\n",
       "<tr><td>9</td>\n",
       "<td>0.4000018</td>\n",
       "<td>0.0255474</td>\n",
       "<td>0.0877203</td>\n",
       "<td>2.4757597</td>\n",
       "<td>0.0103737</td>\n",
       "<td>0.0274375</td>\n",
       "<td>0.2927813</td>\n",
       "<td>0.2692303</td>\n",
       "<td>0.0087723</td>\n",
       "<td>0.9903084</td>\n",
       "<td>-91.2279726</td>\n",
       "<td>147.5759710</td>\n",
       "<td>0.6694785</td></tr>\n",
       "<tr><td>10</td>\n",
       "<td>0.5</td>\n",
       "<td>0.0227505</td>\n",
       "<td>0.0417552</td>\n",
       "<td>1.9889676</td>\n",
       "<td>0.0049379</td>\n",
       "<td>0.0240940</td>\n",
       "<td>0.2352136</td>\n",
       "<td>0.2202039</td>\n",
       "<td>0.0041754</td>\n",
       "<td>0.9944838</td>\n",
       "<td>-95.8244790</td>\n",
       "<td>98.8967631</td>\n",
       "<td>0.5608040</td></tr>\n",
       "<tr><td>11</td>\n",
       "<td>0.5999982</td>\n",
       "<td>0.0202302</td>\n",
       "<td>0.0287306</td>\n",
       "<td>1.6622664</td>\n",
       "<td>0.0033977</td>\n",
       "<td>0.0214862</td>\n",
       "<td>0.1965782</td>\n",
       "<td>0.1870848</td>\n",
       "<td>0.0028730</td>\n",
       "<td>0.9973568</td>\n",
       "<td>-97.1269351</td>\n",
       "<td>66.2266401</td>\n",
       "<td>0.4506524</td></tr>\n",
       "<tr><td>12</td>\n",
       "<td>0.7000009</td>\n",
       "<td>0.0176692</td>\n",
       "<td>0.0141731</td>\n",
       "<td>1.4268184</td>\n",
       "<td>0.0016761</td>\n",
       "<td>0.0189540</td>\n",
       "<td>0.1687343</td>\n",
       "<td>0.1630655</td>\n",
       "<td>0.0014174</td>\n",
       "<td>0.9987742</td>\n",
       "<td>-98.5826855</td>\n",
       "<td>42.6818412</td>\n",
       "<td>0.3388448</td></tr>\n",
       "<tr><td>13</td>\n",
       "<td>0.7999991</td>\n",
       "<td>0.0149531</td>\n",
       "<td>0.0103430</td>\n",
       "<td>1.2497620</td>\n",
       "<td>0.0012232</td>\n",
       "<td>0.0163162</td>\n",
       "<td>0.1477958</td>\n",
       "<td>0.1447221</td>\n",
       "<td>0.0010343</td>\n",
       "<td>0.9998085</td>\n",
       "<td>-98.9656966</td>\n",
       "<td>24.9761998</td>\n",
       "<td>0.2266078</td></tr>\n",
       "<tr><td>14</td>\n",
       "<td>0.8999973</td>\n",
       "<td>0.0116953</td>\n",
       "<td>0.0019154</td>\n",
       "<td>1.1111145</td>\n",
       "<td>0.0002265</td>\n",
       "<td>0.0134797</td>\n",
       "<td>0.1313995</td>\n",
       "<td>0.1301399</td>\n",
       "<td>0.0001915</td>\n",
       "<td>1.0</td>\n",
       "<td>-99.8084623</td>\n",
       "<td>11.1114467</td>\n",
       "<td>0.1134151</td></tr>\n",
       "<tr><td>15</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0079988</td>\n",
       "<td>0.1182592</td>\n",
       "<td>0.1179254</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div></div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-43.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-43 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-43 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-43 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-43 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-43 .h2o-table th,\n",
       "#h2o-table-43 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-43 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-43\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Scoring History: </caption>\n",
       "    <thead><tr><th></th>\n",
       "<th>timestamp</th>\n",
       "<th>duration</th>\n",
       "<th>number_of_trees</th>\n",
       "<th>training_rmse</th>\n",
       "<th>training_logloss</th>\n",
       "<th>training_auc</th>\n",
       "<th>training_pr_auc</th>\n",
       "<th>training_lift</th>\n",
       "<th>training_classification_error</th></tr></thead>\n",
       "    <tbody><tr><td></td>\n",
       "<td>2023-03-31 12:17:39</td>\n",
       "<td> 0.003 sec</td>\n",
       "<td>0.0</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:17:39</td>\n",
       "<td> 0.162 sec</td>\n",
       "<td>1.0</td>\n",
       "<td>0.1574063</td>\n",
       "<td>0.1455872</td>\n",
       "<td>0.9284722</td>\n",
       "<td>0.8717686</td>\n",
       "<td>8.3965046</td>\n",
       "<td>0.0267340</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:17:39</td>\n",
       "<td> 0.260 sec</td>\n",
       "<td>2.0</td>\n",
       "<td>0.1432764</td>\n",
       "<td>0.1314690</td>\n",
       "<td>0.9509010</td>\n",
       "<td>0.9028030</td>\n",
       "<td>8.4042331</td>\n",
       "<td>0.0216219</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:17:39</td>\n",
       "<td> 0.334 sec</td>\n",
       "<td>3.0</td>\n",
       "<td>0.1443787</td>\n",
       "<td>0.1222561</td>\n",
       "<td>0.9518251</td>\n",
       "<td>0.9050887</td>\n",
       "<td>8.4177528</td>\n",
       "<td>0.0214600</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:17:39</td>\n",
       "<td> 0.421 sec</td>\n",
       "<td>4.0</td>\n",
       "<td>0.1435711</td>\n",
       "<td>0.1104916</td>\n",
       "<td>0.9552470</td>\n",
       "<td>0.9109370</td>\n",
       "<td>8.4314346</td>\n",
       "<td>0.0208064</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:17:39</td>\n",
       "<td> 0.511 sec</td>\n",
       "<td>5.0</td>\n",
       "<td>0.1388640</td>\n",
       "<td>0.1010150</td>\n",
       "<td>0.9613219</td>\n",
       "<td>0.9219568</td>\n",
       "<td>8.4314475</td>\n",
       "<td>0.0193302</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:17:39</td>\n",
       "<td> 0.600 sec</td>\n",
       "<td>6.0</td>\n",
       "<td>0.1348861</td>\n",
       "<td>0.0922489</td>\n",
       "<td>0.9665993</td>\n",
       "<td>0.9312010</td>\n",
       "<td>8.4371671</td>\n",
       "<td>0.0181969</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:17:39</td>\n",
       "<td> 0.688 sec</td>\n",
       "<td>7.0</td>\n",
       "<td>0.1332771</td>\n",
       "<td>0.0872902</td>\n",
       "<td>0.9686599</td>\n",
       "<td>0.9353213</td>\n",
       "<td>8.4428151</td>\n",
       "<td>0.0179897</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:17:39</td>\n",
       "<td> 0.779 sec</td>\n",
       "<td>8.0</td>\n",
       "<td>0.1325117</td>\n",
       "<td>0.0839838</td>\n",
       "<td>0.9702155</td>\n",
       "<td>0.9381654</td>\n",
       "<td>8.4473516</td>\n",
       "<td>0.0175254</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:17:39</td>\n",
       "<td> 0.876 sec</td>\n",
       "<td>9.0</td>\n",
       "<td>0.1318092</td>\n",
       "<td>0.0829883</td>\n",
       "<td>0.9711842</td>\n",
       "<td>0.9405422</td>\n",
       "<td>8.4475301</td>\n",
       "<td>0.0170744</td></tr>\n",
       "<tr><td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:17:42</td>\n",
       "<td> 3.323 sec</td>\n",
       "<td>34.0</td>\n",
       "<td>0.1246715</td>\n",
       "<td>0.0710023</td>\n",
       "<td>0.9839537</td>\n",
       "<td>0.9624265</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0124488</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:17:42</td>\n",
       "<td> 3.414 sec</td>\n",
       "<td>35.0</td>\n",
       "<td>0.1241519</td>\n",
       "<td>0.0704037</td>\n",
       "<td>0.9846927</td>\n",
       "<td>0.9634613</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0123084</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:17:42</td>\n",
       "<td> 3.512 sec</td>\n",
       "<td>36.0</td>\n",
       "<td>0.1241937</td>\n",
       "<td>0.0705211</td>\n",
       "<td>0.9848243</td>\n",
       "<td>0.9637418</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0122540</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:17:42</td>\n",
       "<td> 3.605 sec</td>\n",
       "<td>37.0</td>\n",
       "<td>0.1240626</td>\n",
       "<td>0.0703471</td>\n",
       "<td>0.9849451</td>\n",
       "<td>0.9638720</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0122585</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:17:42</td>\n",
       "<td> 3.695 sec</td>\n",
       "<td>38.0</td>\n",
       "<td>0.1240050</td>\n",
       "<td>0.0702635</td>\n",
       "<td>0.9850736</td>\n",
       "<td>0.9640524</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0122268</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:17:42</td>\n",
       "<td> 3.789 sec</td>\n",
       "<td>39.0</td>\n",
       "<td>0.1236234</td>\n",
       "<td>0.0698426</td>\n",
       "<td>0.9853275</td>\n",
       "<td>0.9645528</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0119913</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:17:42</td>\n",
       "<td> 3.890 sec</td>\n",
       "<td>40.0</td>\n",
       "<td>0.1237158</td>\n",
       "<td>0.0699359</td>\n",
       "<td>0.9854852</td>\n",
       "<td>0.9647149</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0120366</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:17:43</td>\n",
       "<td> 3.990 sec</td>\n",
       "<td>41.0</td>\n",
       "<td>0.1235707</td>\n",
       "<td>0.0697993</td>\n",
       "<td>0.9854160</td>\n",
       "<td>0.9648054</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0120773</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:17:47</td>\n",
       "<td> 8.054 sec</td>\n",
       "<td>101.0</td>\n",
       "<td>0.1229161</td>\n",
       "<td>0.0692664</td>\n",
       "<td>0.9885262</td>\n",
       "<td>0.9700442</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0111758</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:17:50</td>\n",
       "<td>11.336 sec</td>\n",
       "<td>150.0</td>\n",
       "<td>0.1223215</td>\n",
       "<td>0.0685686</td>\n",
       "<td>0.9895241</td>\n",
       "<td>0.9714033</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0107364</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "<pre style='font-size: smaller; margin-bottom: 1em;'>[44 rows x 10 columns]</pre></div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-44.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-44 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-44 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-44 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-44 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-44 .h2o-table th,\n",
       "#h2o-table-44 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-44 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-44\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Variable Importances: </caption>\n",
       "    <thead><tr><th>variable</th>\n",
       "<th>relative_importance</th>\n",
       "<th>scaled_importance</th>\n",
       "<th>percentage</th></tr></thead>\n",
       "    <tbody><tr><td>RollingMean_400</td>\n",
       "<td>335381.9062500</td>\n",
       "<td>1.0</td>\n",
       "<td>0.1766520</td></tr>\n",
       "<tr><td>RollingMean_800</td>\n",
       "<td>314337.0937500</td>\n",
       "<td>0.9372512</td>\n",
       "<td>0.1655673</td></tr>\n",
       "<tr><td>RollingSD_800</td>\n",
       "<td>279706.3437500</td>\n",
       "<td>0.8339935</td>\n",
       "<td>0.1473266</td></tr>\n",
       "<tr><td>RollingMean_200</td>\n",
       "<td>177549.9218750</td>\n",
       "<td>0.5293962</td>\n",
       "<td>0.0935189</td></tr>\n",
       "<tr><td>RollingSD_400</td>\n",
       "<td>161735.9531250</td>\n",
       "<td>0.4822441</td>\n",
       "<td>0.0851894</td></tr>\n",
       "<tr><td>RollingSD_200</td>\n",
       "<td>80232.1796875</td>\n",
       "<td>0.2392263</td>\n",
       "<td>0.0422598</td></tr>\n",
       "<tr><td>RollingMean_100</td>\n",
       "<td>52751.4296875</td>\n",
       "<td>0.1572876</td>\n",
       "<td>0.0277852</td></tr>\n",
       "<tr><td>RollingSD_100</td>\n",
       "<td>45941.6679688</td>\n",
       "<td>0.1369831</td>\n",
       "<td>0.0241983</td></tr>\n",
       "<tr><td>RollingSD_50</td>\n",
       "<td>42282.7851562</td>\n",
       "<td>0.1260735</td>\n",
       "<td>0.0222711</td></tr>\n",
       "<tr><td>meanaddsd</td>\n",
       "<td>37747.6835938</td>\n",
       "<td>0.1125513</td>\n",
       "<td>0.0198824</td></tr>\n",
       "<tr><td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td></tr>\n",
       "<tr><td>Lag2</td>\n",
       "<td>12841.0175781</td>\n",
       "<td>0.0382877</td>\n",
       "<td>0.0067636</td></tr>\n",
       "<tr><td>Lag1</td>\n",
       "<td>12321.4951172</td>\n",
       "<td>0.0367387</td>\n",
       "<td>0.0064900</td></tr>\n",
       "<tr><td>RollingSD_5</td>\n",
       "<td>12221.3544922</td>\n",
       "<td>0.0364401</td>\n",
       "<td>0.0064372</td></tr>\n",
       "<tr><td>RollingSD_3</td>\n",
       "<td>8937.7216797</td>\n",
       "<td>0.0266494</td>\n",
       "<td>0.0047077</td></tr>\n",
       "<tr><td>diff50</td>\n",
       "<td>8465.0410156</td>\n",
       "<td>0.0252400</td>\n",
       "<td>0.0044587</td></tr>\n",
       "<tr><td>diff30</td>\n",
       "<td>6764.5805664</td>\n",
       "<td>0.0201698</td>\n",
       "<td>0.0035630</td></tr>\n",
       "<tr><td>RollingSD_2</td>\n",
       "<td>5055.5351562</td>\n",
       "<td>0.0150740</td>\n",
       "<td>0.0026628</td></tr>\n",
       "<tr><td>diff5</td>\n",
       "<td>3792.6650391</td>\n",
       "<td>0.0113085</td>\n",
       "<td>0.0019977</td></tr>\n",
       "<tr><td>diff2</td>\n",
       "<td>2534.2268066</td>\n",
       "<td>0.0075562</td>\n",
       "<td>0.0013348</td></tr>\n",
       "<tr><td>diff1</td>\n",
       "<td>1739.4362793</td>\n",
       "<td>0.0051864</td>\n",
       "<td>0.0009162</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "<pre style='font-size: smaller; margin-bottom: 1em;'>[36 rows x 4 columns]</pre></div><pre style=\"font-size: smaller; margin: 1em 0 0 0;\">\n",
       "\n",
       "[tips]\n",
       "Use `model.explain()` to inspect the model.\n",
       "--\n",
       "Use `h2o.display.toggle_user_tips()` to switch on/off this section.</pre>"
      ],
      "text/plain": [
       "Model Details\n",
       "=============\n",
       "H2ORandomForestEstimator : Distributed Random Forest\n",
       "Model Key: DRF_model_python_1680256840616_211\n",
       "\n",
       "\n",
       "Model Summary: \n",
       "    number_of_trees    number_of_internal_trees    model_size_in_bytes    min_depth    max_depth    mean_depth    min_leaves    max_leaves    mean_leaves\n",
       "--  -----------------  --------------------------  ---------------------  -----------  -----------  ------------  ------------  ------------  -------------\n",
       "    150                150                         1.02862e+06            15           15           15            315           912           539.873\n",
       "\n",
       "ModelMetricsBinomial: drf\n",
       "** Reported on train data. **\n",
       "\n",
       "MSE: 0.01496256045483796\n",
       "RMSE: 0.1223215453419305\n",
       "LogLoss: 0.06856855758096712\n",
       "Mean Per-Class Error: 0.040119712474570184\n",
       "AUC: 0.9895241336871461\n",
       "AUCPR: 0.9714032767310343\n",
       "Gini: 0.9790482673742922\n",
       "\n",
       "Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.155163909279127\n",
       "       0       1      Error    Rate\n",
       "-----  ------  -----  -------  -----------------\n",
       "0      194321  318    0.0016   (318.0/194639.0)\n",
       "1      2052    24053  0.0786   (2052.0/26105.0)\n",
       "Total  196373  24371  0.0107   (2370.0/220744.0)\n",
       "\n",
       "Maximum Metrics: Maximum metrics at their respective thresholds\n",
       "metric                       threshold    value     idx\n",
       "---------------------------  -----------  --------  -----\n",
       "max f1                       0.155164     0.953047  244\n",
       "max f2                       0.108594     0.937856  258\n",
       "max f0point5                 0.218037     0.976318  227\n",
       "max accuracy                 0.155164     0.989264  244\n",
       "max precision                0.999977     1         0\n",
       "max recall                   0.0136117    1         374\n",
       "max specificity              0.999977     1         0\n",
       "max absolute_mcc             0.155164     0.947707  244\n",
       "max min_per_class_accuracy   0.05271      0.954338  295\n",
       "max mean_per_class_accuracy  0.0867589    0.964221  269\n",
       "max tns                      0.999977     194639    0\n",
       "max fns                      0.999977     19082     0\n",
       "max fps                      0.000716803  194639    399\n",
       "max tps                      0.0136117    26105     374\n",
       "max tnr                      0.999977     1         0\n",
       "max fnr                      0.999977     0.730971  0\n",
       "max fpr                      0.000716803  1         399\n",
       "max tpr                      0.0136117    1         374\n",
       "\n",
       "Gains/Lift Table: Avg response rate: 11.83 %, avg score: 11.79 %\n",
       "group    cumulative_data_fraction    lower_threshold    lift        cumulative_lift    response_rate    score       cumulative_response_rate    cumulative_score    capture_rate    cumulative_capture_rate    gain      cumulative_gain    kolmogorov_smirnov\n",
       "-------  --------------------------  -----------------  ----------  -----------------  ---------------  ----------  --------------------------  ------------------  --------------  -------------------------  --------  -----------------  --------------------\n",
       "1        0.0278377                   1                  8.456       8.456              1                1           1                           1                   0.235396        0.235396                   745.6     745.6              0.235396\n",
       "2        0.0300031                   0.999849           8.456       8.456              1                0.999909    1                           0.999993            0.0183107       0.253706                   745.6     745.6              0.253706\n",
       "3        0.0400011                   0.996928           8.456       8.456              1                0.998773    1                           0.999688            0.0845432       0.338249                   745.6     745.6              0.338249\n",
       "4        0.0500036                   0.98672            8.456       8.456              1                0.993352    1                           0.998421            0.0845815       0.422831                   745.6     745.6              0.422831\n",
       "5        0.100003                    0.498751           8.44988     8.45294            0.999275         0.865069    0.999638                    0.931748            0.422486        0.845317                   744.988   745.294            0.845276\n",
       "6        0.150002                    0.0529423          2.16974     6.3586             0.256591         0.122704    0.751963                    0.662075            0.108485        0.953802                   116.974   535.86             0.911606\n",
       "7        0.200001                    0.0395958          0.268919    4.83622            0.0318021        0.0450081   0.571927                    0.507812            0.0134457       0.967248                   -73.1081  383.622            0.87015\n",
       "8        0.299999                    0.0297699          0.142887    3.2718             0.0168977        0.0338603   0.38692                     0.34983             0.0142885       0.981536                   -85.7113  227.18             0.772945\n",
       "9        0.400002                    0.0255474          0.0877203   2.47576            0.0103737        0.0274375   0.292781                    0.26923             0.00877227      0.990308                   -91.228   147.576            0.669479\n",
       "10       0.5                         0.0227505          0.0417552   1.98897            0.00493794       0.024094    0.235214                    0.220204            0.00417545      0.994484                   -95.8245  98.8968            0.560804\n",
       "11       0.599998                    0.0202302          0.0287306   1.66227            0.00339766       0.0214862   0.196578                    0.187085            0.00287301      0.997357                   -97.1269  66.2266            0.450652\n",
       "12       0.700001                    0.0176692          0.0141731   1.42682            0.0016761        0.018954    0.168734                    0.163065            0.00141735      0.998774                   -98.5827  42.6818            0.338845\n",
       "13       0.799999                    0.0149531          0.010343    1.24976            0.00122316       0.0163162   0.147796                    0.144722            0.00103428      0.999808                   -98.9657  24.9762            0.226608\n",
       "14       0.899997                    0.0116953          0.00191538  1.11111            0.000226511      0.0134797   0.131399                    0.13014             0.000191534     1                          -99.8085  11.1114            0.113415\n",
       "15       1                           0                  0           1                  0                0.00799879  0.118259                    0.117925            0               1                          -100      0                  0\n",
       "\n",
       "Scoring History: \n",
       "     timestamp            duration    number_of_trees    training_rmse        training_logloss     training_auc        training_pr_auc     training_lift      training_classification_error\n",
       "---  -------------------  ----------  -----------------  -------------------  -------------------  ------------------  ------------------  -----------------  -------------------------------\n",
       "     2023-03-31 12:17:39  0.003 sec   0.0                nan                  nan                  nan                 nan                 nan                nan\n",
       "     2023-03-31 12:17:39  0.162 sec   1.0                0.1574063208421387   0.1455872461279306   0.9284722098487775  0.8717685829699872  8.396504564475354  0.026734024115933477\n",
       "     2023-03-31 12:17:39  0.260 sec   2.0                0.14327639361634512  0.13146896434209804  0.9509009918139275  0.9028029795458034  8.404233140105305  0.021621865280302917\n",
       "     2023-03-31 12:17:39  0.334 sec   3.0                0.14437872630606932  0.12225608845953997  0.9518250721316519  0.9050886770327154  8.417752783973894  0.021460005803269174\n",
       "     2023-03-31 12:17:39  0.421 sec   4.0                0.14357114065226503  0.11049162031517538  0.9552470059485083  0.910936959271854   8.431434551105674  0.020806416881998278\n",
       "     2023-03-31 12:17:39  0.511 sec   5.0                0.13886399125695764  0.10101499632639407  0.9613219080807427  0.9219568000200676  8.431447490045286  0.019330162804156916\n",
       "     2023-03-31 12:17:39  0.600 sec   6.0                0.134886143512205    0.09224888697172161  0.966599289431064   0.9312009873420324  8.43716708883797   0.01819694492546121\n",
       "     2023-03-31 12:17:39  0.688 sec   7.0                0.13327707685378112  0.08729023212547198  0.9686599096712087  0.9353212583403849  8.442815076182383  0.0179897376830955\n",
       "     2023-03-31 12:17:39  0.779 sec   8.0                0.132511656077061    0.08398381620182795  0.9702155206209488  0.9381654495592706  8.447351609512644  0.017525351575933896\n",
       "     2023-03-31 12:17:39  0.876 sec   9.0                0.13180918415280324  0.08298825222794891  0.9711841577517661  0.9405422040131678  8.447530054161646  0.017074417383856663\n",
       "---  ---                  ---         ---                ---                  ---                  ---                 ---                 ---                ---\n",
       "     2023-03-31 12:17:42  3.323 sec   34.0               0.12467152022007755  0.07100234908390904  0.9839536552623469  0.9624265142968197  8.456004596820533  0.012448809480665386\n",
       "     2023-03-31 12:17:42  3.414 sec   35.0               0.12415188420537882  0.07040369100491614  0.9846927144510441  0.9634613217799415  8.456004596820533  0.012308375312579277\n",
       "     2023-03-31 12:17:42  3.512 sec   36.0               0.1241937496750656   0.0705211222145962   0.9848243203899526  0.9637418413529162  8.456004596820533  0.01225401369912659\n",
       "     2023-03-31 12:17:42  3.605 sec   37.0               0.12406261365226756  0.0703471353213998   0.9849450502327609  0.9638720458201931  8.456004596820533  0.01225854383358098\n",
       "     2023-03-31 12:17:42  3.695 sec   38.0               0.12400499171158319  0.07026353490944998  0.9850736478374265  0.9640523528304471  8.456004596820533  0.012226832892400246\n",
       "     2023-03-31 12:17:42  3.789 sec   39.0               0.1236233862460194   0.06984263007786067  0.9853275335937159  0.9645528201714523  8.456004596820533  0.011991265900771935\n",
       "     2023-03-31 12:17:42  3.890 sec   40.0               0.12371582013147941  0.06993594852701282  0.9854851785346985  0.9647149090837426  8.456004596820533  0.01203656724531584\n",
       "     2023-03-31 12:17:43  3.990 sec   41.0               0.12357067401634717  0.06979930770686929  0.9854160052488116  0.9648054486149535  8.456004596820533  0.012077338455405356\n",
       "     2023-03-31 12:17:47  8.054 sec   101.0              0.12291613096414601  0.06926637267891626  0.9885261524810547  0.9700442055365676  8.456004596820533  0.011175841698981625\n",
       "     2023-03-31 12:17:50  11.336 sec  150.0              0.1223215453419305   0.06856855758096712  0.9895241336871461  0.9714032767310343  8.456004596820533  0.010736418656905737\n",
       "[44 rows x 10 columns]\n",
       "\n",
       "\n",
       "Variable Importances: \n",
       "variable         relative_importance    scaled_importance     percentage\n",
       "---------------  ---------------------  --------------------  ---------------------\n",
       "RollingMean_400  335381.90625           1.0                   0.17665201440520745\n",
       "RollingMean_800  314337.09375           0.9372511989829505    0.16556731230403413\n",
       "RollingSD_800    279706.34375           0.8339935415045963    0.14732663910771993\n",
       "RollingMean_200  177549.921875          0.5293962451947272    0.09351891313220168\n",
       "RollingSD_400    161735.953125          0.4822441226284848    0.08518939569739373\n",
       "RollingSD_200    80232.1796875          0.2392263213737399    0.04225981156941869\n",
       "RollingMean_100  52751.4296875          0.15728764344304874   0.02778517905526258\n",
       "RollingSD_100    45941.66796875         0.1369831440295536    0.024198348332379313\n",
       "RollingSD_50     42282.78515625         0.12607354293207343   0.022271145322152176\n",
       "meanaddsd        37747.68359375         0.11255134188906471   0.01988242126871249\n",
       "---              ---                    ---                   ---\n",
       "Lag2             12841.017578125        0.03828774700968234   0.006763607636297343\n",
       "Lag1             12321.4951171875       0.036738699636356724  0.006489965297390278\n",
       "RollingSD_5      12221.3544921875       0.03644011279212383   0.006437219329881643\n",
       "RollingSD_3      8937.7216796875        0.026649385411463295  0.004707667615595739\n",
       "diff50           8465.041015625         0.025240005074438927  0.0044586977399972945\n",
       "diff30           6764.58056640625       0.020169783880242497  0.0035630329525625187\n",
       "RollingSD_2      5055.53515625          0.015073965118683263  0.0026628463032892305\n",
       "diff5            3792.6650390625        0.011308496279567854  0.0019976686476794554\n",
       "diff2            2534.226806640625      0.007556241882505028  0.0013348253498775102\n",
       "diff1            1739.436279296875      0.005186434470320729  0.0009161940967627618\n",
       "[36 rows x 4 columns]\n",
       "\n",
       "\n",
       "[tips]\n",
       "Use `model.explain()` to inspect the model.\n",
       "--\n",
       "Use `h2o.display.toggle_user_tips()` to switch on/off this section."
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf_2model = H2ORandomForestEstimator(seed=1,mtries=5,max_depth=15,ntrees=150,histogram_type='Random')\n",
    "rf_2model.train(x=predictors, y=response, training_frame=train_h2o)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "69dd6fe4-7529-460a-a589-7d355f78ef9a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<pre style='margin: 1em 0 1em 0;'>ModelMetricsBinomial: drf\n",
       "** Reported on test data. **\n",
       "\n",
       "MSE: 0.011199105885831109\n",
       "RMSE: 0.10582582806588904\n",
       "LogLoss: 0.05972223229578234\n",
       "Mean Per-Class Error: 0.08179715699380495\n",
       "AUC: 0.9913642944793434\n",
       "AUCPR: 0.9180687174741737\n",
       "Gini: 0.9827285889586868</pre>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-45.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-45 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-45 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-45 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-45 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-45 .h2o-table th,\n",
       "#h2o-table-45 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-45 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-45\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.4568725212427132</caption>\n",
       "    <thead><tr><th></th>\n",
       "<th>0</th>\n",
       "<th>1</th>\n",
       "<th>Error</th>\n",
       "<th>Rate</th></tr></thead>\n",
       "    <tbody><tr><td>0</td>\n",
       "<td>95150.0</td>\n",
       "<td>648.0</td>\n",
       "<td>0.0068</td>\n",
       "<td> (648.0/95798.0)</td></tr>\n",
       "<tr><td>1</td>\n",
       "<td>659.0</td>\n",
       "<td>3543.0</td>\n",
       "<td>0.1568</td>\n",
       "<td> (659.0/4202.0)</td></tr>\n",
       "<tr><td>Total</td>\n",
       "<td>95809.0</td>\n",
       "<td>4191.0</td>\n",
       "<td>0.0131</td>\n",
       "<td> (1307.0/100000.0)</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-46.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-46 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-46 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-46 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-46 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-46 .h2o-table th,\n",
       "#h2o-table-46 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-46 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-46\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Maximum Metrics: Maximum metrics at their respective thresholds</caption>\n",
       "    <thead><tr><th>metric</th>\n",
       "<th>threshold</th>\n",
       "<th>value</th>\n",
       "<th>idx</th></tr></thead>\n",
       "    <tbody><tr><td>max f1</td>\n",
       "<td>0.4568725</td>\n",
       "<td>0.8442750</td>\n",
       "<td>126.0</td></tr>\n",
       "<tr><td>max f2</td>\n",
       "<td>0.3601684</td>\n",
       "<td>0.8555027</td>\n",
       "<td>151.0</td></tr>\n",
       "<tr><td>max f0point5</td>\n",
       "<td>0.8752289</td>\n",
       "<td>0.8912378</td>\n",
       "<td>49.0</td></tr>\n",
       "<tr><td>max accuracy</td>\n",
       "<td>0.5584166</td>\n",
       "<td>0.98703</td>\n",
       "<td>104.0</td></tr>\n",
       "<tr><td>max precision</td>\n",
       "<td>0.9987217</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max recall</td>\n",
       "<td>0.0218799</td>\n",
       "<td>1.0</td>\n",
       "<td>360.0</td></tr>\n",
       "<tr><td>max specificity</td>\n",
       "<td>0.9987217</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max absolute_mcc</td>\n",
       "<td>0.4568725</td>\n",
       "<td>0.8374545</td>\n",
       "<td>126.0</td></tr>\n",
       "<tr><td>max min_per_class_accuracy</td>\n",
       "<td>0.1141716</td>\n",
       "<td>0.9576392</td>\n",
       "<td>244.0</td></tr>\n",
       "<tr><td>max mean_per_class_accuracy</td>\n",
       "<td>0.1282548</td>\n",
       "<td>0.9589237</td>\n",
       "<td>236.0</td></tr>\n",
       "<tr><td>max tns</td>\n",
       "<td>0.9987217</td>\n",
       "<td>95798.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fns</td>\n",
       "<td>0.9987217</td>\n",
       "<td>4053.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fps</td>\n",
       "<td>0.0016808</td>\n",
       "<td>95798.0</td>\n",
       "<td>399.0</td></tr>\n",
       "<tr><td>max tps</td>\n",
       "<td>0.0218799</td>\n",
       "<td>4202.0</td>\n",
       "<td>360.0</td></tr>\n",
       "<tr><td>max tnr</td>\n",
       "<td>0.9987217</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fnr</td>\n",
       "<td>0.9987217</td>\n",
       "<td>0.9645407</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fpr</td>\n",
       "<td>0.0016808</td>\n",
       "<td>1.0</td>\n",
       "<td>399.0</td></tr>\n",
       "<tr><td>max tpr</td>\n",
       "<td>0.0218799</td>\n",
       "<td>1.0</td>\n",
       "<td>360.0</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-47.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-47 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-47 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-47 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-47 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-47 .h2o-table th,\n",
       "#h2o-table-47 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-47 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-47\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Gains/Lift Table: Avg response rate:  4.20 %, avg score:  7.24 %</caption>\n",
       "    <thead><tr><th>group</th>\n",
       "<th>cumulative_data_fraction</th>\n",
       "<th>lower_threshold</th>\n",
       "<th>lift</th>\n",
       "<th>cumulative_lift</th>\n",
       "<th>response_rate</th>\n",
       "<th>score</th>\n",
       "<th>cumulative_response_rate</th>\n",
       "<th>cumulative_score</th>\n",
       "<th>capture_rate</th>\n",
       "<th>cumulative_capture_rate</th>\n",
       "<th>gain</th>\n",
       "<th>cumulative_gain</th>\n",
       "<th>kolmogorov_smirnov</th></tr></thead>\n",
       "    <tbody><tr><td>1</td>\n",
       "<td>0.01012</td>\n",
       "<td>0.9938727</td>\n",
       "<td>23.7981913</td>\n",
       "<td>23.7981913</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9971179</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9971179</td>\n",
       "<td>0.2408377</td>\n",
       "<td>0.2408377</td>\n",
       "<td>2279.8191337</td>\n",
       "<td>2279.8191337</td>\n",
       "<td>0.2408377</td></tr>\n",
       "<tr><td>2</td>\n",
       "<td>0.02</td>\n",
       "<td>0.9721521</td>\n",
       "<td>23.6536679</td>\n",
       "<td>23.7267968</td>\n",
       "<td>0.9939271</td>\n",
       "<td>0.9851422</td>\n",
       "<td>0.997</td>\n",
       "<td>0.9912019</td>\n",
       "<td>0.2336982</td>\n",
       "<td>0.4745359</td>\n",
       "<td>2265.3667908</td>\n",
       "<td>2272.6796763</td>\n",
       "<td>0.4744733</td></tr>\n",
       "<tr><td>3</td>\n",
       "<td>0.03</td>\n",
       "<td>0.8942300</td>\n",
       "<td>20.6330319</td>\n",
       "<td>22.6955418</td>\n",
       "<td>0.867</td>\n",
       "<td>0.9375335</td>\n",
       "<td>0.9536667</td>\n",
       "<td>0.9733124</td>\n",
       "<td>0.2063303</td>\n",
       "<td>0.6808663</td>\n",
       "<td>1963.3031890</td>\n",
       "<td>2169.5541805</td>\n",
       "<td>0.6794153</td></tr>\n",
       "<tr><td>4</td>\n",
       "<td>0.04</td>\n",
       "<td>0.5098831</td>\n",
       "<td>13.9695383</td>\n",
       "<td>20.5140409</td>\n",
       "<td>0.587</td>\n",
       "<td>0.7039227</td>\n",
       "<td>0.862</td>\n",
       "<td>0.9059650</td>\n",
       "<td>0.1396954</td>\n",
       "<td>0.8205616</td>\n",
       "<td>1296.9538315</td>\n",
       "<td>1951.4040933</td>\n",
       "<td>0.8147995</td></tr>\n",
       "<tr><td>5</td>\n",
       "<td>0.05</td>\n",
       "<td>0.2986572</td>\n",
       "<td>6.2827225</td>\n",
       "<td>17.6677772</td>\n",
       "<td>0.264</td>\n",
       "<td>0.3875810</td>\n",
       "<td>0.7424</td>\n",
       "<td>0.8022882</td>\n",
       "<td>0.0628272</td>\n",
       "<td>0.8833889</td>\n",
       "<td>528.2722513</td>\n",
       "<td>1666.7777249</td>\n",
       "<td>0.8699439</td></tr>\n",
       "<tr><td>6</td>\n",
       "<td>0.1</td>\n",
       "<td>0.0823182</td>\n",
       "<td>1.6991909</td>\n",
       "<td>9.6834841</td>\n",
       "<td>0.0714</td>\n",
       "<td>0.1497297</td>\n",
       "<td>0.4069</td>\n",
       "<td>0.4760089</td>\n",
       "<td>0.0849595</td>\n",
       "<td>0.9683484</td>\n",
       "<td>69.9190861</td>\n",
       "<td>868.3484055</td>\n",
       "<td>0.9064369</td></tr>\n",
       "<tr><td>7</td>\n",
       "<td>0.15</td>\n",
       "<td>0.0557939</td>\n",
       "<td>0.3188958</td>\n",
       "<td>6.5619546</td>\n",
       "<td>0.0134</td>\n",
       "<td>0.0664621</td>\n",
       "<td>0.2757333</td>\n",
       "<td>0.3394933</td>\n",
       "<td>0.0159448</td>\n",
       "<td>0.9842932</td>\n",
       "<td>-68.1104236</td>\n",
       "<td>556.1954625</td>\n",
       "<td>0.8708879</td></tr>\n",
       "<tr><td>8</td>\n",
       "<td>0.2</td>\n",
       "<td>0.0455887</td>\n",
       "<td>0.1570681</td>\n",
       "<td>4.9607330</td>\n",
       "<td>0.0066</td>\n",
       "<td>0.0501555</td>\n",
       "<td>0.20845</td>\n",
       "<td>0.2671589</td>\n",
       "<td>0.0078534</td>\n",
       "<td>0.9921466</td>\n",
       "<td>-84.2931937</td>\n",
       "<td>396.0732984</td>\n",
       "<td>0.8268926</td></tr>\n",
       "<tr><td>9</td>\n",
       "<td>0.3</td>\n",
       "<td>0.0352078</td>\n",
       "<td>0.0214184</td>\n",
       "<td>3.3142948</td>\n",
       "<td>0.0009</td>\n",
       "<td>0.0397475</td>\n",
       "<td>0.1392667</td>\n",
       "<td>0.1913551</td>\n",
       "<td>0.0021418</td>\n",
       "<td>0.9942884</td>\n",
       "<td>-97.8581628</td>\n",
       "<td>231.4294780</td>\n",
       "<td>0.7247421</td></tr>\n",
       "<tr><td>10</td>\n",
       "<td>0.4</td>\n",
       "<td>0.0293740</td>\n",
       "<td>0.0095193</td>\n",
       "<td>2.4881009</td>\n",
       "<td>0.0004</td>\n",
       "<td>0.0320013</td>\n",
       "<td>0.10455</td>\n",
       "<td>0.1515166</td>\n",
       "<td>0.0009519</td>\n",
       "<td>0.9952404</td>\n",
       "<td>-99.0480723</td>\n",
       "<td>148.8100904</td>\n",
       "<td>0.6213495</td></tr>\n",
       "<tr><td>11</td>\n",
       "<td>0.5</td>\n",
       "<td>0.0258892</td>\n",
       "<td>0.0285578</td>\n",
       "<td>1.9961923</td>\n",
       "<td>0.0012</td>\n",
       "<td>0.0275217</td>\n",
       "<td>0.08388</td>\n",
       "<td>0.1267176</td>\n",
       "<td>0.0028558</td>\n",
       "<td>0.9980961</td>\n",
       "<td>-97.1442170</td>\n",
       "<td>99.6192289</td>\n",
       "<td>0.5199442</td></tr>\n",
       "<tr><td>12</td>\n",
       "<td>0.6</td>\n",
       "<td>0.0230568</td>\n",
       "<td>0.0142789</td>\n",
       "<td>1.6658734</td>\n",
       "<td>0.0006</td>\n",
       "<td>0.0244556</td>\n",
       "<td>0.07</td>\n",
       "<td>0.1096740</td>\n",
       "<td>0.0014279</td>\n",
       "<td>0.9995240</td>\n",
       "<td>-98.5721085</td>\n",
       "<td>66.5873394</td>\n",
       "<td>0.4170484</td></tr>\n",
       "<tr><td>13</td>\n",
       "<td>0.7</td>\n",
       "<td>0.0201106</td>\n",
       "<td>0.0047596</td>\n",
       "<td>1.4285714</td>\n",
       "<td>0.0002</td>\n",
       "<td>0.0216229</td>\n",
       "<td>0.0600286</td>\n",
       "<td>0.0970952</td>\n",
       "<td>0.0004760</td>\n",
       "<td>1.0</td>\n",
       "<td>-99.5240362</td>\n",
       "<td>42.8571429</td>\n",
       "<td>0.3131589</td></tr>\n",
       "<tr><td>14</td>\n",
       "<td>0.8</td>\n",
       "<td>0.0169616</td>\n",
       "<td>0.0</td>\n",
       "<td>1.25</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0185468</td>\n",
       "<td>0.052525</td>\n",
       "<td>0.0872767</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>25.0</td>\n",
       "<td>0.2087726</td></tr>\n",
       "<tr><td>15</td>\n",
       "<td>0.9</td>\n",
       "<td>0.0136802</td>\n",
       "<td>0.0</td>\n",
       "<td>1.1111111</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0153023</td>\n",
       "<td>0.0466889</td>\n",
       "<td>0.0792795</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>11.1111111</td>\n",
       "<td>0.1043863</td></tr>\n",
       "<tr><td>16</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0003406</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0105228</td>\n",
       "<td>0.04202</td>\n",
       "<td>0.0724039</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>"
      ],
      "text/plain": [
       "ModelMetricsBinomial: drf\n",
       "** Reported on test data. **\n",
       "\n",
       "MSE: 0.011199105885831109\n",
       "RMSE: 0.10582582806588904\n",
       "LogLoss: 0.05972223229578234\n",
       "Mean Per-Class Error: 0.08179715699380495\n",
       "AUC: 0.9913642944793434\n",
       "AUCPR: 0.9180687174741737\n",
       "Gini: 0.9827285889586868\n",
       "\n",
       "Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.4568725212427132\n",
       "       0      1     Error    Rate\n",
       "-----  -----  ----  -------  -----------------\n",
       "0      95150  648   0.0068   (648.0/95798.0)\n",
       "1      659    3543  0.1568   (659.0/4202.0)\n",
       "Total  95809  4191  0.0131   (1307.0/100000.0)\n",
       "\n",
       "Maximum Metrics: Maximum metrics at their respective thresholds\n",
       "metric                       threshold    value     idx\n",
       "---------------------------  -----------  --------  -----\n",
       "max f1                       0.456873     0.844275  126\n",
       "max f2                       0.360168     0.855503  151\n",
       "max f0point5                 0.875229     0.891238  49\n",
       "max accuracy                 0.558417     0.98703   104\n",
       "max precision                0.998722     1         0\n",
       "max recall                   0.0218799    1         360\n",
       "max specificity              0.998722     1         0\n",
       "max absolute_mcc             0.456873     0.837455  126\n",
       "max min_per_class_accuracy   0.114172     0.957639  244\n",
       "max mean_per_class_accuracy  0.128255     0.958924  236\n",
       "max tns                      0.998722     95798     0\n",
       "max fns                      0.998722     4053      0\n",
       "max fps                      0.00168079   95798     399\n",
       "max tps                      0.0218799    4202      360\n",
       "max tnr                      0.998722     1         0\n",
       "max fnr                      0.998722     0.964541  0\n",
       "max fpr                      0.00168079   1         399\n",
       "max tpr                      0.0218799    1         360\n",
       "\n",
       "Gains/Lift Table: Avg response rate:  4.20 %, avg score:  7.24 %\n",
       "group    cumulative_data_fraction    lower_threshold    lift        cumulative_lift    response_rate    score      cumulative_response_rate    cumulative_score    capture_rate    cumulative_capture_rate    gain      cumulative_gain    kolmogorov_smirnov\n",
       "-------  --------------------------  -----------------  ----------  -----------------  ---------------  ---------  --------------------------  ------------------  --------------  -------------------------  --------  -----------------  --------------------\n",
       "1        0.01012                     0.993873           23.7982     23.7982            1                0.997118   1                           0.997118            0.240838        0.240838                   2279.82   2279.82            0.240838\n",
       "2        0.02                        0.972152           23.6537     23.7268            0.993927         0.985142   0.997                       0.991202            0.233698        0.474536                   2265.37   2272.68            0.474473\n",
       "3        0.03                        0.89423            20.633      22.6955            0.867            0.937534   0.953667                    0.973312            0.20633         0.680866                   1963.3    2169.55            0.679415\n",
       "4        0.04                        0.509883           13.9695     20.514             0.587            0.703923   0.862                       0.905965            0.139695        0.820562                   1296.95   1951.4             0.8148\n",
       "5        0.05                        0.298657           6.28272     17.6678            0.264            0.387581   0.7424                      0.802288            0.0628272       0.883389                   528.272   1666.78            0.869944\n",
       "6        0.1                         0.0823182          1.69919     9.68348            0.0714           0.14973    0.4069                      0.476009            0.0849595       0.968348                   69.9191   868.348            0.906437\n",
       "7        0.15                        0.0557939          0.318896    6.56195            0.0134           0.0664621  0.275733                    0.339493            0.0159448       0.984293                   -68.1104  556.195            0.870888\n",
       "8        0.2                         0.0455887          0.157068    4.96073            0.0066           0.0501555  0.20845                     0.267159            0.0078534       0.992147                   -84.2932  396.073            0.826893\n",
       "9        0.3                         0.0352078          0.0214184   3.31429            0.0009           0.0397475  0.139267                    0.191355            0.00214184      0.994288                   -97.8582  231.429            0.724742\n",
       "10       0.4                         0.029374           0.00951928  2.4881             0.0004           0.0320013  0.10455                     0.151517            0.000951928     0.99524                    -99.0481  148.81             0.621349\n",
       "11       0.5                         0.0258892          0.0285578   1.99619            0.0012           0.0275217  0.08388                     0.126718            0.00285578      0.998096                   -97.1442  99.6192            0.519944\n",
       "12       0.6                         0.0230568          0.0142789   1.66587            0.0006           0.0244556  0.07                        0.109674            0.00142789      0.999524                   -98.5721  66.5873            0.417048\n",
       "13       0.7                         0.0201106          0.00475964  1.42857            0.0002           0.0216229  0.0600286                   0.0970952           0.000475964     1                          -99.524   42.8571            0.313159\n",
       "14       0.8                         0.0169616          0           1.25               0                0.0185468  0.052525                    0.0872767           0               1                          -100      25                 0.208773\n",
       "15       0.9                         0.0136802          0           1.11111            0                0.0153023  0.0466889                   0.0792795           0               1                          -100      11.1111            0.104386\n",
       "16       1                           0.000340628        0           1                  0                0.0105228  0.04202                     0.0724039           0               1                          -100      0                  0"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf_2model.model_performance(validation_h2o)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "24c1cc97",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<h2o.plot._plot_result._MObject at 0x27fc7c84340>"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1600x900 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "rf_2model.learning_curve_plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "c7fc49ee-f6dc-407c-a13f-1a191e8cb835",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "drf Model Build progress: |██████████████████████████████████████████████████████| (done) 100%\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<pre style='margin: 1em 0 1em 0;'>Model Details\n",
       "=============\n",
       "H2ORandomForestEstimator : Distributed Random Forest\n",
       "Model Key: DRF_model_python_1680256840616_441\n",
       "</pre>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-81.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-81 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-81 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-81 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-81 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-81 .h2o-table th,\n",
       "#h2o-table-81 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-81 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-81\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Model Summary: </caption>\n",
       "    <thead><tr><th></th>\n",
       "<th>number_of_trees</th>\n",
       "<th>number_of_internal_trees</th>\n",
       "<th>model_size_in_bytes</th>\n",
       "<th>min_depth</th>\n",
       "<th>max_depth</th>\n",
       "<th>mean_depth</th>\n",
       "<th>min_leaves</th>\n",
       "<th>max_leaves</th>\n",
       "<th>mean_leaves</th></tr></thead>\n",
       "    <tbody><tr><td></td>\n",
       "<td>100.0</td>\n",
       "<td>100.0</td>\n",
       "<td>920646.0</td>\n",
       "<td>15.0</td>\n",
       "<td>15.0</td>\n",
       "<td>15.0</td>\n",
       "<td>254.0</td>\n",
       "<td>1385.0</td>\n",
       "<td>726.72</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>\n",
       "<div style='margin: 1em 0 1em 0;'><pre style='margin: 1em 0 1em 0;'>ModelMetricsBinomial: drf\n",
       "** Reported on train data. **\n",
       "\n",
       "MSE: 0.017223142490170724\n",
       "RMSE: 0.1312369707444161\n",
       "LogLoss: 0.07906734934360947\n",
       "Mean Per-Class Error: 0.04805870152344926\n",
       "AUC: 0.9847752409779693\n",
       "AUCPR: 0.9633830374600646\n",
       "Gini: 0.9695504819559386</pre>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-82.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-82 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-82 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-82 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-82 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-82 .h2o-table th,\n",
       "#h2o-table-82 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-82 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-82\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.17775036469054584</caption>\n",
       "    <thead><tr><th></th>\n",
       "<th>0</th>\n",
       "<th>1</th>\n",
       "<th>Error</th>\n",
       "<th>Rate</th></tr></thead>\n",
       "    <tbody><tr><td>0</td>\n",
       "<td>194116.0</td>\n",
       "<td>523.0</td>\n",
       "<td>0.0027</td>\n",
       "<td> (523.0/194639.0)</td></tr>\n",
       "<tr><td>1</td>\n",
       "<td>2439.0</td>\n",
       "<td>23666.0</td>\n",
       "<td>0.0934</td>\n",
       "<td> (2439.0/26105.0)</td></tr>\n",
       "<tr><td>Total</td>\n",
       "<td>196555.0</td>\n",
       "<td>24189.0</td>\n",
       "<td>0.0134</td>\n",
       "<td> (2962.0/220744.0)</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-83.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-83 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-83 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-83 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-83 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-83 .h2o-table th,\n",
       "#h2o-table-83 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-83 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-83\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Maximum Metrics: Maximum metrics at their respective thresholds</caption>\n",
       "    <thead><tr><th>metric</th>\n",
       "<th>threshold</th>\n",
       "<th>value</th>\n",
       "<th>idx</th></tr></thead>\n",
       "    <tbody><tr><td>max f1</td>\n",
       "<td>0.1777504</td>\n",
       "<td>0.9411063</td>\n",
       "<td>241.0</td></tr>\n",
       "<tr><td>max f2</td>\n",
       "<td>0.1107457</td>\n",
       "<td>0.9264197</td>\n",
       "<td>262.0</td></tr>\n",
       "<tr><td>max f0point5</td>\n",
       "<td>0.2974986</td>\n",
       "<td>0.9666050</td>\n",
       "<td>213.0</td></tr>\n",
       "<tr><td>max accuracy</td>\n",
       "<td>0.1777504</td>\n",
       "<td>0.9865817</td>\n",
       "<td>241.0</td></tr>\n",
       "<tr><td>max precision</td>\n",
       "<td>0.9999732</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max recall</td>\n",
       "<td>0.0129979</td>\n",
       "<td>1.0</td>\n",
       "<td>380.0</td></tr>\n",
       "<tr><td>max specificity</td>\n",
       "<td>0.9999732</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max absolute_mcc</td>\n",
       "<td>0.1777504</td>\n",
       "<td>0.9344107</td>\n",
       "<td>241.0</td></tr>\n",
       "<tr><td>max min_per_class_accuracy</td>\n",
       "<td>0.0630709</td>\n",
       "<td>0.9487455</td>\n",
       "<td>294.0</td></tr>\n",
       "<tr><td>max mean_per_class_accuracy</td>\n",
       "<td>0.0980931</td>\n",
       "<td>0.9584978</td>\n",
       "<td>269.0</td></tr>\n",
       "<tr><td>max tns</td>\n",
       "<td>0.9999732</td>\n",
       "<td>194639.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fns</td>\n",
       "<td>0.9999732</td>\n",
       "<td>21189.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fps</td>\n",
       "<td>0.0006725</td>\n",
       "<td>194639.0</td>\n",
       "<td>399.0</td></tr>\n",
       "<tr><td>max tps</td>\n",
       "<td>0.0129979</td>\n",
       "<td>26105.0</td>\n",
       "<td>380.0</td></tr>\n",
       "<tr><td>max tnr</td>\n",
       "<td>0.9999732</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fnr</td>\n",
       "<td>0.9999732</td>\n",
       "<td>0.8116836</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fpr</td>\n",
       "<td>0.0006725</td>\n",
       "<td>1.0</td>\n",
       "<td>399.0</td></tr>\n",
       "<tr><td>max tpr</td>\n",
       "<td>0.0129979</td>\n",
       "<td>1.0</td>\n",
       "<td>380.0</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-84.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-84 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-84 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-84 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-84 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-84 .h2o-table th,\n",
       "#h2o-table-84 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-84 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-84\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Gains/Lift Table: Avg response rate: 11.83 %, avg score: 11.79 %</caption>\n",
       "    <thead><tr><th>group</th>\n",
       "<th>cumulative_data_fraction</th>\n",
       "<th>lower_threshold</th>\n",
       "<th>lift</th>\n",
       "<th>cumulative_lift</th>\n",
       "<th>response_rate</th>\n",
       "<th>score</th>\n",
       "<th>cumulative_response_rate</th>\n",
       "<th>cumulative_score</th>\n",
       "<th>capture_rate</th>\n",
       "<th>cumulative_capture_rate</th>\n",
       "<th>gain</th>\n",
       "<th>cumulative_gain</th>\n",
       "<th>kolmogorov_smirnov</th></tr></thead>\n",
       "    <tbody><tr><td>1</td>\n",
       "<td>0.0188136</td>\n",
       "<td>1.0</td>\n",
       "<td>8.4560046</td>\n",
       "<td>8.4560046</td>\n",
       "<td>1.0</td>\n",
       "<td>1.0</td>\n",
       "<td>1.0</td>\n",
       "<td>1.0</td>\n",
       "<td>0.1590883</td>\n",
       "<td>0.1590883</td>\n",
       "<td>745.6004597</td>\n",
       "<td>745.6004597</td>\n",
       "<td>0.1590883</td></tr>\n",
       "<tr><td>2</td>\n",
       "<td>0.0200323</td>\n",
       "<td>0.9999365</td>\n",
       "<td>8.4560046</td>\n",
       "<td>8.4560046</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9999442</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9999966</td>\n",
       "<td>0.0103045</td>\n",
       "<td>0.1693928</td>\n",
       "<td>745.6004597</td>\n",
       "<td>745.6004597</td>\n",
       "<td>0.1693928</td></tr>\n",
       "<tr><td>3</td>\n",
       "<td>0.0300031</td>\n",
       "<td>0.9902415</td>\n",
       "<td>8.4560046</td>\n",
       "<td>8.4560046</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9972483</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9990833</td>\n",
       "<td>0.0843133</td>\n",
       "<td>0.2537062</td>\n",
       "<td>745.6004597</td>\n",
       "<td>745.6004597</td>\n",
       "<td>0.2537062</td></tr>\n",
       "<tr><td>4</td>\n",
       "<td>0.0400011</td>\n",
       "<td>0.9719905</td>\n",
       "<td>8.4560046</td>\n",
       "<td>8.4560046</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9797796</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9942585</td>\n",
       "<td>0.0845432</td>\n",
       "<td>0.3382494</td>\n",
       "<td>745.6004597</td>\n",
       "<td>745.6004597</td>\n",
       "<td>0.3382494</td></tr>\n",
       "<tr><td>5</td>\n",
       "<td>0.0500036</td>\n",
       "<td>0.9545787</td>\n",
       "<td>8.4560046</td>\n",
       "<td>8.4560046</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9641455</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9882348</td>\n",
       "<td>0.0845815</td>\n",
       "<td>0.4228309</td>\n",
       "<td>745.6004597</td>\n",
       "<td>745.6004597</td>\n",
       "<td>0.4228309</td></tr>\n",
       "<tr><td>6</td>\n",
       "<td>0.1000027</td>\n",
       "<td>0.4601459</td>\n",
       "<td>8.4054387</td>\n",
       "<td>8.4307228</td>\n",
       "<td>0.9940201</td>\n",
       "<td>0.8028230</td>\n",
       "<td>0.9970102</td>\n",
       "<td>0.8955331</td>\n",
       "<td>0.4202643</td>\n",
       "<td>0.8430952</td>\n",
       "<td>740.5438655</td>\n",
       "<td>743.0722771</td>\n",
       "<td>0.8427561</td></tr>\n",
       "<tr><td>7</td>\n",
       "<td>0.1500018</td>\n",
       "<td>0.0656330</td>\n",
       "<td>2.0594129</td>\n",
       "<td>6.3070170</td>\n",
       "<td>0.2435444</td>\n",
       "<td>0.1323688</td>\n",
       "<td>0.7458625</td>\n",
       "<td>0.6411527</td>\n",
       "<td>0.1029688</td>\n",
       "<td>0.9460640</td>\n",
       "<td>105.9412916</td>\n",
       "<td>530.7016958</td>\n",
       "<td>0.9028301</td></tr>\n",
       "<tr><td>8</td>\n",
       "<td>0.2000009</td>\n",
       "<td>0.0480428</td>\n",
       "<td>0.2696850</td>\n",
       "<td>4.7977182</td>\n",
       "<td>0.0318927</td>\n",
       "<td>0.0554065</td>\n",
       "<td>0.5673741</td>\n",
       "<td>0.4947194</td>\n",
       "<td>0.0134840</td>\n",
       "<td>0.9595480</td>\n",
       "<td>-73.0314975</td>\n",
       "<td>379.7718162</td>\n",
       "<td>0.8614176</td></tr>\n",
       "<tr><td>9</td>\n",
       "<td>0.2999991</td>\n",
       "<td>0.0359408</td>\n",
       "<td>0.1287133</td>\n",
       "<td>3.2414067</td>\n",
       "<td>0.0152215</td>\n",
       "<td>0.0409827</td>\n",
       "<td>0.3833260</td>\n",
       "<td>0.3434761</td>\n",
       "<td>0.0128711</td>\n",
       "<td>0.9724191</td>\n",
       "<td>-87.1286693</td>\n",
       "<td>224.1406712</td>\n",
       "<td>0.7626050</td></tr>\n",
       "<tr><td>10</td>\n",
       "<td>0.4000018</td>\n",
       "<td>0.0304205</td>\n",
       "<td>0.0892525</td>\n",
       "<td>2.4533503</td>\n",
       "<td>0.0105549</td>\n",
       "<td>0.0329448</td>\n",
       "<td>0.2901311</td>\n",
       "<td>0.2658415</td>\n",
       "<td>0.0089255</td>\n",
       "<td>0.9813446</td>\n",
       "<td>-91.0747494</td>\n",
       "<td>145.3350311</td>\n",
       "<td>0.6593125</td></tr>\n",
       "<tr><td>11</td>\n",
       "<td>0.5</td>\n",
       "<td>0.0263531</td>\n",
       "<td>0.0777643</td>\n",
       "<td>1.9782417</td>\n",
       "<td>0.0091963</td>\n",
       "<td>0.0283021</td>\n",
       "<td>0.2339452</td>\n",
       "<td>0.2183345</td>\n",
       "<td>0.0077763</td>\n",
       "<td>0.9891209</td>\n",
       "<td>-92.2235710</td>\n",
       "<td>97.8241716</td>\n",
       "<td>0.5547218</td></tr>\n",
       "<tr><td>12</td>\n",
       "<td>0.5999982</td>\n",
       "<td>0.0230816</td>\n",
       "<td>0.0475013</td>\n",
       "<td>1.6564565</td>\n",
       "<td>0.0056175</td>\n",
       "<td>0.0246605</td>\n",
       "<td>0.1958912</td>\n",
       "<td>0.1860560</td>\n",
       "<td>0.0047500</td>\n",
       "<td>0.9938709</td>\n",
       "<td>-95.2498660</td>\n",
       "<td>65.6456513</td>\n",
       "<td>0.4466990</td></tr>\n",
       "<tr><td>13</td>\n",
       "<td>0.7000009</td>\n",
       "<td>0.0201146</td>\n",
       "<td>0.0314108</td>\n",
       "<td>1.4243011</td>\n",
       "<td>0.0037146</td>\n",
       "<td>0.0215943</td>\n",
       "<td>0.1684367</td>\n",
       "<td>0.1625609</td>\n",
       "<td>0.0031412</td>\n",
       "<td>0.9970121</td>\n",
       "<td>-96.8589247</td>\n",
       "<td>42.4301109</td>\n",
       "<td>0.3368463</td></tr>\n",
       "<tr><td>14</td>\n",
       "<td>0.7999991</td>\n",
       "<td>0.0168511</td>\n",
       "<td>0.0187707</td>\n",
       "<td>1.2486128</td>\n",
       "<td>0.0022198</td>\n",
       "<td>0.0185265</td>\n",
       "<td>0.1476599</td>\n",
       "<td>0.1445569</td>\n",
       "<td>0.0018770</td>\n",
       "<td>0.9988891</td>\n",
       "<td>-98.1229309</td>\n",
       "<td>24.8612791</td>\n",
       "<td>0.2255652</td></tr>\n",
       "<tr><td>15</td>\n",
       "<td>0.8999973</td>\n",
       "<td>0.0120496</td>\n",
       "<td>0.0111092</td>\n",
       "<td>1.1111145</td>\n",
       "<td>0.0013138</td>\n",
       "<td>0.0146833</td>\n",
       "<td>0.1313995</td>\n",
       "<td>0.1301267</td>\n",
       "<td>0.0011109</td>\n",
       "<td>1.0</td>\n",
       "<td>-98.8890816</td>\n",
       "<td>11.1114467</td>\n",
       "<td>0.1134151</td></tr>\n",
       "<tr><td>16</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0077083</td>\n",
       "<td>0.1182592</td>\n",
       "<td>0.1178845</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div></div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-85.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-85 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-85 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-85 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-85 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-85 .h2o-table th,\n",
       "#h2o-table-85 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-85 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-85\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Scoring History: </caption>\n",
       "    <thead><tr><th></th>\n",
       "<th>timestamp</th>\n",
       "<th>duration</th>\n",
       "<th>number_of_trees</th>\n",
       "<th>training_rmse</th>\n",
       "<th>training_logloss</th>\n",
       "<th>training_auc</th>\n",
       "<th>training_pr_auc</th>\n",
       "<th>training_lift</th>\n",
       "<th>training_classification_error</th></tr></thead>\n",
       "    <tbody><tr><td></td>\n",
       "<td>2023-03-31 12:23:57</td>\n",
       "<td> 0.001 sec</td>\n",
       "<td>0.0</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:23:57</td>\n",
       "<td> 0.069 sec</td>\n",
       "<td>1.0</td>\n",
       "<td>0.1599810</td>\n",
       "<td>0.1772992</td>\n",
       "<td>0.9464337</td>\n",
       "<td>0.8844741</td>\n",
       "<td>8.3397215</td>\n",
       "<td>0.0286392</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:23:57</td>\n",
       "<td> 0.144 sec</td>\n",
       "<td>2.0</td>\n",
       "<td>0.1567302</td>\n",
       "<td>0.1763927</td>\n",
       "<td>0.9456982</td>\n",
       "<td>0.8872538</td>\n",
       "<td>8.3449362</td>\n",
       "<td>0.0265728</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:23:57</td>\n",
       "<td> 0.219 sec</td>\n",
       "<td>3.0</td>\n",
       "<td>0.1508431</td>\n",
       "<td>0.1528991</td>\n",
       "<td>0.9512210</td>\n",
       "<td>0.9002682</td>\n",
       "<td>8.3690674</td>\n",
       "<td>0.0247243</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:23:57</td>\n",
       "<td> 0.303 sec</td>\n",
       "<td>4.0</td>\n",
       "<td>0.1514469</td>\n",
       "<td>0.1394974</td>\n",
       "<td>0.9515033</td>\n",
       "<td>0.9029504</td>\n",
       "<td>8.4022575</td>\n",
       "<td>0.0244240</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:23:57</td>\n",
       "<td> 0.392 sec</td>\n",
       "<td>5.0</td>\n",
       "<td>0.1498596</td>\n",
       "<td>0.1234611</td>\n",
       "<td>0.9557570</td>\n",
       "<td>0.9108908</td>\n",
       "<td>8.4231284</td>\n",
       "<td>0.0237186</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:23:57</td>\n",
       "<td> 0.472 sec</td>\n",
       "<td>6.0</td>\n",
       "<td>0.1471990</td>\n",
       "<td>0.1134622</td>\n",
       "<td>0.9591720</td>\n",
       "<td>0.9176134</td>\n",
       "<td>8.4341199</td>\n",
       "<td>0.0224584</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:23:57</td>\n",
       "<td> 0.561 sec</td>\n",
       "<td>7.0</td>\n",
       "<td>0.1440669</td>\n",
       "<td>0.1053019</td>\n",
       "<td>0.9625341</td>\n",
       "<td>0.9244037</td>\n",
       "<td>8.4390171</td>\n",
       "<td>0.0211666</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:23:57</td>\n",
       "<td> 0.645 sec</td>\n",
       "<td>8.0</td>\n",
       "<td>0.1432628</td>\n",
       "<td>0.1005826</td>\n",
       "<td>0.9641284</td>\n",
       "<td>0.9269401</td>\n",
       "<td>8.4423091</td>\n",
       "<td>0.0206438</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:23:58</td>\n",
       "<td> 0.728 sec</td>\n",
       "<td>9.0</td>\n",
       "<td>0.1440298</td>\n",
       "<td>0.0978742</td>\n",
       "<td>0.9655278</td>\n",
       "<td>0.9290023</td>\n",
       "<td>8.4455401</td>\n",
       "<td>0.0206707</td></tr>\n",
       "<tr><td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:24:00</td>\n",
       "<td> 3.279 sec</td>\n",
       "<td>38.0</td>\n",
       "<td>0.1324668</td>\n",
       "<td>0.0802251</td>\n",
       "<td>0.9801670</td>\n",
       "<td>0.9562776</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0146958</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:24:00</td>\n",
       "<td> 3.382 sec</td>\n",
       "<td>39.0</td>\n",
       "<td>0.1319750</td>\n",
       "<td>0.0797206</td>\n",
       "<td>0.9805302</td>\n",
       "<td>0.9568684</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0145825</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:24:00</td>\n",
       "<td> 3.467 sec</td>\n",
       "<td>40.0</td>\n",
       "<td>0.1323018</td>\n",
       "<td>0.0803051</td>\n",
       "<td>0.9802591</td>\n",
       "<td>0.9564083</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0146776</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:24:00</td>\n",
       "<td> 3.553 sec</td>\n",
       "<td>41.0</td>\n",
       "<td>0.1317557</td>\n",
       "<td>0.0797153</td>\n",
       "<td>0.9807060</td>\n",
       "<td>0.9572073</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0145055</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:24:00</td>\n",
       "<td> 3.638 sec</td>\n",
       "<td>42.0</td>\n",
       "<td>0.1320714</td>\n",
       "<td>0.0800269</td>\n",
       "<td>0.9806405</td>\n",
       "<td>0.9571984</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0145146</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:24:01</td>\n",
       "<td> 3.732 sec</td>\n",
       "<td>43.0</td>\n",
       "<td>0.1318760</td>\n",
       "<td>0.0798526</td>\n",
       "<td>0.9809900</td>\n",
       "<td>0.9576144</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0143741</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:24:01</td>\n",
       "<td> 3.825 sec</td>\n",
       "<td>44.0</td>\n",
       "<td>0.1321847</td>\n",
       "<td>0.0800925</td>\n",
       "<td>0.9810728</td>\n",
       "<td>0.9578108</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0143198</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:24:01</td>\n",
       "<td> 3.914 sec</td>\n",
       "<td>45.0</td>\n",
       "<td>0.1322692</td>\n",
       "<td>0.0801752</td>\n",
       "<td>0.9810347</td>\n",
       "<td>0.9578679</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0142473</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:24:01</td>\n",
       "<td> 4.001 sec</td>\n",
       "<td>46.0</td>\n",
       "<td>0.1321338</td>\n",
       "<td>0.0800518</td>\n",
       "<td>0.9811794</td>\n",
       "<td>0.9580782</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0142337</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:24:04</td>\n",
       "<td> 7.325 sec</td>\n",
       "<td>100.0</td>\n",
       "<td>0.1312370</td>\n",
       "<td>0.0790673</td>\n",
       "<td>0.9847752</td>\n",
       "<td>0.9633830</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0134183</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "<pre style='font-size: smaller; margin-bottom: 1em;'>[48 rows x 10 columns]</pre></div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-86.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-86 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-86 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-86 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-86 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-86 .h2o-table th,\n",
       "#h2o-table-86 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-86 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-86\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Variable Importances: </caption>\n",
       "    <thead><tr><th>variable</th>\n",
       "<th>relative_importance</th>\n",
       "<th>scaled_importance</th>\n",
       "<th>percentage</th></tr></thead>\n",
       "    <tbody><tr><td>RollingMean_800</td>\n",
       "<td>175369.2343750</td>\n",
       "<td>1.0</td>\n",
       "<td>0.1472910</td></tr>\n",
       "<tr><td>RollingSD_800</td>\n",
       "<td>164485.0312500</td>\n",
       "<td>0.9379355</td>\n",
       "<td>0.1381495</td></tr>\n",
       "<tr><td>RollingMean_400</td>\n",
       "<td>160214.0</td>\n",
       "<td>0.9135810</td>\n",
       "<td>0.1345623</td></tr>\n",
       "<tr><td>RollingMean_200</td>\n",
       "<td>117217.8437500</td>\n",
       "<td>0.6684060</td>\n",
       "<td>0.0984502</td></tr>\n",
       "<tr><td>RollingSD_400</td>\n",
       "<td>102123.5937500</td>\n",
       "<td>0.5823347</td>\n",
       "<td>0.0857727</td></tr>\n",
       "<tr><td>RollingSD_200</td>\n",
       "<td>57513.2851562</td>\n",
       "<td>0.3279554</td>\n",
       "<td>0.0483049</td></tr>\n",
       "<tr><td>RollingMean_100</td>\n",
       "<td>35860.7343750</td>\n",
       "<td>0.2044870</td>\n",
       "<td>0.0301191</td></tr>\n",
       "<tr><td>RollingSD_100</td>\n",
       "<td>32440.8574219</td>\n",
       "<td>0.1849860</td>\n",
       "<td>0.0272468</td></tr>\n",
       "<tr><td>meanaddsd</td>\n",
       "<td>28556.5683594</td>\n",
       "<td>0.1628368</td>\n",
       "<td>0.0239844</td></tr>\n",
       "<tr><td>meandiffsd</td>\n",
       "<td>21476.5664062</td>\n",
       "<td>0.1224648</td>\n",
       "<td>0.0180380</td></tr>\n",
       "<tr><td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td></tr>\n",
       "<tr><td>RollingMean_5</td>\n",
       "<td>10222.2861328</td>\n",
       "<td>0.0582901</td>\n",
       "<td>0.0085856</td></tr>\n",
       "<tr><td>RollingSD_5</td>\n",
       "<td>9366.3613281</td>\n",
       "<td>0.0534094</td>\n",
       "<td>0.0078667</td></tr>\n",
       "<tr><td>diff50</td>\n",
       "<td>8207.2412109</td>\n",
       "<td>0.0467998</td>\n",
       "<td>0.0068932</td></tr>\n",
       "<tr><td>RollingSD_4</td>\n",
       "<td>7528.5673828</td>\n",
       "<td>0.0429298</td>\n",
       "<td>0.0063232</td></tr>\n",
       "<tr><td>diff30</td>\n",
       "<td>6408.7924805</td>\n",
       "<td>0.0365446</td>\n",
       "<td>0.0053827</td></tr>\n",
       "<tr><td>RollingSD_3</td>\n",
       "<td>5001.7128906</td>\n",
       "<td>0.0285210</td>\n",
       "<td>0.0042009</td></tr>\n",
       "<tr><td>RollingSD_2</td>\n",
       "<td>3857.1816406</td>\n",
       "<td>0.0219946</td>\n",
       "<td>0.0032396</td></tr>\n",
       "<tr><td>diff5</td>\n",
       "<td>3346.4116211</td>\n",
       "<td>0.0190821</td>\n",
       "<td>0.0028106</td></tr>\n",
       "<tr><td>diff2</td>\n",
       "<td>2175.0688477</td>\n",
       "<td>0.0124028</td>\n",
       "<td>0.0018268</td></tr>\n",
       "<tr><td>diff1</td>\n",
       "<td>1630.1914062</td>\n",
       "<td>0.0092958</td>\n",
       "<td>0.0013692</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "<pre style='font-size: smaller; margin-bottom: 1em;'>[36 rows x 4 columns]</pre></div><pre style=\"font-size: smaller; margin: 1em 0 0 0;\">\n",
       "\n",
       "[tips]\n",
       "Use `model.explain()` to inspect the model.\n",
       "--\n",
       "Use `h2o.display.toggle_user_tips()` to switch on/off this section.</pre>"
      ],
      "text/plain": [
       "Model Details\n",
       "=============\n",
       "H2ORandomForestEstimator : Distributed Random Forest\n",
       "Model Key: DRF_model_python_1680256840616_441\n",
       "\n",
       "\n",
       "Model Summary: \n",
       "    number_of_trees    number_of_internal_trees    model_size_in_bytes    min_depth    max_depth    mean_depth    min_leaves    max_leaves    mean_leaves\n",
       "--  -----------------  --------------------------  ---------------------  -----------  -----------  ------------  ------------  ------------  -------------\n",
       "    100                100                         920646                 15           15           15            254           1385          726.72\n",
       "\n",
       "ModelMetricsBinomial: drf\n",
       "** Reported on train data. **\n",
       "\n",
       "MSE: 0.017223142490170724\n",
       "RMSE: 0.1312369707444161\n",
       "LogLoss: 0.07906734934360947\n",
       "Mean Per-Class Error: 0.04805870152344926\n",
       "AUC: 0.9847752409779693\n",
       "AUCPR: 0.9633830374600646\n",
       "Gini: 0.9695504819559386\n",
       "\n",
       "Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.17775036469054584\n",
       "       0       1      Error    Rate\n",
       "-----  ------  -----  -------  -----------------\n",
       "0      194116  523    0.0027   (523.0/194639.0)\n",
       "1      2439    23666  0.0934   (2439.0/26105.0)\n",
       "Total  196555  24189  0.0134   (2962.0/220744.0)\n",
       "\n",
       "Maximum Metrics: Maximum metrics at their respective thresholds\n",
       "metric                       threshold    value     idx\n",
       "---------------------------  -----------  --------  -----\n",
       "max f1                       0.17775      0.941106  241\n",
       "max f2                       0.110746     0.92642   262\n",
       "max f0point5                 0.297499     0.966605  213\n",
       "max accuracy                 0.17775      0.986582  241\n",
       "max precision                0.999973     1         0\n",
       "max recall                   0.0129979    1         380\n",
       "max specificity              0.999973     1         0\n",
       "max absolute_mcc             0.17775      0.934411  241\n",
       "max min_per_class_accuracy   0.0630709    0.948745  294\n",
       "max mean_per_class_accuracy  0.0980931    0.958498  269\n",
       "max tns                      0.999973     194639    0\n",
       "max fns                      0.999973     21189     0\n",
       "max fps                      0.000672459  194639    399\n",
       "max tps                      0.0129979    26105     380\n",
       "max tnr                      0.999973     1         0\n",
       "max fnr                      0.999973     0.811684  0\n",
       "max fpr                      0.000672459  1         399\n",
       "max tpr                      0.0129979    1         380\n",
       "\n",
       "Gains/Lift Table: Avg response rate: 11.83 %, avg score: 11.79 %\n",
       "group    cumulative_data_fraction    lower_threshold    lift       cumulative_lift    response_rate    score       cumulative_response_rate    cumulative_score    capture_rate    cumulative_capture_rate    gain      cumulative_gain    kolmogorov_smirnov\n",
       "-------  --------------------------  -----------------  ---------  -----------------  ---------------  ----------  --------------------------  ------------------  --------------  -------------------------  --------  -----------------  --------------------\n",
       "1        0.0188136                   1                  8.456      8.456              1                1           1                           1                   0.159088        0.159088                   745.6     745.6              0.159088\n",
       "2        0.0200323                   0.999937           8.456      8.456              1                0.999944    1                           0.999997            0.0103045       0.169393                   745.6     745.6              0.169393\n",
       "3        0.0300031                   0.990242           8.456      8.456              1                0.997248    1                           0.999083            0.0843133       0.253706                   745.6     745.6              0.253706\n",
       "4        0.0400011                   0.97199            8.456      8.456              1                0.97978     1                           0.994258            0.0845432       0.338249                   745.6     745.6              0.338249\n",
       "5        0.0500036                   0.954579           8.456      8.456              1                0.964146    1                           0.988235            0.0845815       0.422831                   745.6     745.6              0.422831\n",
       "6        0.100003                    0.460146           8.40544    8.43072            0.99402          0.802823    0.99701                     0.895533            0.420264        0.843095                   740.544   743.072            0.842756\n",
       "7        0.150002                    0.065633           2.05941    6.30702            0.243544         0.132369    0.745863                    0.641153            0.102969        0.946064                   105.941   530.702            0.90283\n",
       "8        0.200001                    0.0480428          0.269685   4.79772            0.0318927        0.0554065   0.567374                    0.494719            0.013484        0.959548                   -73.0315  379.772            0.861418\n",
       "9        0.299999                    0.0359408          0.128713   3.24141            0.0152215        0.0409827   0.383326                    0.343476            0.0128711       0.972419                   -87.1287  224.141            0.762605\n",
       "10       0.400002                    0.0304205          0.0892525  2.45335            0.0105549        0.0329448   0.290131                    0.265842            0.00892549      0.981345                   -91.0747  145.335            0.659313\n",
       "11       0.5                         0.0263531          0.0777643  1.97824            0.00919634       0.0283021   0.233945                    0.218335            0.00777629      0.989121                   -92.2236  97.8242            0.554722\n",
       "12       0.599998                    0.0230816          0.0475013  1.65646            0.00561747       0.0246605   0.195891                    0.186056            0.00475005      0.993871                   -95.2499  65.6457            0.446699\n",
       "13       0.700001                    0.0201146          0.0314108  1.4243             0.00371461       0.0215943   0.168437                    0.162561            0.00314116      0.997012                   -96.8589  42.4301            0.336846\n",
       "14       0.799999                    0.0168511          0.0187707  1.24861            0.00221981       0.0185265   0.14766                     0.144557            0.00187704      0.998889                   -98.1229  24.8613            0.225565\n",
       "15       0.899997                    0.0120496          0.0111092  1.11111            0.00131376       0.0146833   0.131399                    0.130127            0.0011109       1                          -98.8891  11.1114            0.113415\n",
       "16       1                           0                  0          1                  0                0.00770834  0.118259                    0.117885            0               1                          -100      0                  0\n",
       "\n",
       "Scoring History: \n",
       "     timestamp            duration    number_of_trees    training_rmse        training_logloss     training_auc        training_pr_auc     training_lift      training_classification_error\n",
       "---  -------------------  ----------  -----------------  -------------------  -------------------  ------------------  ------------------  -----------------  -------------------------------\n",
       "     2023-03-31 12:23:57  0.001 sec   0.0                nan                  nan                  nan                 nan                 nan                nan\n",
       "     2023-03-31 12:23:57  0.069 sec   1.0                0.15998099369869656  0.17729917426458558  0.946433731318102   0.8844740801809722  8.339721468094169  0.028639207443735635\n",
       "     2023-03-31 12:23:57  0.144 sec   2.0                0.15673015072212412  0.17639269395068707  0.9456981570086617  0.8872537830989309  8.344936235216059  0.02657280663531321\n",
       "     2023-03-31 12:23:57  0.219 sec   3.0                0.15084308151021666  0.15289913354714935  0.9512210157477604  0.9002682131655597  8.369067360805028  0.02472434471418899\n",
       "     2023-03-31 12:23:57  0.303 sec   4.0                0.15144689508462972  0.13949743996462943  0.9515033183433383  0.9029503720049593  8.402257465232276  0.024423987941429803\n",
       "     2023-03-31 12:23:57  0.392 sec   5.0                0.14985959111332384  0.12346107400989226  0.9557569538545668  0.9108908150194558  8.423128405239584  0.023718577791198008\n",
       "     2023-03-31 12:23:57  0.472 sec   6.0                0.14719896569463312  0.11346222216265693  0.9591720319106649  0.9176134281731149  8.434119897648316  0.022458377269781074\n",
       "     2023-03-31 12:23:57  0.561 sec   7.0                0.14406692606957028  0.10530186130436651  0.9625340683575153  0.9244037105260434  8.439017087585848  0.021166618675150937\n",
       "     2023-03-31 12:23:57  0.645 sec   8.0                0.14326282207025307  0.10058257683257397  0.9641284495950238  0.9269401032694083  8.442309078038084  0.020643758074860345\n",
       "     2023-03-31 12:23:58  0.728 sec   9.0                0.1440297552516849   0.09787418679233587  0.9655277888558934  0.929002265145232   8.445540108003414  0.020670728057209427\n",
       "---  ---                  ---         ---                ---                  ---                  ---                 ---                 ---                ---\n",
       "     2023-03-31 12:24:00  3.279 sec   38.0               0.1324667932788369   0.08022514364902018  0.9801669946599897  0.956277580296344   8.456004596820533  0.014695756170043126\n",
       "     2023-03-31 12:24:00  3.382 sec   39.0               0.13197501510052345  0.07972057349484743  0.980530221080172   0.9568683746113846  8.456004596820533  0.014582502808683362\n",
       "     2023-03-31 12:24:00  3.467 sec   40.0               0.13230180276931994  0.08030512035169787  0.9802591329776816  0.9564083148211702  8.456004596820533  0.014677635632225565\n",
       "     2023-03-31 12:24:00  3.553 sec   41.0               0.13175567796885448  0.07971525955378637  0.9807059964233641  0.9572072564965031  8.456004596820533  0.01450549052295872\n",
       "     2023-03-31 12:24:00  3.638 sec   42.0               0.13207135537366962  0.08002691962604186  0.9806404708079401  0.9571983689653335  8.456004596820533  0.014514550791867503\n",
       "     2023-03-31 12:24:01  3.732 sec   43.0               0.13187601799953436  0.0798526446190322   0.9809900171846235  0.957614436041913   8.456004596820533  0.014374116623781394\n",
       "     2023-03-31 12:24:01  3.825 sec   44.0               0.1321846649248098   0.08009246199211788  0.9810728138328236  0.9578107629789451  8.456004596820533  0.014319755010328707\n",
       "     2023-03-31 12:24:01  3.914 sec   45.0               0.13226916841608144  0.08017516568838128  0.9810346813684226  0.9578678915869199  8.456004596820533  0.014247272859058456\n",
       "     2023-03-31 12:24:01  4.001 sec   46.0               0.13213381232664673  0.08005183070836955  0.9811793722967866  0.9580781851928095  8.456004596820533  0.014233682455695285\n",
       "     2023-03-31 12:24:04  7.325 sec   100.0              0.1312369707444161   0.07906734934360947  0.9847752409779693  0.9633830374600646  8.456004596820533  0.013418258253904976\n",
       "[48 rows x 10 columns]\n",
       "\n",
       "\n",
       "Variable Importances: \n",
       "variable         relative_importance    scaled_importance     percentage\n",
       "---------------  ---------------------  --------------------  ---------------------\n",
       "RollingMean_800  175369.234375          1.0                   0.14729103873119173\n",
       "RollingSD_800    164485.03125           0.9379355041162704    0.1381494946641494\n",
       "RollingMean_400  160214.0               0.9135809970944341    0.13456229402711703\n",
       "RollingMean_200  117217.84375           0.6684059730759145    0.09845021006848442\n",
       "RollingSD_400    102123.59375           0.5823347185950786    0.08577268559110535\n",
       "RollingSD_200    57513.28515625         0.32795538716481326   0.04830488963299548\n",
       "RollingMean_100  35860.734375           0.20448703276150115   0.030119107462500738\n",
       "RollingSD_100    32440.857421875        0.18498602413069354   0.027246783644963146\n",
       "meanaddsd        28556.568359375        0.1628368194749097    0.023984404284142995\n",
       "meandiffsd       21476.56640625         0.12246484671493565   0.018037974480699045\n",
       "---              ---                    ---                   ---\n",
       "RollingMean_5    10222.2861328125       0.05829007675858196   0.008585605953492434\n",
       "RollingSD_5      9366.361328125         0.053409375718071986  0.007866722427499312\n",
       "diff50           8207.2412109375        0.04679977785263967   0.0068931878923043175\n",
       "RollingSD_4      7528.5673828125        0.04292980698492315   0.006323175863338901\n",
       "diff30           6408.79248046875       0.036544565546568666  0.00538268701933422\n",
       "RollingSD_3      5001.712890625         0.02852103967067342   0.004200893558787015\n",
       "RollingSD_2      3857.181640625         0.02199463123832207   0.003239612081601976\n",
       "diff5            3346.41162109375       0.019082090613100164  0.002810620947566246\n",
       "diff2            2175.06884765625       0.012402796051473895  0.0018268207135927132\n",
       "diff1            1630.19140625          0.009295766227524765  0.001369183063454454\n",
       "[36 rows x 4 columns]\n",
       "\n",
       "\n",
       "[tips]\n",
       "Use `model.explain()` to inspect the model.\n",
       "--\n",
       "Use `h2o.display.toggle_user_tips()` to switch on/off this section."
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf_3model = H2ORandomForestEstimator(seed=1,mtries=3,max_depth=15,ntrees=100,histogram_type='Random')\n",
    "rf_3model.train(x=predictors, y=response, training_frame=train_h2o)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "551a1867",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<pre style='margin: 1em 0 1em 0;'>ModelMetricsBinomial: drf\n",
       "** Reported on test data. **\n",
       "\n",
       "MSE: 0.010869869336460194\n",
       "RMSE: 0.10425866552215309\n",
       "LogLoss: 0.06270051806376054\n",
       "Mean Per-Class Error: 0.08321662204917754\n",
       "AUC: 0.9899913933708621\n",
       "AUCPR: 0.9168687082576839\n",
       "Gini: 0.9799827867417241</pre>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-90.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-90 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-90 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-90 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-90 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-90 .h2o-table th,\n",
       "#h2o-table-90 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-90 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-90\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.49762779177502836</caption>\n",
       "    <thead><tr><th></th>\n",
       "<th>0</th>\n",
       "<th>1</th>\n",
       "<th>Error</th>\n",
       "<th>Rate</th></tr></thead>\n",
       "    <tbody><tr><td>0</td>\n",
       "<td>95334.0</td>\n",
       "<td>464.0</td>\n",
       "<td>0.0048</td>\n",
       "<td> (464.0/95798.0)</td></tr>\n",
       "<tr><td>1</td>\n",
       "<td>679.0</td>\n",
       "<td>3523.0</td>\n",
       "<td>0.1616</td>\n",
       "<td> (679.0/4202.0)</td></tr>\n",
       "<tr><td>Total</td>\n",
       "<td>96013.0</td>\n",
       "<td>3987.0</td>\n",
       "<td>0.0114</td>\n",
       "<td> (1143.0/100000.0)</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-91.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-91 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-91 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-91 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-91 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-91 .h2o-table th,\n",
       "#h2o-table-91 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-91 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-91\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Maximum Metrics: Maximum metrics at their respective thresholds</caption>\n",
       "    <thead><tr><th>metric</th>\n",
       "<th>threshold</th>\n",
       "<th>value</th>\n",
       "<th>idx</th></tr></thead>\n",
       "    <tbody><tr><td>max f1</td>\n",
       "<td>0.4976278</td>\n",
       "<td>0.8604225</td>\n",
       "<td>121.0</td></tr>\n",
       "<tr><td>max f2</td>\n",
       "<td>0.4155728</td>\n",
       "<td>0.8578152</td>\n",
       "<td>138.0</td></tr>\n",
       "<tr><td>max f0point5</td>\n",
       "<td>0.7959362</td>\n",
       "<td>0.9005678</td>\n",
       "<td>64.0</td></tr>\n",
       "<tr><td>max accuracy</td>\n",
       "<td>0.4976278</td>\n",
       "<td>0.98857</td>\n",
       "<td>121.0</td></tr>\n",
       "<tr><td>max precision</td>\n",
       "<td>0.9527430</td>\n",
       "<td>0.9988031</td>\n",
       "<td>16.0</td></tr>\n",
       "<tr><td>max recall</td>\n",
       "<td>0.0263027</td>\n",
       "<td>1.0</td>\n",
       "<td>351.0</td></tr>\n",
       "<tr><td>max specificity</td>\n",
       "<td>0.9986957</td>\n",
       "<td>0.9999896</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max absolute_mcc</td>\n",
       "<td>0.4976278</td>\n",
       "<td>0.8547882</td>\n",
       "<td>121.0</td></tr>\n",
       "<tr><td>max min_per_class_accuracy</td>\n",
       "<td>0.1059807</td>\n",
       "<td>0.9516691</td>\n",
       "<td>245.0</td></tr>\n",
       "<tr><td>max mean_per_class_accuracy</td>\n",
       "<td>0.1037161</td>\n",
       "<td>0.9520071</td>\n",
       "<td>247.0</td></tr>\n",
       "<tr><td>max tns</td>\n",
       "<td>0.9986957</td>\n",
       "<td>95797.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fns</td>\n",
       "<td>0.9986957</td>\n",
       "<td>4202.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fps</td>\n",
       "<td>0.0019873</td>\n",
       "<td>95798.0</td>\n",
       "<td>399.0</td></tr>\n",
       "<tr><td>max tps</td>\n",
       "<td>0.0263027</td>\n",
       "<td>4202.0</td>\n",
       "<td>351.0</td></tr>\n",
       "<tr><td>max tnr</td>\n",
       "<td>0.9986957</td>\n",
       "<td>0.9999896</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fnr</td>\n",
       "<td>0.9986957</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fpr</td>\n",
       "<td>0.0019873</td>\n",
       "<td>1.0</td>\n",
       "<td>399.0</td></tr>\n",
       "<tr><td>max tpr</td>\n",
       "<td>0.0263027</td>\n",
       "<td>1.0</td>\n",
       "<td>351.0</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-92.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-92 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-92 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-92 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-92 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-92 .h2o-table th,\n",
       "#h2o-table-92 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-92 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-92\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Gains/Lift Table: Avg response rate:  4.20 %, avg score:  7.43 %</caption>\n",
       "    <thead><tr><th>group</th>\n",
       "<th>cumulative_data_fraction</th>\n",
       "<th>lower_threshold</th>\n",
       "<th>lift</th>\n",
       "<th>cumulative_lift</th>\n",
       "<th>response_rate</th>\n",
       "<th>score</th>\n",
       "<th>cumulative_response_rate</th>\n",
       "<th>cumulative_score</th>\n",
       "<th>capture_rate</th>\n",
       "<th>cumulative_capture_rate</th>\n",
       "<th>gain</th>\n",
       "<th>cumulative_gain</th>\n",
       "<th>kolmogorov_smirnov</th></tr></thead>\n",
       "    <tbody><tr><td>1</td>\n",
       "<td>0.01</td>\n",
       "<td>0.9626961</td>\n",
       "<td>23.7505950</td>\n",
       "<td>23.7505950</td>\n",
       "<td>0.998</td>\n",
       "<td>0.9691264</td>\n",
       "<td>0.998</td>\n",
       "<td>0.9691264</td>\n",
       "<td>0.2375059</td>\n",
       "<td>0.2375059</td>\n",
       "<td>2275.0594955</td>\n",
       "<td>2275.0594955</td>\n",
       "<td>0.2374851</td></tr>\n",
       "<tr><td>2</td>\n",
       "<td>0.02</td>\n",
       "<td>0.9433391</td>\n",
       "<td>23.7743931</td>\n",
       "<td>23.7624941</td>\n",
       "<td>0.999</td>\n",
       "<td>0.9546235</td>\n",
       "<td>0.9985</td>\n",
       "<td>0.9618749</td>\n",
       "<td>0.2377439</td>\n",
       "<td>0.4752499</td>\n",
       "<td>2277.4393146</td>\n",
       "<td>2276.2494050</td>\n",
       "<td>0.4752186</td></tr>\n",
       "<tr><td>3</td>\n",
       "<td>0.03</td>\n",
       "<td>0.8127500</td>\n",
       "<td>21.5373632</td>\n",
       "<td>23.0207838</td>\n",
       "<td>0.905</td>\n",
       "<td>0.8906454</td>\n",
       "<td>0.9673333</td>\n",
       "<td>0.9381317</td>\n",
       "<td>0.2153736</td>\n",
       "<td>0.6906235</td>\n",
       "<td>2053.7363160</td>\n",
       "<td>2202.0783754</td>\n",
       "<td>0.6896005</td></tr>\n",
       "<tr><td>4</td>\n",
       "<td>0.04</td>\n",
       "<td>0.4925265</td>\n",
       "<td>14.8976678</td>\n",
       "<td>20.9900048</td>\n",
       "<td>0.626</td>\n",
       "<td>0.6527370</td>\n",
       "<td>0.882</td>\n",
       "<td>0.8667830</td>\n",
       "<td>0.1489767</td>\n",
       "<td>0.8396002</td>\n",
       "<td>1389.7667777</td>\n",
       "<td>1999.0004760</td>\n",
       "<td>0.8346732</td></tr>\n",
       "<tr><td>5</td>\n",
       "<td>0.05</td>\n",
       "<td>0.2820315</td>\n",
       "<td>4.5216564</td>\n",
       "<td>17.6963351</td>\n",
       "<td>0.19</td>\n",
       "<td>0.3649629</td>\n",
       "<td>0.7436</td>\n",
       "<td>0.7664190</td>\n",
       "<td>0.0452166</td>\n",
       "<td>0.8848168</td>\n",
       "<td>352.1656354</td>\n",
       "<td>1669.6335079</td>\n",
       "<td>0.8714344</td></tr>\n",
       "<tr><td>6</td>\n",
       "<td>0.1</td>\n",
       "<td>0.0913986</td>\n",
       "<td>1.5564017</td>\n",
       "<td>9.6263684</td>\n",
       "<td>0.0654</td>\n",
       "<td>0.1466950</td>\n",
       "<td>0.4045</td>\n",
       "<td>0.4565570</td>\n",
       "<td>0.0778201</td>\n",
       "<td>0.9626368</td>\n",
       "<td>55.6401713</td>\n",
       "<td>862.6368396</td>\n",
       "<td>0.9004748</td></tr>\n",
       "<tr><td>7</td>\n",
       "<td>0.15</td>\n",
       "<td>0.0653466</td>\n",
       "<td>0.4093289</td>\n",
       "<td>6.5540219</td>\n",
       "<td>0.0172</td>\n",
       "<td>0.0763117</td>\n",
       "<td>0.2754</td>\n",
       "<td>0.3298086</td>\n",
       "<td>0.0204664</td>\n",
       "<td>0.9831033</td>\n",
       "<td>-59.0671109</td>\n",
       "<td>555.4021894</td>\n",
       "<td>0.8696458</td></tr>\n",
       "<tr><td>8</td>\n",
       "<td>0.2</td>\n",
       "<td>0.0535977</td>\n",
       "<td>0.1285102</td>\n",
       "<td>4.9476440</td>\n",
       "<td>0.0054</td>\n",
       "<td>0.0589899</td>\n",
       "<td>0.2079</td>\n",
       "<td>0.2621039</td>\n",
       "<td>0.0064255</td>\n",
       "<td>0.9895288</td>\n",
       "<td>-87.1489767</td>\n",
       "<td>394.7643979</td>\n",
       "<td>0.8241600</td></tr>\n",
       "<tr><td>9</td>\n",
       "<td>0.3</td>\n",
       "<td>0.0414853</td>\n",
       "<td>0.0285578</td>\n",
       "<td>3.3079486</td>\n",
       "<td>0.0012</td>\n",
       "<td>0.0467581</td>\n",
       "<td>0.139</td>\n",
       "<td>0.1903220</td>\n",
       "<td>0.0028558</td>\n",
       "<td>0.9923846</td>\n",
       "<td>-97.1442170</td>\n",
       "<td>230.7948596</td>\n",
       "<td>0.7227547</td></tr>\n",
       "<tr><td>10</td>\n",
       "<td>0.4</td>\n",
       "<td>0.0350769</td>\n",
       "<td>0.0166587</td>\n",
       "<td>2.4851261</td>\n",
       "<td>0.0007</td>\n",
       "<td>0.0379959</td>\n",
       "<td>0.104425</td>\n",
       "<td>0.1522405</td>\n",
       "<td>0.0016659</td>\n",
       "<td>0.9940505</td>\n",
       "<td>-98.3341266</td>\n",
       "<td>148.5126130</td>\n",
       "<td>0.6201074</td></tr>\n",
       "<tr><td>11</td>\n",
       "<td>0.5</td>\n",
       "<td>0.0304279</td>\n",
       "<td>0.0118991</td>\n",
       "<td>1.9904807</td>\n",
       "<td>0.0005</td>\n",
       "<td>0.0326663</td>\n",
       "<td>0.08364</td>\n",
       "<td>0.1283256</td>\n",
       "<td>0.0011899</td>\n",
       "<td>0.9952404</td>\n",
       "<td>-98.8100904</td>\n",
       "<td>99.0480723</td>\n",
       "<td>0.5169632</td></tr>\n",
       "<tr><td>12</td>\n",
       "<td>0.6</td>\n",
       "<td>0.0263867</td>\n",
       "<td>0.0428367</td>\n",
       "<td>1.6658734</td>\n",
       "<td>0.0018</td>\n",
       "<td>0.0283567</td>\n",
       "<td>0.07</td>\n",
       "<td>0.1116641</td>\n",
       "<td>0.0042837</td>\n",
       "<td>0.9995240</td>\n",
       "<td>-95.7163256</td>\n",
       "<td>66.5873394</td>\n",
       "<td>0.4170484</td></tr>\n",
       "<tr><td>13</td>\n",
       "<td>0.7</td>\n",
       "<td>0.0227004</td>\n",
       "<td>0.0047596</td>\n",
       "<td>1.4285714</td>\n",
       "<td>0.0002</td>\n",
       "<td>0.0245369</td>\n",
       "<td>0.0600286</td>\n",
       "<td>0.0992174</td>\n",
       "<td>0.0004760</td>\n",
       "<td>1.0</td>\n",
       "<td>-99.5240362</td>\n",
       "<td>42.8571429</td>\n",
       "<td>0.3131589</td></tr>\n",
       "<tr><td>14</td>\n",
       "<td>0.8</td>\n",
       "<td>0.0188387</td>\n",
       "<td>0.0</td>\n",
       "<td>1.25</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0208410</td>\n",
       "<td>0.052525</td>\n",
       "<td>0.0894203</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>25.0</td>\n",
       "<td>0.2087726</td></tr>\n",
       "<tr><td>15</td>\n",
       "<td>0.9</td>\n",
       "<td>0.0144461</td>\n",
       "<td>0.0</td>\n",
       "<td>1.1111111</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0168100</td>\n",
       "<td>0.0466889</td>\n",
       "<td>0.0813525</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>11.1111111</td>\n",
       "<td>0.1043863</td></tr>\n",
       "<tr><td>16</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0011704</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0104776</td>\n",
       "<td>0.04202</td>\n",
       "<td>0.0742650</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>"
      ],
      "text/plain": [
       "ModelMetricsBinomial: drf\n",
       "** Reported on test data. **\n",
       "\n",
       "MSE: 0.010869869336460194\n",
       "RMSE: 0.10425866552215309\n",
       "LogLoss: 0.06270051806376054\n",
       "Mean Per-Class Error: 0.08321662204917754\n",
       "AUC: 0.9899913933708621\n",
       "AUCPR: 0.9168687082576839\n",
       "Gini: 0.9799827867417241\n",
       "\n",
       "Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.49762779177502836\n",
       "       0      1     Error    Rate\n",
       "-----  -----  ----  -------  -----------------\n",
       "0      95334  464   0.0048   (464.0/95798.0)\n",
       "1      679    3523  0.1616   (679.0/4202.0)\n",
       "Total  96013  3987  0.0114   (1143.0/100000.0)\n",
       "\n",
       "Maximum Metrics: Maximum metrics at their respective thresholds\n",
       "metric                       threshold    value     idx\n",
       "---------------------------  -----------  --------  -----\n",
       "max f1                       0.497628     0.860423  121\n",
       "max f2                       0.415573     0.857815  138\n",
       "max f0point5                 0.795936     0.900568  64\n",
       "max accuracy                 0.497628     0.98857   121\n",
       "max precision                0.952743     0.998803  16\n",
       "max recall                   0.0263027    1         351\n",
       "max specificity              0.998696     0.99999   0\n",
       "max absolute_mcc             0.497628     0.854788  121\n",
       "max min_per_class_accuracy   0.105981     0.951669  245\n",
       "max mean_per_class_accuracy  0.103716     0.952007  247\n",
       "max tns                      0.998696     95797     0\n",
       "max fns                      0.998696     4202      0\n",
       "max fps                      0.00198726   95798     399\n",
       "max tps                      0.0263027    4202      351\n",
       "max tnr                      0.998696     0.99999   0\n",
       "max fnr                      0.998696     1         0\n",
       "max fpr                      0.00198726   1         399\n",
       "max tpr                      0.0263027    1         351\n",
       "\n",
       "Gains/Lift Table: Avg response rate:  4.20 %, avg score:  7.43 %\n",
       "group    cumulative_data_fraction    lower_threshold    lift        cumulative_lift    response_rate    score      cumulative_response_rate    cumulative_score    capture_rate    cumulative_capture_rate    gain      cumulative_gain    kolmogorov_smirnov\n",
       "-------  --------------------------  -----------------  ----------  -----------------  ---------------  ---------  --------------------------  ------------------  --------------  -------------------------  --------  -----------------  --------------------\n",
       "1        0.01                        0.962696           23.7506     23.7506            0.998            0.969126   0.998                       0.969126            0.237506        0.237506                   2275.06   2275.06            0.237485\n",
       "2        0.02                        0.943339           23.7744     23.7625            0.999            0.954623   0.9985                      0.961875            0.237744        0.47525                    2277.44   2276.25            0.475219\n",
       "3        0.03                        0.81275            21.5374     23.0208            0.905            0.890645   0.967333                    0.938132            0.215374        0.690624                   2053.74   2202.08            0.689601\n",
       "4        0.04                        0.492526           14.8977     20.99              0.626            0.652737   0.882                       0.866783            0.148977        0.8396                     1389.77   1999               0.834673\n",
       "5        0.05                        0.282032           4.52166     17.6963            0.19             0.364963   0.7436                      0.766419            0.0452166       0.884817                   352.166   1669.63            0.871434\n",
       "6        0.1                         0.0913986          1.5564      9.62637            0.0654           0.146695   0.4045                      0.456557            0.0778201       0.962637                   55.6402   862.637            0.900475\n",
       "7        0.15                        0.0653466          0.409329    6.55402            0.0172           0.0763117  0.2754                      0.329809            0.0204664       0.983103                   -59.0671  555.402            0.869646\n",
       "8        0.2                         0.0535977          0.12851     4.94764            0.0054           0.0589899  0.2079                      0.262104            0.00642551      0.989529                   -87.149   394.764            0.82416\n",
       "9        0.3                         0.0414853          0.0285578   3.30795            0.0012           0.0467581  0.139                       0.190322            0.00285578      0.992385                   -97.1442  230.795            0.722755\n",
       "10       0.4                         0.0350769          0.0166587   2.48513            0.0007           0.0379959  0.104425                    0.15224             0.00166587      0.99405                    -98.3341  148.513            0.620107\n",
       "11       0.5                         0.0304279          0.0118991   1.99048            0.0005           0.0326663  0.08364                     0.128326            0.00118991      0.99524                    -98.8101  99.0481            0.516963\n",
       "12       0.6                         0.0263867          0.0428367   1.66587            0.0018           0.0283567  0.07                        0.111664            0.00428367      0.999524                   -95.7163  66.5873            0.417048\n",
       "13       0.7                         0.0227004          0.00475964  1.42857            0.0002           0.0245369  0.0600286                   0.0992174           0.000475964     1                          -99.524   42.8571            0.313159\n",
       "14       0.8                         0.0188387          0           1.25               0                0.020841   0.052525                    0.0894203           0               1                          -100      25                 0.208773\n",
       "15       0.9                         0.0144461          0           1.11111            0                0.01681    0.0466889                   0.0813525           0               1                          -100      11.1111            0.104386\n",
       "16       1                           0.00117042         0           1                  0                0.0104776  0.04202                     0.074265            0               1                          -100      0                  0"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf_3model.model_performance(validation_h2o)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "7076c3ab",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<h2o.plot._plot_result._MObject at 0x27fc8009fd0>"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1600x900 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "rf_3model.learning_curve_plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "83090338",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "drf Model Build progress: |██████████████████████████████████████████████████████| (done) 100%\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<pre style='margin: 1em 0 1em 0;'>Model Details\n",
       "=============\n",
       "H2ORandomForestEstimator : Distributed Random Forest\n",
       "Model Key: DRF_model_python_1680256840616_491\n",
       "</pre>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-93.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-93 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-93 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-93 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-93 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-93 .h2o-table th,\n",
       "#h2o-table-93 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-93 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-93\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Model Summary: </caption>\n",
       "    <thead><tr><th></th>\n",
       "<th>number_of_trees</th>\n",
       "<th>number_of_internal_trees</th>\n",
       "<th>model_size_in_bytes</th>\n",
       "<th>min_depth</th>\n",
       "<th>max_depth</th>\n",
       "<th>mean_depth</th>\n",
       "<th>min_leaves</th>\n",
       "<th>max_leaves</th>\n",
       "<th>mean_leaves</th></tr></thead>\n",
       "    <tbody><tr><td></td>\n",
       "<td>50.0</td>\n",
       "<td>50.0</td>\n",
       "<td>472123.0</td>\n",
       "<td>15.0</td>\n",
       "<td>15.0</td>\n",
       "<td>15.0</td>\n",
       "<td>254.0</td>\n",
       "<td>1385.0</td>\n",
       "<td>745.46</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>\n",
       "<div style='margin: 1em 0 1em 0;'><pre style='margin: 1em 0 1em 0;'>ModelMetricsBinomial: drf\n",
       "** Reported on train data. **\n",
       "\n",
       "MSE: 0.017505605640559002\n",
       "RMSE: 0.1323087511866052\n",
       "LogLoss: 0.08028558473349331\n",
       "Mean Per-Class Error: 0.0509734941958894\n",
       "AUC: 0.9811273599286645\n",
       "AUCPR: 0.9580454776260502\n",
       "Gini: 0.962254719857329</pre>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-94.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-94 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-94 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-94 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-94 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-94 .h2o-table th,\n",
       "#h2o-table-94 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-94 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-94\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.19283058598703115</caption>\n",
       "    <thead><tr><th></th>\n",
       "<th>0</th>\n",
       "<th>1</th>\n",
       "<th>Error</th>\n",
       "<th>Rate</th></tr></thead>\n",
       "    <tbody><tr><td>0</td>\n",
       "<td>194055.0</td>\n",
       "<td>584.0</td>\n",
       "<td>0.003</td>\n",
       "<td> (584.0/194639.0)</td></tr>\n",
       "<tr><td>1</td>\n",
       "<td>2583.0</td>\n",
       "<td>23522.0</td>\n",
       "<td>0.0989</td>\n",
       "<td> (2583.0/26105.0)</td></tr>\n",
       "<tr><td>Total</td>\n",
       "<td>196638.0</td>\n",
       "<td>24106.0</td>\n",
       "<td>0.0143</td>\n",
       "<td> (3167.0/220744.0)</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-95.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-95 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-95 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-95 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-95 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-95 .h2o-table th,\n",
       "#h2o-table-95 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-95 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-95\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Maximum Metrics: Maximum metrics at their respective thresholds</caption>\n",
       "    <thead><tr><th>metric</th>\n",
       "<th>threshold</th>\n",
       "<th>value</th>\n",
       "<th>idx</th></tr></thead>\n",
       "    <tbody><tr><td>max f1</td>\n",
       "<td>0.1928306</td>\n",
       "<td>0.9369262</td>\n",
       "<td>228.0</td></tr>\n",
       "<tr><td>max f2</td>\n",
       "<td>0.1242883</td>\n",
       "<td>0.9219676</td>\n",
       "<td>251.0</td></tr>\n",
       "<tr><td>max f0point5</td>\n",
       "<td>0.3390765</td>\n",
       "<td>0.9647272</td>\n",
       "<td>195.0</td></tr>\n",
       "<tr><td>max accuracy</td>\n",
       "<td>0.2006577</td>\n",
       "<td>0.9856757</td>\n",
       "<td>226.0</td></tr>\n",
       "<tr><td>max precision</td>\n",
       "<td>0.9999894</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max recall</td>\n",
       "<td>0.0118797</td>\n",
       "<td>1.0</td>\n",
       "<td>381.0</td></tr>\n",
       "<tr><td>max specificity</td>\n",
       "<td>0.9999894</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max absolute_mcc</td>\n",
       "<td>0.2006577</td>\n",
       "<td>0.9298737</td>\n",
       "<td>226.0</td></tr>\n",
       "<tr><td>max min_per_class_accuracy</td>\n",
       "<td>0.0646510</td>\n",
       "<td>0.9436506</td>\n",
       "<td>291.0</td></tr>\n",
       "<tr><td>max mean_per_class_accuracy</td>\n",
       "<td>0.1110112</td>\n",
       "<td>0.9554698</td>\n",
       "<td>257.0</td></tr>\n",
       "<tr><td>max tns</td>\n",
       "<td>0.9999894</td>\n",
       "<td>194639.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fns</td>\n",
       "<td>0.9999894</td>\n",
       "<td>19609.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fps</td>\n",
       "<td>0.0004478</td>\n",
       "<td>194639.0</td>\n",
       "<td>399.0</td></tr>\n",
       "<tr><td>max tps</td>\n",
       "<td>0.0118797</td>\n",
       "<td>26105.0</td>\n",
       "<td>381.0</td></tr>\n",
       "<tr><td>max tnr</td>\n",
       "<td>0.9999894</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fnr</td>\n",
       "<td>0.9999894</td>\n",
       "<td>0.7511588</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fpr</td>\n",
       "<td>0.0004478</td>\n",
       "<td>1.0</td>\n",
       "<td>399.0</td></tr>\n",
       "<tr><td>max tpr</td>\n",
       "<td>0.0118797</td>\n",
       "<td>1.0</td>\n",
       "<td>381.0</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-96.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-96 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-96 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-96 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-96 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-96 .h2o-table th,\n",
       "#h2o-table-96 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-96 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-96\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Gains/Lift Table: Avg response rate: 11.83 %, avg score: 11.79 %</caption>\n",
       "    <thead><tr><th>group</th>\n",
       "<th>cumulative_data_fraction</th>\n",
       "<th>lower_threshold</th>\n",
       "<th>lift</th>\n",
       "<th>cumulative_lift</th>\n",
       "<th>response_rate</th>\n",
       "<th>score</th>\n",
       "<th>cumulative_response_rate</th>\n",
       "<th>cumulative_score</th>\n",
       "<th>capture_rate</th>\n",
       "<th>cumulative_capture_rate</th>\n",
       "<th>gain</th>\n",
       "<th>cumulative_gain</th>\n",
       "<th>kolmogorov_smirnov</th></tr></thead>\n",
       "    <tbody><tr><td>1</td>\n",
       "<td>0.0274843</td>\n",
       "<td>1.0</td>\n",
       "<td>8.4560046</td>\n",
       "<td>8.4560046</td>\n",
       "<td>1.0</td>\n",
       "<td>1.0</td>\n",
       "<td>1.0</td>\n",
       "<td>1.0</td>\n",
       "<td>0.2324076</td>\n",
       "<td>0.2324076</td>\n",
       "<td>745.6004597</td>\n",
       "<td>745.6004597</td>\n",
       "<td>0.2324076</td></tr>\n",
       "<tr><td>2</td>\n",
       "<td>0.0301027</td>\n",
       "<td>0.9995836</td>\n",
       "<td>8.4560046</td>\n",
       "<td>8.4560046</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9998052</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9999831</td>\n",
       "<td>0.0221414</td>\n",
       "<td>0.2545489</td>\n",
       "<td>745.6004597</td>\n",
       "<td>745.6004597</td>\n",
       "<td>0.2545489</td></tr>\n",
       "<tr><td>3</td>\n",
       "<td>0.0400011</td>\n",
       "<td>0.9788196</td>\n",
       "<td>8.4560046</td>\n",
       "<td>8.4560046</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9917001</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9979334</td>\n",
       "<td>0.0837004</td>\n",
       "<td>0.3382494</td>\n",
       "<td>745.6004597</td>\n",
       "<td>745.6004597</td>\n",
       "<td>0.3382494</td></tr>\n",
       "<tr><td>4</td>\n",
       "<td>0.0500036</td>\n",
       "<td>0.9496454</td>\n",
       "<td>8.4560046</td>\n",
       "<td>8.4560046</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9619785</td>\n",
       "<td>1.0</td>\n",
       "<td>0.9907411</td>\n",
       "<td>0.0845815</td>\n",
       "<td>0.4228309</td>\n",
       "<td>745.6004597</td>\n",
       "<td>745.6004597</td>\n",
       "<td>0.4228309</td></tr>\n",
       "<tr><td>5</td>\n",
       "<td>0.1000027</td>\n",
       "<td>0.4568346</td>\n",
       "<td>8.3931802</td>\n",
       "<td>8.4245938</td>\n",
       "<td>0.9925704</td>\n",
       "<td>0.8024556</td>\n",
       "<td>0.9962854</td>\n",
       "<td>0.8966026</td>\n",
       "<td>0.4196514</td>\n",
       "<td>0.8424823</td>\n",
       "<td>739.3180244</td>\n",
       "<td>742.4593844</td>\n",
       "<td>0.8420610</td></tr>\n",
       "<tr><td>6</td>\n",
       "<td>0.1500018</td>\n",
       "<td>0.0704989</td>\n",
       "<td>1.9621118</td>\n",
       "<td>6.2704982</td>\n",
       "<td>0.2320377</td>\n",
       "<td>0.1392311</td>\n",
       "<td>0.7415439</td>\n",
       "<td>0.6441531</td>\n",
       "<td>0.0981038</td>\n",
       "<td>0.9405861</td>\n",
       "<td>96.2111785</td>\n",
       "<td>527.0498214</td>\n",
       "<td>0.8966175</td></tr>\n",
       "<tr><td>7</td>\n",
       "<td>0.2000009</td>\n",
       "<td>0.0500207</td>\n",
       "<td>0.2421036</td>\n",
       "<td>4.7634337</td>\n",
       "<td>0.0286310</td>\n",
       "<td>0.0583603</td>\n",
       "<td>0.5633197</td>\n",
       "<td>0.4977082</td>\n",
       "<td>0.0121050</td>\n",
       "<td>0.9526911</td>\n",
       "<td>-75.7896398</td>\n",
       "<td>376.3433698</td>\n",
       "<td>0.8536410</td></tr>\n",
       "<tr><td>8</td>\n",
       "<td>0.2999991</td>\n",
       "<td>0.0357801</td>\n",
       "<td>0.1379071</td>\n",
       "<td>3.2216148</td>\n",
       "<td>0.0163088</td>\n",
       "<td>0.0414786</td>\n",
       "<td>0.3809855</td>\n",
       "<td>0.3456340</td>\n",
       "<td>0.0137905</td>\n",
       "<td>0.9664815</td>\n",
       "<td>-86.2092885</td>\n",
       "<td>222.1614786</td>\n",
       "<td>0.7558711</td></tr>\n",
       "<tr><td>9</td>\n",
       "<td>0.4000018</td>\n",
       "<td>0.0296955</td>\n",
       "<td>0.0869542</td>\n",
       "<td>2.4379319</td>\n",
       "<td>0.0102831</td>\n",
       "<td>0.0325084</td>\n",
       "<td>0.2883078</td>\n",
       "<td>0.2673508</td>\n",
       "<td>0.0086957</td>\n",
       "<td>0.9751772</td>\n",
       "<td>-91.3045842</td>\n",
       "<td>143.7931879</td>\n",
       "<td>0.6523179</td></tr>\n",
       "<tr><td>10</td>\n",
       "<td>0.5</td>\n",
       "<td>0.0254608</td>\n",
       "<td>0.0861919</td>\n",
       "<td>1.9675924</td>\n",
       "<td>0.0101930</td>\n",
       "<td>0.0274769</td>\n",
       "<td>0.2326858</td>\n",
       "<td>0.2193769</td>\n",
       "<td>0.0086190</td>\n",
       "<td>0.9837962</td>\n",
       "<td>-91.3808053</td>\n",
       "<td>96.7592415</td>\n",
       "<td>0.5486830</td></tr>\n",
       "<tr><td>11</td>\n",
       "<td>0.5999982</td>\n",
       "<td>0.0221245</td>\n",
       "<td>0.0697197</td>\n",
       "<td>1.6512851</td>\n",
       "<td>0.0082450</td>\n",
       "<td>0.0237433</td>\n",
       "<td>0.1952796</td>\n",
       "<td>0.1867718</td>\n",
       "<td>0.0069718</td>\n",
       "<td>0.9907681</td>\n",
       "<td>-93.0280292</td>\n",
       "<td>65.1285074</td>\n",
       "<td>0.4431800</td></tr>\n",
       "<tr><td>12</td>\n",
       "<td>0.7000009</td>\n",
       "<td>0.0190307</td>\n",
       "<td>0.0467331</td>\n",
       "<td>1.4220574</td>\n",
       "<td>0.0055266</td>\n",
       "<td>0.0205692</td>\n",
       "<td>0.1681713</td>\n",
       "<td>0.1630279</td>\n",
       "<td>0.0046734</td>\n",
       "<td>0.9954415</td>\n",
       "<td>-95.3266928</td>\n",
       "<td>42.2057426</td>\n",
       "<td>0.3350651</td></tr>\n",
       "<tr><td>13</td>\n",
       "<td>0.7999991</td>\n",
       "<td>0.0158058</td>\n",
       "<td>0.0317953</td>\n",
       "<td>1.2482776</td>\n",
       "<td>0.0037601</td>\n",
       "<td>0.0174439</td>\n",
       "<td>0.1476203</td>\n",
       "<td>0.1448302</td>\n",
       "<td>0.0031795</td>\n",
       "<td>0.9986210</td>\n",
       "<td>-96.8204749</td>\n",
       "<td>24.8277606</td>\n",
       "<td>0.2252610</td></tr>\n",
       "<tr><td>14</td>\n",
       "<td>0.8999973</td>\n",
       "<td>0.0112004</td>\n",
       "<td>0.0137907</td>\n",
       "<td>1.1111145</td>\n",
       "<td>0.0016309</td>\n",
       "<td>0.0137879</td>\n",
       "<td>0.1313995</td>\n",
       "<td>0.1302702</td>\n",
       "<td>0.0013790</td>\n",
       "<td>1.0</td>\n",
       "<td>-98.6209289</td>\n",
       "<td>11.1114467</td>\n",
       "<td>0.1134151</td></tr>\n",
       "<tr><td>15</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0068893</td>\n",
       "<td>0.1182592</td>\n",
       "<td>0.1179318</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div></div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-97.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-97 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-97 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-97 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-97 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-97 .h2o-table th,\n",
       "#h2o-table-97 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-97 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-97\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Scoring History: </caption>\n",
       "    <thead><tr><th></th>\n",
       "<th>timestamp</th>\n",
       "<th>duration</th>\n",
       "<th>number_of_trees</th>\n",
       "<th>training_rmse</th>\n",
       "<th>training_logloss</th>\n",
       "<th>training_auc</th>\n",
       "<th>training_pr_auc</th>\n",
       "<th>training_lift</th>\n",
       "<th>training_classification_error</th></tr></thead>\n",
       "    <tbody><tr><td></td>\n",
       "<td>2023-03-31 12:25:51</td>\n",
       "<td> 0.003 sec</td>\n",
       "<td>0.0</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td>\n",
       "<td>nan</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:25:51</td>\n",
       "<td> 0.070 sec</td>\n",
       "<td>1.0</td>\n",
       "<td>0.1599810</td>\n",
       "<td>0.1772992</td>\n",
       "<td>0.9464337</td>\n",
       "<td>0.8844741</td>\n",
       "<td>8.3397215</td>\n",
       "<td>0.0286392</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:25:51</td>\n",
       "<td> 0.139 sec</td>\n",
       "<td>2.0</td>\n",
       "<td>0.1567302</td>\n",
       "<td>0.1763927</td>\n",
       "<td>0.9456982</td>\n",
       "<td>0.8872538</td>\n",
       "<td>8.3449362</td>\n",
       "<td>0.0265728</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:25:51</td>\n",
       "<td> 0.214 sec</td>\n",
       "<td>3.0</td>\n",
       "<td>0.1508431</td>\n",
       "<td>0.1528991</td>\n",
       "<td>0.9512210</td>\n",
       "<td>0.9002682</td>\n",
       "<td>8.3690674</td>\n",
       "<td>0.0247243</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:25:51</td>\n",
       "<td> 0.299 sec</td>\n",
       "<td>4.0</td>\n",
       "<td>0.1514469</td>\n",
       "<td>0.1394974</td>\n",
       "<td>0.9515033</td>\n",
       "<td>0.9029504</td>\n",
       "<td>8.4022575</td>\n",
       "<td>0.0244240</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:25:51</td>\n",
       "<td> 0.381 sec</td>\n",
       "<td>5.0</td>\n",
       "<td>0.1498596</td>\n",
       "<td>0.1234611</td>\n",
       "<td>0.9557570</td>\n",
       "<td>0.9108908</td>\n",
       "<td>8.4231284</td>\n",
       "<td>0.0237186</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:25:51</td>\n",
       "<td> 0.464 sec</td>\n",
       "<td>6.0</td>\n",
       "<td>0.1471990</td>\n",
       "<td>0.1134622</td>\n",
       "<td>0.9591720</td>\n",
       "<td>0.9176134</td>\n",
       "<td>8.4341199</td>\n",
       "<td>0.0224584</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:25:51</td>\n",
       "<td> 0.554 sec</td>\n",
       "<td>7.0</td>\n",
       "<td>0.1440669</td>\n",
       "<td>0.1053019</td>\n",
       "<td>0.9625341</td>\n",
       "<td>0.9244037</td>\n",
       "<td>8.4390171</td>\n",
       "<td>0.0211666</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:25:52</td>\n",
       "<td> 0.639 sec</td>\n",
       "<td>8.0</td>\n",
       "<td>0.1432628</td>\n",
       "<td>0.1005826</td>\n",
       "<td>0.9641284</td>\n",
       "<td>0.9269401</td>\n",
       "<td>8.4423091</td>\n",
       "<td>0.0206438</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:25:52</td>\n",
       "<td> 0.720 sec</td>\n",
       "<td>9.0</td>\n",
       "<td>0.1440298</td>\n",
       "<td>0.0978742</td>\n",
       "<td>0.9655278</td>\n",
       "<td>0.9290023</td>\n",
       "<td>8.4455401</td>\n",
       "<td>0.0206707</td></tr>\n",
       "<tr><td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:25:54</td>\n",
       "<td> 3.229 sec</td>\n",
       "<td>37.0</td>\n",
       "<td>0.1329746</td>\n",
       "<td>0.0807346</td>\n",
       "<td>0.9797901</td>\n",
       "<td>0.9556463</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0148045</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:25:54</td>\n",
       "<td> 3.324 sec</td>\n",
       "<td>38.0</td>\n",
       "<td>0.1324668</td>\n",
       "<td>0.0802251</td>\n",
       "<td>0.9801670</td>\n",
       "<td>0.9562776</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0146958</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:25:54</td>\n",
       "<td> 3.412 sec</td>\n",
       "<td>39.0</td>\n",
       "<td>0.1319750</td>\n",
       "<td>0.0797206</td>\n",
       "<td>0.9805302</td>\n",
       "<td>0.9568684</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0145825</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:25:54</td>\n",
       "<td> 3.507 sec</td>\n",
       "<td>40.0</td>\n",
       "<td>0.1323018</td>\n",
       "<td>0.0803051</td>\n",
       "<td>0.9802591</td>\n",
       "<td>0.9564083</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0146776</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:25:54</td>\n",
       "<td> 3.593 sec</td>\n",
       "<td>41.0</td>\n",
       "<td>0.1317557</td>\n",
       "<td>0.0797153</td>\n",
       "<td>0.9807060</td>\n",
       "<td>0.9572073</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0145055</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:25:55</td>\n",
       "<td> 3.677 sec</td>\n",
       "<td>42.0</td>\n",
       "<td>0.1320714</td>\n",
       "<td>0.0800269</td>\n",
       "<td>0.9806405</td>\n",
       "<td>0.9571984</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0145146</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:25:55</td>\n",
       "<td> 3.770 sec</td>\n",
       "<td>43.0</td>\n",
       "<td>0.1318760</td>\n",
       "<td>0.0798526</td>\n",
       "<td>0.9809900</td>\n",
       "<td>0.9576144</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0143741</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:25:55</td>\n",
       "<td> 3.861 sec</td>\n",
       "<td>44.0</td>\n",
       "<td>0.1321847</td>\n",
       "<td>0.0800925</td>\n",
       "<td>0.9810728</td>\n",
       "<td>0.9578108</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0143198</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:25:55</td>\n",
       "<td> 3.952 sec</td>\n",
       "<td>45.0</td>\n",
       "<td>0.1322692</td>\n",
       "<td>0.0801752</td>\n",
       "<td>0.9810347</td>\n",
       "<td>0.9578679</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0142473</td></tr>\n",
       "<tr><td></td>\n",
       "<td>2023-03-31 12:25:55</td>\n",
       "<td> 4.278 sec</td>\n",
       "<td>50.0</td>\n",
       "<td>0.1323088</td>\n",
       "<td>0.0802856</td>\n",
       "<td>0.9811274</td>\n",
       "<td>0.9580455</td>\n",
       "<td>8.4560046</td>\n",
       "<td>0.0143469</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "<pre style='font-size: smaller; margin-bottom: 1em;'>[47 rows x 10 columns]</pre></div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-98.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-98 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-98 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-98 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-98 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-98 .h2o-table th,\n",
       "#h2o-table-98 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-98 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-98\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Variable Importances: </caption>\n",
       "    <thead><tr><th>variable</th>\n",
       "<th>relative_importance</th>\n",
       "<th>scaled_importance</th>\n",
       "<th>percentage</th></tr></thead>\n",
       "    <tbody><tr><td>RollingMean_400</td>\n",
       "<td>85204.4375000</td>\n",
       "<td>1.0</td>\n",
       "<td>0.1431248</td></tr>\n",
       "<tr><td>RollingMean_800</td>\n",
       "<td>84269.3125000</td>\n",
       "<td>0.9890249</td>\n",
       "<td>0.1415540</td></tr>\n",
       "<tr><td>RollingSD_800</td>\n",
       "<td>77863.6796875</td>\n",
       "<td>0.9138454</td>\n",
       "<td>0.1307939</td></tr>\n",
       "<tr><td>RollingMean_200</td>\n",
       "<td>59378.4687500</td>\n",
       "<td>0.6968941</td>\n",
       "<td>0.0997428</td></tr>\n",
       "<tr><td>RollingSD_400</td>\n",
       "<td>50173.2656250</td>\n",
       "<td>0.5888574</td>\n",
       "<td>0.0842801</td></tr>\n",
       "<tr><td>RollingSD_200</td>\n",
       "<td>31352.2578125</td>\n",
       "<td>0.3679651</td>\n",
       "<td>0.0526649</td></tr>\n",
       "<tr><td>RollingMean_100</td>\n",
       "<td>17826.7675781</td>\n",
       "<td>0.2092235</td>\n",
       "<td>0.0299451</td></tr>\n",
       "<tr><td>RollingSD_100</td>\n",
       "<td>16165.5419922</td>\n",
       "<td>0.1897265</td>\n",
       "<td>0.0271546</td></tr>\n",
       "<tr><td>meanaddsd</td>\n",
       "<td>14138.1113281</td>\n",
       "<td>0.1659316</td>\n",
       "<td>0.0237489</td></tr>\n",
       "<tr><td>Lag50</td>\n",
       "<td>10032.2060547</td>\n",
       "<td>0.1177428</td>\n",
       "<td>0.0168519</td></tr>\n",
       "<tr><td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td>\n",
       "<td>---</td></tr>\n",
       "<tr><td>RollingMean_2</td>\n",
       "<td>5518.7656250</td>\n",
       "<td>0.0647709</td>\n",
       "<td>0.0092703</td></tr>\n",
       "<tr><td>RollingSD_5</td>\n",
       "<td>3948.0102539</td>\n",
       "<td>0.0463357</td>\n",
       "<td>0.0066318</td></tr>\n",
       "<tr><td>RollingSD_4</td>\n",
       "<td>3489.4306641</td>\n",
       "<td>0.0409536</td>\n",
       "<td>0.0058615</td></tr>\n",
       "<tr><td>RollingSD_3</td>\n",
       "<td>3127.0729980</td>\n",
       "<td>0.0367008</td>\n",
       "<td>0.0052528</td></tr>\n",
       "<tr><td>diff30</td>\n",
       "<td>3079.6169434</td>\n",
       "<td>0.0361439</td>\n",
       "<td>0.0051731</td></tr>\n",
       "<tr><td>diff50</td>\n",
       "<td>2938.3535156</td>\n",
       "<td>0.0344859</td>\n",
       "<td>0.0049358</td></tr>\n",
       "<tr><td>diff5</td>\n",
       "<td>1759.5457764</td>\n",
       "<td>0.0206509</td>\n",
       "<td>0.0029557</td></tr>\n",
       "<tr><td>RollingSD_2</td>\n",
       "<td>1551.4398193</td>\n",
       "<td>0.0182084</td>\n",
       "<td>0.0026061</td></tr>\n",
       "<tr><td>diff2</td>\n",
       "<td>1293.2437744</td>\n",
       "<td>0.0151781</td>\n",
       "<td>0.0021724</td></tr>\n",
       "<tr><td>diff1</td>\n",
       "<td>837.2649536</td>\n",
       "<td>0.0098265</td>\n",
       "<td>0.0014064</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "<pre style='font-size: smaller; margin-bottom: 1em;'>[36 rows x 4 columns]</pre></div><pre style=\"font-size: smaller; margin: 1em 0 0 0;\">\n",
       "\n",
       "[tips]\n",
       "Use `model.explain()` to inspect the model.\n",
       "--\n",
       "Use `h2o.display.toggle_user_tips()` to switch on/off this section.</pre>"
      ],
      "text/plain": [
       "Model Details\n",
       "=============\n",
       "H2ORandomForestEstimator : Distributed Random Forest\n",
       "Model Key: DRF_model_python_1680256840616_491\n",
       "\n",
       "\n",
       "Model Summary: \n",
       "    number_of_trees    number_of_internal_trees    model_size_in_bytes    min_depth    max_depth    mean_depth    min_leaves    max_leaves    mean_leaves\n",
       "--  -----------------  --------------------------  ---------------------  -----------  -----------  ------------  ------------  ------------  -------------\n",
       "    50                 50                          472123                 15           15           15            254           1385          745.46\n",
       "\n",
       "ModelMetricsBinomial: drf\n",
       "** Reported on train data. **\n",
       "\n",
       "MSE: 0.017505605640559002\n",
       "RMSE: 0.1323087511866052\n",
       "LogLoss: 0.08028558473349331\n",
       "Mean Per-Class Error: 0.0509734941958894\n",
       "AUC: 0.9811273599286645\n",
       "AUCPR: 0.9580454776260502\n",
       "Gini: 0.962254719857329\n",
       "\n",
       "Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.19283058598703115\n",
       "       0       1      Error    Rate\n",
       "-----  ------  -----  -------  -----------------\n",
       "0      194055  584    0.003    (584.0/194639.0)\n",
       "1      2583    23522  0.0989   (2583.0/26105.0)\n",
       "Total  196638  24106  0.0143   (3167.0/220744.0)\n",
       "\n",
       "Maximum Metrics: Maximum metrics at their respective thresholds\n",
       "metric                       threshold    value     idx\n",
       "---------------------------  -----------  --------  -----\n",
       "max f1                       0.192831     0.936926  228\n",
       "max f2                       0.124288     0.921968  251\n",
       "max f0point5                 0.339077     0.964727  195\n",
       "max accuracy                 0.200658     0.985676  226\n",
       "max precision                0.999989     1         0\n",
       "max recall                   0.0118797    1         381\n",
       "max specificity              0.999989     1         0\n",
       "max absolute_mcc             0.200658     0.929874  226\n",
       "max min_per_class_accuracy   0.064651     0.943651  291\n",
       "max mean_per_class_accuracy  0.111011     0.95547   257\n",
       "max tns                      0.999989     194639    0\n",
       "max fns                      0.999989     19609     0\n",
       "max fps                      0.000447754  194639    399\n",
       "max tps                      0.0118797    26105     381\n",
       "max tnr                      0.999989     1         0\n",
       "max fnr                      0.999989     0.751159  0\n",
       "max fpr                      0.000447754  1         399\n",
       "max tpr                      0.0118797    1         381\n",
       "\n",
       "Gains/Lift Table: Avg response rate: 11.83 %, avg score: 11.79 %\n",
       "group    cumulative_data_fraction    lower_threshold    lift       cumulative_lift    response_rate    score       cumulative_response_rate    cumulative_score    capture_rate    cumulative_capture_rate    gain      cumulative_gain    kolmogorov_smirnov\n",
       "-------  --------------------------  -----------------  ---------  -----------------  ---------------  ----------  --------------------------  ------------------  --------------  -------------------------  --------  -----------------  --------------------\n",
       "1        0.0274843                   1                  8.456      8.456              1                1           1                           1                   0.232408        0.232408                   745.6     745.6              0.232408\n",
       "2        0.0301027                   0.999584           8.456      8.456              1                0.999805    1                           0.999983            0.0221414       0.254549                   745.6     745.6              0.254549\n",
       "3        0.0400011                   0.97882            8.456      8.456              1                0.9917      1                           0.997933            0.0837004       0.338249                   745.6     745.6              0.338249\n",
       "4        0.0500036                   0.949645           8.456      8.456              1                0.961978    1                           0.990741            0.0845815       0.422831                   745.6     745.6              0.422831\n",
       "5        0.100003                    0.456835           8.39318    8.42459            0.99257          0.802456    0.996285                    0.896603            0.419651        0.842482                   739.318   742.459            0.842061\n",
       "6        0.150002                    0.0704989          1.96211    6.2705             0.232038         0.139231    0.741544                    0.644153            0.0981038       0.940586                   96.2112   527.05             0.896618\n",
       "7        0.200001                    0.0500207          0.242104   4.76343            0.028631         0.0583603   0.56332                     0.497708            0.012105        0.952691                   -75.7896  376.343            0.853641\n",
       "8        0.299999                    0.0357801          0.137907   3.22161            0.0163088        0.0414786   0.380985                    0.345634            0.0137905       0.966482                   -86.2093  222.161            0.755871\n",
       "9        0.400002                    0.0296955          0.0869542  2.43793            0.0102831        0.0325084   0.288308                    0.267351            0.00869565      0.975177                   -91.3046  143.793            0.652318\n",
       "10       0.5                         0.0254608          0.0861919  1.96759            0.010193         0.0274769   0.232686                    0.219377            0.00861904      0.983796                   -91.3808  96.7592            0.548683\n",
       "11       0.599998                    0.0221245          0.0697197  1.65129            0.00824499       0.0237433   0.19528                     0.186772            0.00697184      0.990768                   -93.028   65.1285            0.44318\n",
       "12       0.700001                    0.0190307          0.0467331  1.42206            0.00552661       0.0205692   0.168171                    0.163028            0.00467343      0.995441                   -95.3267  42.2057            0.335065\n",
       "13       0.799999                    0.0158058          0.0317953  1.24828            0.00376008       0.0174439   0.14762                     0.14483             0.00317947      0.998621                   -96.8205  24.8278            0.225261\n",
       "14       0.899997                    0.0112004          0.0137907  1.11111            0.00163088       0.0137879   0.131399                    0.13027             0.00137905      1                          -98.6209  11.1114            0.113415\n",
       "15       1                           0                  0          1                  0                0.00688929  0.118259                    0.117932            0               1                          -100      0                  0\n",
       "\n",
       "Scoring History: \n",
       "     timestamp            duration    number_of_trees    training_rmse        training_logloss     training_auc        training_pr_auc     training_lift      training_classification_error\n",
       "---  -------------------  ----------  -----------------  -------------------  -------------------  ------------------  ------------------  -----------------  -------------------------------\n",
       "     2023-03-31 12:25:51  0.003 sec   0.0                nan                  nan                  nan                 nan                 nan                nan\n",
       "     2023-03-31 12:25:51  0.070 sec   1.0                0.15998099369869656  0.17729917426458558  0.946433731318102   0.8844740801809722  8.339721468094169  0.028639207443735635\n",
       "     2023-03-31 12:25:51  0.139 sec   2.0                0.15673015072212412  0.17639269395068707  0.9456981570086617  0.8872537830989309  8.344936235216059  0.02657280663531321\n",
       "     2023-03-31 12:25:51  0.214 sec   3.0                0.15084308151021666  0.15289913354714935  0.9512210157477604  0.9002682131655597  8.369067360805028  0.02472434471418899\n",
       "     2023-03-31 12:25:51  0.299 sec   4.0                0.15144689508462972  0.13949743996462943  0.9515033183433383  0.9029503720049593  8.402257465232276  0.024423987941429803\n",
       "     2023-03-31 12:25:51  0.381 sec   5.0                0.14985959111332384  0.12346107400989226  0.9557569538545668  0.9108908150194558  8.423128405239584  0.023718577791198008\n",
       "     2023-03-31 12:25:51  0.464 sec   6.0                0.14719896569463312  0.11346222216265693  0.9591720319106649  0.9176134281731149  8.434119897648316  0.022458377269781074\n",
       "     2023-03-31 12:25:51  0.554 sec   7.0                0.14406692606957028  0.10530186130436651  0.9625340683575153  0.9244037105260434  8.439017087585848  0.021166618675150937\n",
       "     2023-03-31 12:25:52  0.639 sec   8.0                0.14326282207025307  0.10058257683257397  0.9641284495950238  0.9269401032694083  8.442309078038084  0.020643758074860345\n",
       "     2023-03-31 12:25:52  0.720 sec   9.0                0.1440297552516849   0.09787418679233587  0.9655277888558934  0.929002265145232   8.445540108003414  0.020670728057209427\n",
       "---  ---                  ---         ---                ---                  ---                  ---                 ---                 ---                ---\n",
       "     2023-03-31 12:25:54  3.229 sec   37.0               0.13297455174233225  0.08073461499898979  0.9797900759941088  0.9556463062478541  8.456004596820533  0.0148044793969485\n",
       "     2023-03-31 12:25:54  3.324 sec   38.0               0.1324667932788369   0.08022514364902018  0.9801669946599897  0.956277580296344   8.456004596820533  0.014695756170043126\n",
       "     2023-03-31 12:25:54  3.412 sec   39.0               0.13197501510052345  0.07972057349484743  0.980530221080172   0.9568683746113846  8.456004596820533  0.014582502808683362\n",
       "     2023-03-31 12:25:54  3.507 sec   40.0               0.13230180276931994  0.08030512035169787  0.9802591329776816  0.9564083148211702  8.456004596820533  0.014677635632225565\n",
       "     2023-03-31 12:25:54  3.593 sec   41.0               0.13175567796885448  0.07971525955378637  0.9807059964233641  0.9572072564965031  8.456004596820533  0.01450549052295872\n",
       "     2023-03-31 12:25:55  3.677 sec   42.0               0.13207135537366962  0.08002691962604186  0.9806404708079401  0.9571983689653335  8.456004596820533  0.014514550791867503\n",
       "     2023-03-31 12:25:55  3.770 sec   43.0               0.13187601799953436  0.0798526446190322   0.9809900171846235  0.957614436041913   8.456004596820533  0.014374116623781394\n",
       "     2023-03-31 12:25:55  3.861 sec   44.0               0.1321846649248098   0.08009246199211788  0.9810728138328236  0.9578107629789451  8.456004596820533  0.014319755010328707\n",
       "     2023-03-31 12:25:55  3.952 sec   45.0               0.13226916841608144  0.08017516568838128  0.9810346813684226  0.9578678915869199  8.456004596820533  0.014247272859058456\n",
       "     2023-03-31 12:25:55  4.278 sec   50.0               0.1323087511866052   0.08028558473349331  0.9811273599286645  0.9580454776260502  8.456004596820533  0.01434693581705505\n",
       "[47 rows x 10 columns]\n",
       "\n",
       "\n",
       "Variable Importances: \n",
       "variable         relative_importance    scaled_importance     percentage\n",
       "---------------  ---------------------  --------------------  ---------------------\n",
       "RollingMean_400  85204.4375             1.0                   0.14312479017508348\n",
       "RollingMean_800  84269.3125             0.9890249260785273    0.1415539850229167\n",
       "RollingSD_800    77863.6796875          0.9138453579662444    0.1307939251113928\n",
       "RollingMean_200  59378.46875            0.6968940878225973    0.09974282009386544\n",
       "RollingSD_400    50173.265625           0.5888574245326131    0.0842800953292703\n",
       "RollingSD_200    31352.2578125          0.3679650817775776    0.05266492512117323\n",
       "RollingMean_100  17826.767578125        0.20922346419017201   0.029945064411922465\n",
       "RollingSD_100    16165.5419921875       0.18972652676907234   0.027154569334470836\n",
       "meanaddsd        14138.111328125        0.16593163153180843   0.023748929946399346\n",
       "Lag50            10032.2060546875       0.11774276491981418   0.016851908523782586\n",
       "---              ---                    ---                   ---\n",
       "RollingMean_2    5518.765625            0.06477087094202107   0.009270317313034177\n",
       "RollingSD_5      3948.01025390625       0.046335735200484715  0.006631792378177605\n",
       "RollingSD_4      3489.4306640625        0.0409536259665173    0.005861479123366638\n",
       "RollingSD_3      3127.072998046875      0.03670082321765078   0.005252797622279101\n",
       "diff30           3079.616943359375      0.036143856279309106  0.005173081846094489\n",
       "diff50           2938.353515625         0.034485921177814245  0.004935790232569131\n",
       "diff5            1759.5457763671875     0.020650870165854773  0.0029556514594208556\n",
       "RollingSD_2      1551.4398193359375     0.018208439194683228  0.0026060790391548034\n",
       "diff2            1293.2437744140625     0.015178127012622581  0.0021723662439323737\n",
       "diff1            837.2649536132812      0.009826541647121151  0.0014064217113909342\n",
       "[36 rows x 4 columns]\n",
       "\n",
       "\n",
       "[tips]\n",
       "Use `model.explain()` to inspect the model.\n",
       "--\n",
       "Use `h2o.display.toggle_user_tips()` to switch on/off this section."
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf_4model = H2ORandomForestEstimator(seed=1,mtries=3,max_depth=15,ntrees=50,histogram_type='Random')\n",
    "rf_4model.train(x=predictors, y=response, training_frame=train_h2o)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "5aab2a2d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<pre style='margin: 1em 0 1em 0;'>ModelMetricsBinomial: drf\n",
       "** Reported on test data. **\n",
       "\n",
       "MSE: 0.010749700438463487\n",
       "RMSE: 0.10368076214256668\n",
       "LogLoss: 0.06250668514959247\n",
       "Mean Per-Class Error: 0.07635168177081796\n",
       "AUC: 0.9890371839746609\n",
       "AUCPR: 0.9135610684081078\n",
       "Gini: 0.9780743679493218</pre>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-105.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-105 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-105 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-105 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-105 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-105 .h2o-table th,\n",
       "#h2o-table-105 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-105 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-105\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.4670059429746205</caption>\n",
       "    <thead><tr><th></th>\n",
       "<th>0</th>\n",
       "<th>1</th>\n",
       "<th>Error</th>\n",
       "<th>Rate</th></tr></thead>\n",
       "    <tbody><tr><td>0</td>\n",
       "<td>95327.0</td>\n",
       "<td>471.0</td>\n",
       "<td>0.0049</td>\n",
       "<td> (471.0/95798.0)</td></tr>\n",
       "<tr><td>1</td>\n",
       "<td>621.0</td>\n",
       "<td>3581.0</td>\n",
       "<td>0.1478</td>\n",
       "<td> (621.0/4202.0)</td></tr>\n",
       "<tr><td>Total</td>\n",
       "<td>95948.0</td>\n",
       "<td>4052.0</td>\n",
       "<td>0.0109</td>\n",
       "<td> (1092.0/100000.0)</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-106.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-106 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-106 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-106 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-106 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-106 .h2o-table th,\n",
       "#h2o-table-106 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-106 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-106\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Maximum Metrics: Maximum metrics at their respective thresholds</caption>\n",
       "    <thead><tr><th>metric</th>\n",
       "<th>threshold</th>\n",
       "<th>value</th>\n",
       "<th>idx</th></tr></thead>\n",
       "    <tbody><tr><td>max f1</td>\n",
       "<td>0.4670059</td>\n",
       "<td>0.8677005</td>\n",
       "<td>126.0</td></tr>\n",
       "<tr><td>max f2</td>\n",
       "<td>0.3769764</td>\n",
       "<td>0.8630420</td>\n",
       "<td>147.0</td></tr>\n",
       "<tr><td>max f0point5</td>\n",
       "<td>0.8159793</td>\n",
       "<td>0.9134674</td>\n",
       "<td>59.0</td></tr>\n",
       "<tr><td>max accuracy</td>\n",
       "<td>0.4670059</td>\n",
       "<td>0.98908</td>\n",
       "<td>126.0</td></tr>\n",
       "<tr><td>max precision</td>\n",
       "<td>0.9336894</td>\n",
       "<td>0.9940185</td>\n",
       "<td>24.0</td></tr>\n",
       "<tr><td>max recall</td>\n",
       "<td>0.0213527</td>\n",
       "<td>1.0</td>\n",
       "<td>361.0</td></tr>\n",
       "<tr><td>max specificity</td>\n",
       "<td>0.9994118</td>\n",
       "<td>0.9999791</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max absolute_mcc</td>\n",
       "<td>0.4670059</td>\n",
       "<td>0.8621623</td>\n",
       "<td>126.0</td></tr>\n",
       "<tr><td>max min_per_class_accuracy</td>\n",
       "<td>0.1110710</td>\n",
       "<td>0.9521389</td>\n",
       "<td>242.0</td></tr>\n",
       "<tr><td>max mean_per_class_accuracy</td>\n",
       "<td>0.1110710</td>\n",
       "<td>0.9532232</td>\n",
       "<td>242.0</td></tr>\n",
       "<tr><td>max tns</td>\n",
       "<td>0.9994118</td>\n",
       "<td>95796.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fns</td>\n",
       "<td>0.9994118</td>\n",
       "<td>4202.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fps</td>\n",
       "<td>0.0015519</td>\n",
       "<td>95798.0</td>\n",
       "<td>399.0</td></tr>\n",
       "<tr><td>max tps</td>\n",
       "<td>0.0213527</td>\n",
       "<td>4202.0</td>\n",
       "<td>361.0</td></tr>\n",
       "<tr><td>max tnr</td>\n",
       "<td>0.9994118</td>\n",
       "<td>0.9999791</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fnr</td>\n",
       "<td>0.9994118</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fpr</td>\n",
       "<td>0.0015519</td>\n",
       "<td>1.0</td>\n",
       "<td>399.0</td></tr>\n",
       "<tr><td>max tpr</td>\n",
       "<td>0.0213527</td>\n",
       "<td>1.0</td>\n",
       "<td>361.0</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-107.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-107 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-107 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-107 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-107 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-107 .h2o-table th,\n",
       "#h2o-table-107 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-107 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-107\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Gains/Lift Table: Avg response rate:  4.20 %, avg score:  7.36 %</caption>\n",
       "    <thead><tr><th>group</th>\n",
       "<th>cumulative_data_fraction</th>\n",
       "<th>lower_threshold</th>\n",
       "<th>lift</th>\n",
       "<th>cumulative_lift</th>\n",
       "<th>response_rate</th>\n",
       "<th>score</th>\n",
       "<th>cumulative_response_rate</th>\n",
       "<th>cumulative_score</th>\n",
       "<th>capture_rate</th>\n",
       "<th>cumulative_capture_rate</th>\n",
       "<th>gain</th>\n",
       "<th>cumulative_gain</th>\n",
       "<th>kolmogorov_smirnov</th></tr></thead>\n",
       "    <tbody><tr><td>1</td>\n",
       "<td>0.01</td>\n",
       "<td>0.9520255</td>\n",
       "<td>23.5602094</td>\n",
       "<td>23.5602094</td>\n",
       "<td>0.99</td>\n",
       "<td>0.9602755</td>\n",
       "<td>0.99</td>\n",
       "<td>0.9602755</td>\n",
       "<td>0.2356021</td>\n",
       "<td>0.2356021</td>\n",
       "<td>2256.0209424</td>\n",
       "<td>2256.0209424</td>\n",
       "<td>0.2354977</td></tr>\n",
       "<tr><td>2</td>\n",
       "<td>0.02</td>\n",
       "<td>0.9253598</td>\n",
       "<td>23.7267968</td>\n",
       "<td>23.6435031</td>\n",
       "<td>0.997</td>\n",
       "<td>0.9391723</td>\n",
       "<td>0.9935</td>\n",
       "<td>0.9497239</td>\n",
       "<td>0.2372680</td>\n",
       "<td>0.4728701</td>\n",
       "<td>2272.6796763</td>\n",
       "<td>2264.3503094</td>\n",
       "<td>0.4727344</td></tr>\n",
       "<tr><td>3</td>\n",
       "<td>0.03002</td>\n",
       "<td>0.8263441</td>\n",
       "<td>23.0144186</td>\n",
       "<td>23.4335288</td>\n",
       "<td>0.9670659</td>\n",
       "<td>0.8853695</td>\n",
       "<td>0.9846769</td>\n",
       "<td>0.9282439</td>\n",
       "<td>0.2306045</td>\n",
       "<td>0.7034745</td>\n",
       "<td>2201.4418569</td>\n",
       "<td>2243.3528845</td>\n",
       "<td>0.7029944</td></tr>\n",
       "<tr><td>4</td>\n",
       "<td>0.04</td>\n",
       "<td>0.4791968</td>\n",
       "<td>14.2598381</td>\n",
       "<td>21.1446930</td>\n",
       "<td>0.5991984</td>\n",
       "<td>0.6511404</td>\n",
       "<td>0.8885</td>\n",
       "<td>0.8591065</td>\n",
       "<td>0.1423132</td>\n",
       "<td>0.8457877</td>\n",
       "<td>1325.9838096</td>\n",
       "<td>2014.4693003</td>\n",
       "<td>0.8411321</td></tr>\n",
       "<tr><td>5</td>\n",
       "<td>0.05</td>\n",
       "<td>0.2949319</td>\n",
       "<td>3.6887197</td>\n",
       "<td>17.6534983</td>\n",
       "<td>0.155</td>\n",
       "<td>0.3737082</td>\n",
       "<td>0.7418</td>\n",
       "<td>0.7620269</td>\n",
       "<td>0.0368872</td>\n",
       "<td>0.8826749</td>\n",
       "<td>268.8719657</td>\n",
       "<td>1665.3498334</td>\n",
       "<td>0.8691986</td></tr>\n",
       "<tr><td>6</td>\n",
       "<td>0.1</td>\n",
       "<td>0.0928536</td>\n",
       "<td>1.6182770</td>\n",
       "<td>9.6358877</td>\n",
       "<td>0.068</td>\n",
       "<td>0.1522323</td>\n",
       "<td>0.4049</td>\n",
       "<td>0.4571296</td>\n",
       "<td>0.0809139</td>\n",
       "<td>0.9635888</td>\n",
       "<td>61.8277011</td>\n",
       "<td>863.5887673</td>\n",
       "<td>0.9014685</td></tr>\n",
       "<tr><td>7</td>\n",
       "<td>0.15</td>\n",
       "<td>0.0651489</td>\n",
       "<td>0.3855307</td>\n",
       "<td>6.5524353</td>\n",
       "<td>0.0162</td>\n",
       "<td>0.0767009</td>\n",
       "<td>0.2753333</td>\n",
       "<td>0.3303200</td>\n",
       "<td>0.0192765</td>\n",
       "<td>0.9828653</td>\n",
       "<td>-61.4469300</td>\n",
       "<td>555.2435348</td>\n",
       "<td>0.8693974</td></tr>\n",
       "<tr><td>8</td>\n",
       "<td>0.20001</td>\n",
       "<td>0.0529299</td>\n",
       "<td>0.1189672</td>\n",
       "<td>4.9438271</td>\n",
       "<td>0.0049990</td>\n",
       "<td>0.0582480</td>\n",
       "<td>0.2077396</td>\n",
       "<td>0.2622918</td>\n",
       "<td>0.0059495</td>\n",
       "<td>0.9888149</td>\n",
       "<td>-88.1032837</td>\n",
       "<td>394.3827059</td>\n",
       "<td>0.8234043</td></tr>\n",
       "<tr><td>9</td>\n",
       "<td>0.3</td>\n",
       "<td>0.0412860</td>\n",
       "<td>0.0214205</td>\n",
       "<td>3.3031890</td>\n",
       "<td>0.0009001</td>\n",
       "<td>0.0463643</td>\n",
       "<td>0.1388</td>\n",
       "<td>0.1903232</td>\n",
       "<td>0.0021418</td>\n",
       "<td>0.9909567</td>\n",
       "<td>-97.8579486</td>\n",
       "<td>230.3188958</td>\n",
       "<td>0.7212642</td></tr>\n",
       "<tr><td>10</td>\n",
       "<td>0.4</td>\n",
       "<td>0.0341007</td>\n",
       "<td>0.0047596</td>\n",
       "<td>2.4785816</td>\n",
       "<td>0.0002</td>\n",
       "<td>0.0373828</td>\n",
       "<td>0.10415</td>\n",
       "<td>0.1520881</td>\n",
       "<td>0.0004760</td>\n",
       "<td>0.9914327</td>\n",
       "<td>-99.5240362</td>\n",
       "<td>147.8581628</td>\n",
       "<td>0.6173747</td></tr>\n",
       "<tr><td>11</td>\n",
       "<td>0.5</td>\n",
       "<td>0.0292284</td>\n",
       "<td>0.0261780</td>\n",
       "<td>1.9881009</td>\n",
       "<td>0.0011</td>\n",
       "<td>0.0315618</td>\n",
       "<td>0.08354</td>\n",
       "<td>0.1279828</td>\n",
       "<td>0.0026178</td>\n",
       "<td>0.9940505</td>\n",
       "<td>-97.3821990</td>\n",
       "<td>98.8100904</td>\n",
       "<td>0.5157211</td></tr>\n",
       "<tr><td>12</td>\n",
       "<td>0.6</td>\n",
       "<td>0.0253918</td>\n",
       "<td>0.0118991</td>\n",
       "<td>1.6587339</td>\n",
       "<td>0.0005</td>\n",
       "<td>0.0272588</td>\n",
       "<td>0.0697</td>\n",
       "<td>0.1111955</td>\n",
       "<td>0.0011899</td>\n",
       "<td>0.9952404</td>\n",
       "<td>-98.8100904</td>\n",
       "<td>65.8733936</td>\n",
       "<td>0.4125768</td></tr>\n",
       "<tr><td>13</td>\n",
       "<td>0.7</td>\n",
       "<td>0.0217886</td>\n",
       "<td>0.0452166</td>\n",
       "<td>1.4282315</td>\n",
       "<td>0.0019</td>\n",
       "<td>0.0235679</td>\n",
       "<td>0.0600143</td>\n",
       "<td>0.0986773</td>\n",
       "<td>0.0045217</td>\n",
       "<td>0.9997620</td>\n",
       "<td>-95.4783436</td>\n",
       "<td>42.8231454</td>\n",
       "<td>0.3129105</td></tr>\n",
       "<tr><td>14</td>\n",
       "<td>0.8</td>\n",
       "<td>0.0178631</td>\n",
       "<td>0.0023798</td>\n",
       "<td>1.25</td>\n",
       "<td>0.0001</td>\n",
       "<td>0.0197718</td>\n",
       "<td>0.052525</td>\n",
       "<td>0.0888141</td>\n",
       "<td>0.0002380</td>\n",
       "<td>1.0</td>\n",
       "<td>-99.7620181</td>\n",
       "<td>25.0</td>\n",
       "<td>0.2087726</td></tr>\n",
       "<tr><td>15</td>\n",
       "<td>0.9</td>\n",
       "<td>0.0136696</td>\n",
       "<td>0.0</td>\n",
       "<td>1.1111111</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0158244</td>\n",
       "<td>0.0466889</td>\n",
       "<td>0.0807041</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>11.1111111</td>\n",
       "<td>0.1043863</td></tr>\n",
       "<tr><td>16</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0004474</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0095315</td>\n",
       "<td>0.04202</td>\n",
       "<td>0.0735868</td>\n",
       "<td>0.0</td>\n",
       "<td>1.0</td>\n",
       "<td>-100.0</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>"
      ],
      "text/plain": [
       "ModelMetricsBinomial: drf\n",
       "** Reported on test data. **\n",
       "\n",
       "MSE: 0.010749700438463487\n",
       "RMSE: 0.10368076214256668\n",
       "LogLoss: 0.06250668514959247\n",
       "Mean Per-Class Error: 0.07635168177081796\n",
       "AUC: 0.9890371839746609\n",
       "AUCPR: 0.9135610684081078\n",
       "Gini: 0.9780743679493218\n",
       "\n",
       "Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.4670059429746205\n",
       "       0      1     Error    Rate\n",
       "-----  -----  ----  -------  -----------------\n",
       "0      95327  471   0.0049   (471.0/95798.0)\n",
       "1      621    3581  0.1478   (621.0/4202.0)\n",
       "Total  95948  4052  0.0109   (1092.0/100000.0)\n",
       "\n",
       "Maximum Metrics: Maximum metrics at their respective thresholds\n",
       "metric                       threshold    value     idx\n",
       "---------------------------  -----------  --------  -----\n",
       "max f1                       0.467006     0.867701  126\n",
       "max f2                       0.376976     0.863042  147\n",
       "max f0point5                 0.815979     0.913467  59\n",
       "max accuracy                 0.467006     0.98908   126\n",
       "max precision                0.933689     0.994018  24\n",
       "max recall                   0.0213527    1         361\n",
       "max specificity              0.999412     0.999979  0\n",
       "max absolute_mcc             0.467006     0.862162  126\n",
       "max min_per_class_accuracy   0.111071     0.952139  242\n",
       "max mean_per_class_accuracy  0.111071     0.953223  242\n",
       "max tns                      0.999412     95796     0\n",
       "max fns                      0.999412     4202      0\n",
       "max fps                      0.00155195   95798     399\n",
       "max tps                      0.0213527    4202      361\n",
       "max tnr                      0.999412     0.999979  0\n",
       "max fnr                      0.999412     1         0\n",
       "max fpr                      0.00155195   1         399\n",
       "max tpr                      0.0213527    1         361\n",
       "\n",
       "Gains/Lift Table: Avg response rate:  4.20 %, avg score:  7.36 %\n",
       "group    cumulative_data_fraction    lower_threshold    lift        cumulative_lift    response_rate    score       cumulative_response_rate    cumulative_score    capture_rate    cumulative_capture_rate    gain      cumulative_gain    kolmogorov_smirnov\n",
       "-------  --------------------------  -----------------  ----------  -----------------  ---------------  ----------  --------------------------  ------------------  --------------  -------------------------  --------  -----------------  --------------------\n",
       "1        0.01                        0.952025           23.5602     23.5602            0.99             0.960275    0.99                        0.960275            0.235602        0.235602                   2256.02   2256.02            0.235498\n",
       "2        0.02                        0.92536            23.7268     23.6435            0.997            0.939172    0.9935                      0.949724            0.237268        0.47287                    2272.68   2264.35            0.472734\n",
       "3        0.03002                     0.826344           23.0144     23.4335            0.967066         0.885369    0.984677                    0.928244            0.230604        0.703475                   2201.44   2243.35            0.702994\n",
       "4        0.04                        0.479197           14.2598     21.1447            0.599198         0.65114     0.8885                      0.859107            0.142313        0.845788                   1325.98   2014.47            0.841132\n",
       "5        0.05                        0.294932           3.68872     17.6535            0.155            0.373708    0.7418                      0.762027            0.0368872       0.882675                   268.872   1665.35            0.869199\n",
       "6        0.1                         0.0928536          1.61828     9.63589            0.068            0.152232    0.4049                      0.45713             0.0809139       0.963589                   61.8277   863.589            0.901468\n",
       "7        0.15                        0.0651489          0.385531    6.55244            0.0162           0.0767009   0.275333                    0.33032             0.0192765       0.982865                   -61.4469  555.244            0.869397\n",
       "8        0.20001                     0.0529299          0.118967    4.94383            0.004999         0.058248    0.20774                     0.262292            0.00594955      0.988815                   -88.1033  394.383            0.823404\n",
       "9        0.3                         0.041286           0.0214205   3.30319            0.00090009       0.0463643   0.1388                      0.190323            0.00214184      0.990957                   -97.8579  230.319            0.721264\n",
       "10       0.4                         0.0341007          0.00475964  2.47858            0.0002           0.0373828   0.10415                     0.152088            0.000475964     0.991433                   -99.524   147.858            0.617375\n",
       "11       0.5                         0.0292284          0.026178    1.9881             0.0011           0.0315618   0.08354                     0.127983            0.0026178       0.99405                    -97.3822  98.8101            0.515721\n",
       "12       0.6                         0.0253918          0.0118991   1.65873            0.0005           0.0272588   0.0697                      0.111195            0.00118991      0.99524                    -98.8101  65.8734            0.412577\n",
       "13       0.7                         0.0217886          0.0452166   1.42823            0.0019           0.0235679   0.0600143                   0.0986773           0.00452166      0.999762                   -95.4783  42.8231            0.312911\n",
       "14       0.8                         0.0178631          0.00237982  1.25               0.0001           0.0197718   0.052525                    0.0888141           0.000237982     1                          -99.762   25                 0.208773\n",
       "15       0.9                         0.0136696          0           1.11111            0                0.0158244   0.0466889                   0.0807041           0               1                          -100      11.1111            0.104386\n",
       "16       1                           0.000447427        0           1                  0                0.00953149  0.04202                     0.0735868           0               1                          -100      0                  0"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf_4model.model_performance(validation_h2o)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "68673f72",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<h2o.plot._plot_result._MObject at 0x27fc8226eb0>"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1600x900 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "rf_4model.learning_curve_plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "559e8d71",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<pre style='margin: 1em 0 1em 0;'>ModelMetricsBinomial: drf\n",
       "** Reported on test data. **\n",
       "\n",
       "MSE: 0.08607764751628483\n",
       "RMSE: 0.29338992401969916\n",
       "LogLoss: 0.2850543542126062\n",
       "Mean Per-Class Error: 0.13789716372509686\n",
       "AUC: 0.9149414130727681\n",
       "AUCPR: 0.7308461869304076\n",
       "Gini: 0.8298828261455362</pre>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-108.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-108 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-108 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-108 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-108 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-108 .h2o-table th,\n",
       "#h2o-table-108 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-108 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-108\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.18633682950984531</caption>\n",
       "    <thead><tr><th></th>\n",
       "<th>0</th>\n",
       "<th>1</th>\n",
       "<th>Error</th>\n",
       "<th>Rate</th></tr></thead>\n",
       "    <tbody><tr><td>0</td>\n",
       "<td>235455.0</td>\n",
       "<td>30202.0</td>\n",
       "<td>0.1137</td>\n",
       "<td> (30202.0/265657.0)</td></tr>\n",
       "<tr><td>1</td>\n",
       "<td>9503.0</td>\n",
       "<td>49119.0</td>\n",
       "<td>0.1621</td>\n",
       "<td> (9503.0/58622.0)</td></tr>\n",
       "<tr><td>Total</td>\n",
       "<td>244958.0</td>\n",
       "<td>79321.0</td>\n",
       "<td>0.1224</td>\n",
       "<td> (39705.0/324279.0)</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-109.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-109 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-109 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-109 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-109 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-109 .h2o-table th,\n",
       "#h2o-table-109 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-109 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-109\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Maximum Metrics: Maximum metrics at their respective thresholds</caption>\n",
       "    <thead><tr><th>metric</th>\n",
       "<th>threshold</th>\n",
       "<th>value</th>\n",
       "<th>idx</th></tr></thead>\n",
       "    <tbody><tr><td>max f1</td>\n",
       "<td>0.1863368</td>\n",
       "<td>0.7121637</td>\n",
       "<td>266.0</td></tr>\n",
       "<tr><td>max f2</td>\n",
       "<td>0.1228736</td>\n",
       "<td>0.7948205</td>\n",
       "<td>292.0</td></tr>\n",
       "<tr><td>max f0point5</td>\n",
       "<td>0.8066826</td>\n",
       "<td>0.6768289</td>\n",
       "<td>71.0</td></tr>\n",
       "<tr><td>max accuracy</td>\n",
       "<td>0.2506669</td>\n",
       "<td>0.8795019</td>\n",
       "<td>245.0</td></tr>\n",
       "<tr><td>max precision</td>\n",
       "<td>0.9817030</td>\n",
       "<td>0.9816671</td>\n",
       "<td>5.0</td></tr>\n",
       "<tr><td>max recall</td>\n",
       "<td>0.0069996</td>\n",
       "<td>1.0</td>\n",
       "<td>397.0</td></tr>\n",
       "<tr><td>max specificity</td>\n",
       "<td>0.9999388</td>\n",
       "<td>0.9997214</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max absolute_mcc</td>\n",
       "<td>0.1725827</td>\n",
       "<td>0.6488930</td>\n",
       "<td>270.0</td></tr>\n",
       "<tr><td>max min_per_class_accuracy</td>\n",
       "<td>0.1410400</td>\n",
       "<td>0.8674757</td>\n",
       "<td>283.0</td></tr>\n",
       "<tr><td>max mean_per_class_accuracy</td>\n",
       "<td>0.1436747</td>\n",
       "<td>0.8681719</td>\n",
       "<td>282.0</td></tr>\n",
       "<tr><td>max tns</td>\n",
       "<td>0.9999388</td>\n",
       "<td>265583.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fns</td>\n",
       "<td>0.9999388</td>\n",
       "<td>55934.0</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fps</td>\n",
       "<td>0.0024483</td>\n",
       "<td>265657.0</td>\n",
       "<td>399.0</td></tr>\n",
       "<tr><td>max tps</td>\n",
       "<td>0.0069996</td>\n",
       "<td>58622.0</td>\n",
       "<td>397.0</td></tr>\n",
       "<tr><td>max tnr</td>\n",
       "<td>0.9999388</td>\n",
       "<td>0.9997214</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fnr</td>\n",
       "<td>0.9999388</td>\n",
       "<td>0.9541469</td>\n",
       "<td>0.0</td></tr>\n",
       "<tr><td>max fpr</td>\n",
       "<td>0.0024483</td>\n",
       "<td>1.0</td>\n",
       "<td>399.0</td></tr>\n",
       "<tr><td>max tpr</td>\n",
       "<td>0.0069996</td>\n",
       "<td>1.0</td>\n",
       "<td>397.0</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>\n",
       "<div style='margin: 1em 0 1em 0;'>\n",
       "<style>\n",
       "\n",
       "#h2o-table-110.h2o-container {\n",
       "  overflow-x: auto;\n",
       "}\n",
       "#h2o-table-110 .h2o-table {\n",
       "  /* width: 100%; */\n",
       "  margin-top: 1em;\n",
       "  margin-bottom: 1em;\n",
       "}\n",
       "#h2o-table-110 .h2o-table caption {\n",
       "  white-space: nowrap;\n",
       "  caption-side: top;\n",
       "  text-align: left;\n",
       "  /* margin-left: 1em; */\n",
       "  margin: 0;\n",
       "  font-size: larger;\n",
       "}\n",
       "#h2o-table-110 .h2o-table thead {\n",
       "  white-space: nowrap; \n",
       "  position: sticky;\n",
       "  top: 0;\n",
       "  box-shadow: 0 -1px inset;\n",
       "}\n",
       "#h2o-table-110 .h2o-table tbody {\n",
       "  overflow: auto;\n",
       "}\n",
       "#h2o-table-110 .h2o-table th,\n",
       "#h2o-table-110 .h2o-table td {\n",
       "  text-align: right;\n",
       "  /* border: 1px solid; */\n",
       "}\n",
       "#h2o-table-110 .h2o-table tr:nth-child(even) {\n",
       "  /* background: #F5F5F5 */\n",
       "}\n",
       "\n",
       "</style>      \n",
       "<div id=\"h2o-table-110\" class=\"h2o-container\">\n",
       "  <table class=\"h2o-table\">\n",
       "    <caption>Gains/Lift Table: Avg response rate: 18.08 %, avg score: 19.12 %</caption>\n",
       "    <thead><tr><th>group</th>\n",
       "<th>cumulative_data_fraction</th>\n",
       "<th>lower_threshold</th>\n",
       "<th>lift</th>\n",
       "<th>cumulative_lift</th>\n",
       "<th>response_rate</th>\n",
       "<th>score</th>\n",
       "<th>cumulative_response_rate</th>\n",
       "<th>cumulative_score</th>\n",
       "<th>capture_rate</th>\n",
       "<th>cumulative_capture_rate</th>\n",
       "<th>gain</th>\n",
       "<th>cumulative_gain</th>\n",
       "<th>kolmogorov_smirnov</th></tr></thead>\n",
       "    <tbody><tr><td>1</td>\n",
       "<td>0.0100006</td>\n",
       "<td>0.9932337</td>\n",
       "<td>5.4037646</td>\n",
       "<td>5.4037646</td>\n",
       "<td>0.9768733</td>\n",
       "<td>0.9994378</td>\n",
       "<td>0.9768733</td>\n",
       "<td>0.9994378</td>\n",
       "<td>0.0540411</td>\n",
       "<td>0.0540411</td>\n",
       "<td>440.3764554</td>\n",
       "<td>440.3764554</td>\n",
       "<td>0.0537588</td></tr>\n",
       "<tr><td>2</td>\n",
       "<td>0.0200013</td>\n",
       "<td>0.9632378</td>\n",
       "<td>4.9108075</td>\n",
       "<td>5.1572860</td>\n",
       "<td>0.8877582</td>\n",
       "<td>0.9749781</td>\n",
       "<td>0.9323158</td>\n",
       "<td>0.9872079</td>\n",
       "<td>0.0491113</td>\n",
       "<td>0.1031524</td>\n",
       "<td>391.0807497</td>\n",
       "<td>415.7286025</td>\n",
       "<td>0.1014999</td></tr>\n",
       "<tr><td>3</td>\n",
       "<td>0.0300019</td>\n",
       "<td>0.9565777</td>\n",
       "<td>4.8528125</td>\n",
       "<td>5.0557949</td>\n",
       "<td>0.8772741</td>\n",
       "<td>0.9600168</td>\n",
       "<td>0.9139685</td>\n",
       "<td>0.9781442</td>\n",
       "<td>0.0485313</td>\n",
       "<td>0.1516837</td>\n",
       "<td>385.2812549</td>\n",
       "<td>405.5794866</td>\n",
       "<td>0.1485330</td></tr>\n",
       "<tr><td>4</td>\n",
       "<td>0.0400519</td>\n",
       "<td>0.9401411</td>\n",
       "<td>4.9257372</td>\n",
       "<td>5.0231603</td>\n",
       "<td>0.8904572</td>\n",
       "<td>0.9485588</td>\n",
       "<td>0.9080690</td>\n",
       "<td>0.9707205</td>\n",
       "<td>0.0495036</td>\n",
       "<td>0.2011873</td>\n",
       "<td>392.5737247</td>\n",
       "<td>402.3160297</td>\n",
       "<td>0.1966928</td></tr>\n",
       "<tr><td>5</td>\n",
       "<td>0.0500002</td>\n",
       "<td>0.9271963</td>\n",
       "<td>4.6640450</td>\n",
       "<td>4.9517093</td>\n",
       "<td>0.8431494</td>\n",
       "<td>0.9347513</td>\n",
       "<td>0.8951523</td>\n",
       "<td>0.9635640</td>\n",
       "<td>0.0463990</td>\n",
       "<td>0.2475862</td>\n",
       "<td>366.4045032</td>\n",
       "<td>395.1709339</td>\n",
       "<td>0.2411870</td></tr>\n",
       "<tr><td>6</td>\n",
       "<td>0.1000003</td>\n",
       "<td>0.7651266</td>\n",
       "<td>3.6569776</td>\n",
       "<td>4.3043435</td>\n",
       "<td>0.6610953</td>\n",
       "<td>0.8582987</td>\n",
       "<td>0.7781238</td>\n",
       "<td>0.9109313</td>\n",
       "<td>0.1828494</td>\n",
       "<td>0.4304357</td>\n",
       "<td>265.6977567</td>\n",
       "<td>330.4343453</td>\n",
       "<td>0.4033519</td></tr>\n",
       "<tr><td>7</td>\n",
       "<td>0.1500005</td>\n",
       "<td>0.5497443</td>\n",
       "<td>2.4639138</td>\n",
       "<td>3.6908669</td>\n",
       "<td>0.4454175</td>\n",
       "<td>0.6667315</td>\n",
       "<td>0.6672217</td>\n",
       "<td>0.8295314</td>\n",
       "<td>0.1231961</td>\n",
       "<td>0.5536317</td>\n",
       "<td>146.3913797</td>\n",
       "<td>269.0866901</td>\n",
       "<td>0.4926998</td></tr>\n",
       "<tr><td>8</td>\n",
       "<td>0.2000006</td>\n",
       "<td>0.3520121</td>\n",
       "<td>3.0660750</td>\n",
       "<td>3.5346689</td>\n",
       "<td>0.5542741</td>\n",
       "<td>0.4509449</td>\n",
       "<td>0.6389848</td>\n",
       "<td>0.7348848</td>\n",
       "<td>0.1533042</td>\n",
       "<td>0.7069360</td>\n",
       "<td>206.6074950</td>\n",
       "<td>253.4668913</td>\n",
       "<td>0.6187998</td></tr>\n",
       "<tr><td>9</td>\n",
       "<td>0.3000009</td>\n",
       "<td>0.1033972</td>\n",
       "<td>1.9061045</td>\n",
       "<td>2.9918141</td>\n",
       "<td>0.3445788</td>\n",
       "<td>0.1949264</td>\n",
       "<td>0.5408495</td>\n",
       "<td>0.5548986</td>\n",
       "<td>0.1906110</td>\n",
       "<td>0.8975470</td>\n",
       "<td>90.6104456</td>\n",
       "<td>199.1814094</td>\n",
       "<td>0.7294054</td></tr>\n",
       "<tr><td>10</td>\n",
       "<td>0.4000012</td>\n",
       "<td>0.0580555</td>\n",
       "<td>0.4247539</td>\n",
       "<td>2.3500490</td>\n",
       "<td>0.0767855</td>\n",
       "<td>0.0761308</td>\n",
       "<td>0.4248335</td>\n",
       "<td>0.4352067</td>\n",
       "<td>0.0424755</td>\n",
       "<td>0.9400225</td>\n",
       "<td>-57.5246098</td>\n",
       "<td>135.0049046</td>\n",
       "<td>0.6591867</td></tr>\n",
       "<tr><td>11</td>\n",
       "<td>0.5000015</td>\n",
       "<td>0.0425256</td>\n",
       "<td>0.1717780</td>\n",
       "<td>1.9143948</td>\n",
       "<td>0.0310534</td>\n",
       "<td>0.0494599</td>\n",
       "<td>0.3460775</td>\n",
       "<td>0.3580573</td>\n",
       "<td>0.0171779</td>\n",
       "<td>0.9572004</td>\n",
       "<td>-82.8222017</td>\n",
       "<td>91.4394833</td>\n",
       "<td>0.5580880</td></tr>\n",
       "<tr><td>12</td>\n",
       "<td>0.5999988</td>\n",
       "<td>0.0333775</td>\n",
       "<td>0.1021829</td>\n",
       "<td>1.6123673</td>\n",
       "<td>0.0184723</td>\n",
       "<td>0.0375849</td>\n",
       "<td>0.2914780</td>\n",
       "<td>0.3046466</td>\n",
       "<td>0.0102180</td>\n",
       "<td>0.9674184</td>\n",
       "<td>-89.7817095</td>\n",
       "<td>61.2367274</td>\n",
       "<td>0.4484974</td></tr>\n",
       "<tr><td>13</td>\n",
       "<td>0.6999991</td>\n",
       "<td>0.0269788</td>\n",
       "<td>0.0818803</td>\n",
       "<td>1.3937253</td>\n",
       "<td>0.0148020</td>\n",
       "<td>0.0300302</td>\n",
       "<td>0.2519527</td>\n",
       "<td>0.2654155</td>\n",
       "<td>0.0081881</td>\n",
       "<td>0.9756064</td>\n",
       "<td>-91.8119730</td>\n",
       "<td>39.3725310</td>\n",
       "<td>0.3364251</td></tr>\n",
       "<tr><td>14</td>\n",
       "<td>0.7999994</td>\n",
       "<td>0.0215518</td>\n",
       "<td>0.0810274</td>\n",
       "<td>1.2296374</td>\n",
       "<td>0.0146478</td>\n",
       "<td>0.0242362</td>\n",
       "<td>0.2222895</td>\n",
       "<td>0.2352680</td>\n",
       "<td>0.0081028</td>\n",
       "<td>0.9837092</td>\n",
       "<td>-91.8972649</td>\n",
       "<td>22.9637433</td>\n",
       "<td>0.2242487</td></tr>\n",
       "<tr><td>15</td>\n",
       "<td>0.8999997</td>\n",
       "<td>0.0159245</td>\n",
       "<td>0.0955270</td>\n",
       "<td>1.1036247</td>\n",
       "<td>0.0172690</td>\n",
       "<td>0.0187104</td>\n",
       "<td>0.1995093</td>\n",
       "<td>0.2112060</td>\n",
       "<td>0.0095527</td>\n",
       "<td>0.9932619</td>\n",
       "<td>-90.4473018</td>\n",
       "<td>10.3624729</td>\n",
       "<td>0.1138422</td></tr>\n",
       "<tr><td>16</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0004474</td>\n",
       "<td>0.0673806</td>\n",
       "<td>1.0</td>\n",
       "<td>0.0121808</td>\n",
       "<td>0.0116174</td>\n",
       "<td>0.1807764</td>\n",
       "<td>0.1912470</td>\n",
       "<td>0.0067381</td>\n",
       "<td>1.0</td>\n",
       "<td>-93.2619361</td>\n",
       "<td>0.0</td>\n",
       "<td>0.0</td></tr></tbody>\n",
       "  </table>\n",
       "</div>\n",
       "</div>"
      ],
      "text/plain": [
       "ModelMetricsBinomial: drf\n",
       "** Reported on test data. **\n",
       "\n",
       "MSE: 0.08607764751628483\n",
       "RMSE: 0.29338992401969916\n",
       "LogLoss: 0.2850543542126062\n",
       "Mean Per-Class Error: 0.13789716372509686\n",
       "AUC: 0.9149414130727681\n",
       "AUCPR: 0.7308461869304076\n",
       "Gini: 0.8298828261455362\n",
       "\n",
       "Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.18633682950984531\n",
       "       0       1      Error    Rate\n",
       "-----  ------  -----  -------  ------------------\n",
       "0      235455  30202  0.1137   (30202.0/265657.0)\n",
       "1      9503    49119  0.1621   (9503.0/58622.0)\n",
       "Total  244958  79321  0.1224   (39705.0/324279.0)\n",
       "\n",
       "Maximum Metrics: Maximum metrics at their respective thresholds\n",
       "metric                       threshold    value     idx\n",
       "---------------------------  -----------  --------  -----\n",
       "max f1                       0.186337     0.712164  266\n",
       "max f2                       0.122874     0.79482   292\n",
       "max f0point5                 0.806683     0.676829  71\n",
       "max accuracy                 0.250667     0.879502  245\n",
       "max precision                0.981703     0.981667  5\n",
       "max recall                   0.00699958   1         397\n",
       "max specificity              0.999939     0.999721  0\n",
       "max absolute_mcc             0.172583     0.648893  270\n",
       "max min_per_class_accuracy   0.14104      0.867476  283\n",
       "max mean_per_class_accuracy  0.143675     0.868172  282\n",
       "max tns                      0.999939     265583    0\n",
       "max fns                      0.999939     55934     0\n",
       "max fps                      0.0024483    265657    399\n",
       "max tps                      0.00699958   58622     397\n",
       "max tnr                      0.999939     0.999721  0\n",
       "max fnr                      0.999939     0.954147  0\n",
       "max fpr                      0.0024483    1         399\n",
       "max tpr                      0.00699958   1         397\n",
       "\n",
       "Gains/Lift Table: Avg response rate: 18.08 %, avg score: 19.12 %\n",
       "group    cumulative_data_fraction    lower_threshold    lift       cumulative_lift    response_rate    score      cumulative_response_rate    cumulative_score    capture_rate    cumulative_capture_rate    gain      cumulative_gain    kolmogorov_smirnov\n",
       "-------  --------------------------  -----------------  ---------  -----------------  ---------------  ---------  --------------------------  ------------------  --------------  -------------------------  --------  -----------------  --------------------\n",
       "1        0.0100006                   0.993234           5.40376    5.40376            0.976873         0.999438   0.976873                    0.999438            0.0540411       0.0540411                  440.376   440.376            0.0537588\n",
       "2        0.0200013                   0.963238           4.91081    5.15729            0.887758         0.974978   0.932316                    0.987208            0.0491113       0.103152                   391.081   415.729            0.1015\n",
       "3        0.0300019                   0.956578           4.85281    5.05579            0.877274         0.960017   0.913969                    0.978144            0.0485313       0.151684                   385.281   405.579            0.148533\n",
       "4        0.0400519                   0.940141           4.92574    5.02316            0.890457         0.948559   0.908069                    0.970721            0.0495036       0.201187                   392.574   402.316            0.196693\n",
       "5        0.0500002                   0.927196           4.66405    4.95171            0.843149         0.934751   0.895152                    0.963564            0.046399        0.247586                   366.405   395.171            0.241187\n",
       "6        0.1                         0.765127           3.65698    4.30434            0.661095         0.858299   0.778124                    0.910931            0.182849        0.430436                   265.698   330.434            0.403352\n",
       "7        0.15                        0.549744           2.46391    3.69087            0.445418         0.666732   0.667222                    0.829531            0.123196        0.553632                   146.391   269.087            0.4927\n",
       "8        0.200001                    0.352012           3.06607    3.53467            0.554274         0.450945   0.638985                    0.734885            0.153304        0.706936                   206.607   253.467            0.6188\n",
       "9        0.300001                    0.103397           1.9061     2.99181            0.344579         0.194926   0.540849                    0.554899            0.190611        0.897547                   90.6104   199.181            0.729405\n",
       "10       0.400001                    0.0580555          0.424754   2.35005            0.0767855        0.0761308  0.424833                    0.435207            0.0424755       0.940023                   -57.5246  135.005            0.659187\n",
       "11       0.500002                    0.0425256          0.171778   1.91439            0.0310534        0.0494599  0.346077                    0.358057            0.0171779       0.9572                     -82.8222  91.4395            0.558088\n",
       "12       0.599999                    0.0333775          0.102183   1.61237            0.0184723        0.0375849  0.291478                    0.304647            0.010218        0.967418                   -89.7817  61.2367            0.448497\n",
       "13       0.699999                    0.0269788          0.0818803  1.39373            0.014802         0.0300302  0.251953                    0.265416            0.00818805      0.975606                   -91.812   39.3725            0.336425\n",
       "14       0.799999                    0.0215518          0.0810274  1.22964            0.0146478        0.0242362  0.222289                    0.235268            0.00810276      0.983709                   -91.8973  22.9637            0.224249\n",
       "15       0.9                         0.0159245          0.095527   1.10362            0.017269         0.0187104  0.199509                    0.211206            0.00955273      0.993262                   -90.4473  10.3625            0.113842\n",
       "16       1                           0.000447427        0.0673806  1                  0.0121808        0.0116174  0.180776                    0.191247            0.00673808      1                          -93.2619  0                  0"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf_4model.model_performance(test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "f43ef033",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<h2o.plot._plot_result._MObject at 0x27fc9733c40>"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x1200 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "rf_4model.shap_summary_plot(test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "650d682c-5c82-4a5d-8579-e30734e1f244",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('RollingMean_400', 85204.4375, 1.0, 0.14312479017508348),\n",
       " ('RollingMean_800', 84269.3125, 0.9890249260785273, 0.1415539850229167),\n",
       " ('RollingSD_800', 77863.6796875, 0.9138453579662444, 0.1307939251113928),\n",
       " ('RollingMean_200', 59378.46875, 0.6968940878225973, 0.09974282009386544),\n",
       " ('RollingSD_400', 50173.265625, 0.5888574245326131, 0.0842800953292703),\n",
       " ('RollingSD_200', 31352.2578125, 0.3679650817775776, 0.05266492512117323),\n",
       " ('RollingMean_100',\n",
       "  17826.767578125,\n",
       "  0.20922346419017201,\n",
       "  0.029945064411922465),\n",
       " ('RollingSD_100',\n",
       "  16165.5419921875,\n",
       "  0.18972652676907234,\n",
       "  0.027154569334470836),\n",
       " ('meanaddsd', 14138.111328125, 0.16593163153180843, 0.023748929946399346),\n",
       " ('Lag50', 10032.2060546875, 0.11774276491981418, 0.016851908523782586),\n",
       " ('RollingMean_50', 9354.158203125, 0.10978487127651068, 0.015712936665849143),\n",
       " ('Lag10', 9168.6162109375, 0.1076072617806731, 0.015401266763674119),\n",
       " ('RollingSD_25', 9146.203125, 0.10734421109229199, 0.015363617689094161),\n",
       " ('Lag30', 8775.1923828125, 0.10298985170593375, 0.014740400915574731),\n",
       " ('RollingSD_50', 8772.857421875, 0.1029624474884304, 0.014736478692694652),\n",
       " ('meandiffsd', 8687.59375, 0.1019617522854957, 0.014593254401745411),\n",
       " ('RollingMean_35', 8406.134765625, 0.09865841512802663, 0.014120464964205095),\n",
       " ('Lag5', 7581.0, 0.08897423916448013, 0.012734419311403915),\n",
       " ('RollingMean_3',\n",
       "  7374.55419921875,\n",
       "  0.08655129258049207,\n",
       "  0.012387635589965187),\n",
       " ('RollingMean_25', 7057.078125, 0.08282524164307757, 0.011854345331366064),\n",
       " ('Lag2', 6974.90869140625, 0.08186086189943159, 0.011716318682907633),\n",
       " ('RollingSD_35', 6607.6455078125, 0.07755048565178897, 0.011099396986888119),\n",
       " ('RollingMean_5', 6508.87109375, 0.07639122192139347, 0.01093347760871768),\n",
       " ('Lag1', 5733.09521484375, 0.06728634544232218, 0.009630344073080548),\n",
       " ('RollingMean_4', 5614.94873046875, 0.065899721836304, 0.009431883860417378),\n",
       " ('Station', 5606.0693359375, 0.06579550901838299, 0.009416968422718879),\n",
       " ('RollingMean_2', 5518.765625, 0.06477087094202107, 0.009270317313034177),\n",
       " ('RollingSD_5', 3948.01025390625, 0.046335735200484715, 0.006631792378177605),\n",
       " ('RollingSD_4', 3489.4306640625, 0.0409536259665173, 0.005861479123366638),\n",
       " ('RollingSD_3', 3127.072998046875, 0.03670082321765078, 0.005252797622279101),\n",
       " ('diff30', 3079.616943359375, 0.036143856279309106, 0.005173081846094489),\n",
       " ('diff50', 2938.353515625, 0.034485921177814245, 0.004935790232569131),\n",
       " ('diff5', 1759.5457763671875, 0.020650870165854773, 0.0029556514594208556),\n",
       " ('RollingSD_2',\n",
       "  1551.4398193359375,\n",
       "  0.018208439194683228,\n",
       "  0.0026060790391548034),\n",
       " ('diff2', 1293.2437744140625, 0.015178127012622581, 0.0021723662439323737),\n",
       " ('diff1', 837.2649536132812, 0.009826541647121151, 0.0014064217113909342)]"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf_4model.varimp()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "d9d83145-d9ad-44b6-a972-691178127bec",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "drf prediction progress: |███████████████████████████████████████████████████████| (done) 100%\n",
      "prec: 0.6214539571736112 recall: 0.8322131622940193 f1: 0.7115551504102096\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import precision_score\n",
    "from sklearn.metrics import recall_score, f1_score\n",
    "\n",
    "predictions = rf_4model.predict(test)\n",
    "fixed_threshold = 0.19283058598703115 #treshhold of training\n",
    "predictions['prediction'] = (predictions['p1'] >= fixed_threshold).ifelse(1, 0)\n",
    "y_true = test.as_data_frame()['Anomaly'].values\n",
    "y_pred = predictions.as_data_frame()['prediction'].values\n",
    "precision = precision_score(y_true, y_pred)\n",
    "recall = recall_score(y_true, y_pred)\n",
    "f1 = f1_score(y_true, y_pred)\n",
    "print('prec:',precision,'recall:',recall,'f1:',f1)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}