NCAA-Women Stage 1.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import re\n",
    "import sklearn\n",
    "import numpy as np \n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from collections import Counter\n",
    "from sklearn.metrics import *\n",
    "from sklearn.linear_model import *\n",
    "from sklearn.model_selection import *\n",
    "\n",
    "pd.set_option('display.max_columns', None)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WNCAATourneySlots.csv\n",
      "WTeams.csv\n",
      "WSampleSubmissionStage1.csv\n",
      "WNCAATourneyDetailedResults.csv\n",
      "WNCAATourneyCompactResults.csv\n",
      "WSeasons.csv\n",
      "Cities.csv\n",
      "WRegularSeasonCompactResults.csv\n",
      "WTeamSpellings.csv\n",
      "WRegularSeasonDetailedResults.csv\n",
      "WNCAATourneySeeds.csv\n",
      "WTeamConferences.csv\n",
      "Conferences.csv\n",
      "WGameCities.csv\n"
     ]
    }
   ],
   "source": [
    "DATA_PATH = 'WDataFiles_Stage1/'\n",
    "\n",
    "for filename in os.listdir(DATA_PATH):\n",
    "    print(filename)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Seeds"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Data Section 1 file: WNCAATourneySeeds.csv\n",
    "\n",
    "This file identifies the seeds for all teams in each NCAA® tournament, for all seasons of historical data. Thus, there are exactly 64 rows for each year, since there are no play-in teams in the women's tournament. We will not know the seeds of the respective tournament teams, or even exactly which 64 teams it will be, until Selection Monday on March 14, 2022 (DayNum=133)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Season</th>\n",
       "      <th>Seed</th>\n",
       "      <th>TeamID</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1998</td>\n",
       "      <td>W01</td>\n",
       "      <td>3330</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1998</td>\n",
       "      <td>W02</td>\n",
       "      <td>3163</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1998</td>\n",
       "      <td>W03</td>\n",
       "      <td>3112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1998</td>\n",
       "      <td>W04</td>\n",
       "      <td>3301</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1998</td>\n",
       "      <td>W05</td>\n",
       "      <td>3272</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1467</th>\n",
       "      <td>2021</td>\n",
       "      <td>Z12</td>\n",
       "      <td>3141</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1468</th>\n",
       "      <td>2021</td>\n",
       "      <td>Z13</td>\n",
       "      <td>3226</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1469</th>\n",
       "      <td>2021</td>\n",
       "      <td>Z14</td>\n",
       "      <td>3292</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1470</th>\n",
       "      <td>2021</td>\n",
       "      <td>Z15</td>\n",
       "      <td>3238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1471</th>\n",
       "      <td>2021</td>\n",
       "      <td>Z16</td>\n",
       "      <td>3219</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1472 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Season Seed  TeamID\n",
       "0       1998  W01    3330\n",
       "1       1998  W02    3163\n",
       "2       1998  W03    3112\n",
       "3       1998  W04    3301\n",
       "4       1998  W05    3272\n",
       "...      ...  ...     ...\n",
       "1467    2021  Z12    3141\n",
       "1468    2021  Z13    3226\n",
       "1469    2021  Z14    3292\n",
       "1470    2021  Z15    3238\n",
       "1471    2021  Z16    3219\n",
       "\n",
       "[1472 rows x 3 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_seeds = pd.read_csv(DATA_PATH + \"WNCAATourneySeeds.csv\")\n",
    "df_seeds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['W01', 'W02', 'W03', 'W04', 'W05', 'W06', 'W07', 'W08', 'W09',\n",
       "       'W10', 'W11', 'W12', 'W13', 'W14', 'W15', 'W16', 'X01', 'X02',\n",
       "       'X03', 'X04', 'X05', 'X06', 'X07', 'X08', 'X09', 'X10', 'X11',\n",
       "       'X12', 'X13', 'X14', 'X15', 'X16', 'Y01', 'Y02', 'Y03', 'Y04',\n",
       "       'Y05', 'Y06', 'Y07', 'Y08', 'Y09', 'Y10', 'Y11', 'Y12', 'Y13',\n",
       "       'Y14', 'Y15', 'Y16', 'Z01', 'Z02', 'Z03', 'Z04', 'Z05', 'Z06',\n",
       "       'Z07', 'Z08', 'Z09', 'Z10', 'Z11', 'Z12', 'Z13', 'Z14', 'Z15',\n",
       "       'Z16'], dtype=object)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_seeds.Seed.unique()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Slots"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Slot</th>\n",
       "      <th>StrongSeed</th>\n",
       "      <th>WeakSeed</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>R1W1</td>\n",
       "      <td>W01</td>\n",
       "      <td>W16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>R1W2</td>\n",
       "      <td>W02</td>\n",
       "      <td>W15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>R1W3</td>\n",
       "      <td>W03</td>\n",
       "      <td>W14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>R1W4</td>\n",
       "      <td>W04</td>\n",
       "      <td>W13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>R1W5</td>\n",
       "      <td>W05</td>\n",
       "      <td>W12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58</th>\n",
       "      <td>R4Y1</td>\n",
       "      <td>R3Y1</td>\n",
       "      <td>R3Y2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>59</th>\n",
       "      <td>R4Z1</td>\n",
       "      <td>R3Z1</td>\n",
       "      <td>R3Z2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60</th>\n",
       "      <td>R5WX</td>\n",
       "      <td>R4W1</td>\n",
       "      <td>R4X1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>61</th>\n",
       "      <td>R5YZ</td>\n",
       "      <td>R4Y1</td>\n",
       "      <td>R4Z1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>62</th>\n",
       "      <td>R6CH</td>\n",
       "      <td>R5WX</td>\n",
       "      <td>R5YZ</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>63 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    Slot StrongSeed WeakSeed\n",
       "0   R1W1        W01      W16\n",
       "1   R1W2        W02      W15\n",
       "2   R1W3        W03      W14\n",
       "3   R1W4        W04      W13\n",
       "4   R1W5        W05      W12\n",
       "..   ...        ...      ...\n",
       "58  R4Y1       R3Y1     R3Y2\n",
       "59  R4Z1       R3Z1     R3Z2\n",
       "60  R5WX       R4W1     R4X1\n",
       "61  R5YZ       R4Y1     R4Z1\n",
       "62  R6CH       R5WX     R5YZ\n",
       "\n",
       "[63 rows x 3 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_slot = pd.read_csv(DATA_PATH + \"WNCAATourneySlots.csv\")\n",
    "df_slot"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Season results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Data Section 1 file: WRegularSeasonCompactResults.csv\n",
    "\n",
    "This file identifies the game-by-game results for many seasons of historical data, starting with the 1998 season. For each season, the file includes all games played from DayNum 0 through 132. It is important to realize that the \"Regular Season\" games are simply defined to be all games played on DayNum=132 or earlier (DayNum=133 is Selection Monday). Thus a game played before Selection Monday will show up here whether it was a pre-season tournament, a non-conference game, a regular conference game, a conference tournament game, or whatever."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Season</th>\n",
       "      <th>DayNum</th>\n",
       "      <th>WTeamID</th>\n",
       "      <th>WScore</th>\n",
       "      <th>LTeamID</th>\n",
       "      <th>LScore</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1998</td>\n",
       "      <td>18</td>\n",
       "      <td>3104</td>\n",
       "      <td>91</td>\n",
       "      <td>3202</td>\n",
       "      <td>41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1998</td>\n",
       "      <td>18</td>\n",
       "      <td>3163</td>\n",
       "      <td>87</td>\n",
       "      <td>3221</td>\n",
       "      <td>76</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1998</td>\n",
       "      <td>18</td>\n",
       "      <td>3222</td>\n",
       "      <td>66</td>\n",
       "      <td>3261</td>\n",
       "      <td>59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1998</td>\n",
       "      <td>18</td>\n",
       "      <td>3307</td>\n",
       "      <td>69</td>\n",
       "      <td>3365</td>\n",
       "      <td>62</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1998</td>\n",
       "      <td>18</td>\n",
       "      <td>3349</td>\n",
       "      <td>115</td>\n",
       "      <td>3411</td>\n",
       "      <td>35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119272</th>\n",
       "      <td>2022</td>\n",
       "      <td>98</td>\n",
       "      <td>3359</td>\n",
       "      <td>73</td>\n",
       "      <td>3190</td>\n",
       "      <td>61</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119273</th>\n",
       "      <td>2022</td>\n",
       "      <td>98</td>\n",
       "      <td>3360</td>\n",
       "      <td>81</td>\n",
       "      <td>3334</td>\n",
       "      <td>77</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119274</th>\n",
       "      <td>2022</td>\n",
       "      <td>98</td>\n",
       "      <td>3379</td>\n",
       "      <td>60</td>\n",
       "      <td>3427</td>\n",
       "      <td>57</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119275</th>\n",
       "      <td>2022</td>\n",
       "      <td>98</td>\n",
       "      <td>3388</td>\n",
       "      <td>88</td>\n",
       "      <td>3365</td>\n",
       "      <td>81</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>119276</th>\n",
       "      <td>2022</td>\n",
       "      <td>98</td>\n",
       "      <td>3453</td>\n",
       "      <td>65</td>\n",
       "      <td>3454</td>\n",
       "      <td>56</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>119277 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        Season  DayNum  WTeamID  WScore  LTeamID  LScore\n",
       "0         1998      18     3104      91     3202      41\n",
       "1         1998      18     3163      87     3221      76\n",
       "2         1998      18     3222      66     3261      59\n",
       "3         1998      18     3307      69     3365      62\n",
       "4         1998      18     3349     115     3411      35\n",
       "...        ...     ...      ...     ...      ...     ...\n",
       "119272    2022      98     3359      73     3190      61\n",
       "119273    2022      98     3360      81     3334      77\n",
       "119274    2022      98     3379      60     3427      57\n",
       "119275    2022      98     3388      88     3365      81\n",
       "119276    2022      98     3453      65     3454      56\n",
       "\n",
       "[119277 rows x 6 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_season = pd.read_csv(DATA_PATH + \"WRegularSeasonCompactResults.csv\")\n",
    "df_season.drop(['NumOT', 'WLoc'], axis=1, inplace=True)\n",
    "df_season"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Season</th>\n",
       "      <th>DayNum</th>\n",
       "      <th>WTeamID</th>\n",
       "      <th>WScore</th>\n",
       "      <th>LTeamID</th>\n",
       "      <th>LScore</th>\n",
       "      <th>ScoreGap</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1998</td>\n",
       "      <td>18</td>\n",
       "      <td>3104</td>\n",
       "      <td>91</td>\n",
       "      <td>3202</td>\n",
       "      <td>41</td>\n",
       "      <td>50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1998</td>\n",
       "      <td>18</td>\n",
       "      <td>3163</td>\n",
       "      <td>87</td>\n",
       "      <td>3221</td>\n",
       "      <td>76</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1998</td>\n",
       "      <td>18</td>\n",
       "      <td>3222</td>\n",
       "      <td>66</td>\n",
       "      <td>3261</td>\n",
       "      <td>59</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1998</td>\n",
       "      <td>18</td>\n",
       "      <td>3307</td>\n",
       "      <td>69</td>\n",
       "      <td>3365</td>\n",
       "      <td>62</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1998</td>\n",
       "      <td>18</td>\n",
       "      <td>3349</td>\n",
       "      <td>115</td>\n",
       "      <td>3411</td>\n",
       "      <td>35</td>\n",
       "      <td>80</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Season  DayNum  WTeamID  WScore  LTeamID  LScore  ScoreGap\n",
       "0    1998      18     3104      91     3202      41        50\n",
       "1    1998      18     3163      87     3221      76        11\n",
       "2    1998      18     3222      66     3261      59         7\n",
       "3    1998      18     3307      69     3365      62         7\n",
       "4    1998      18     3349     115     3411      35        80"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_season['ScoreGap'] = df_season['WScore'] - df_season['LScore']\n",
    "df_season.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "win=df_season.groupby(['Season', 'WTeamID']).agg(NumWins=('DayNum','count'),\n",
    "                                             AvgWinsGap=('ScoreGap','mean'),\n",
    "                                             W_TotalPoints=('WScore', 'sum'),\n",
    "                                             W_MaxPoints=('WScore', 'max'),\n",
    "                                             W_MinPoints=('WScore', 'min')).reset_index().rename(columns={\"WTeamID\": \"TeamID\"})\n",
    "                                             \n",
    "loss=df_season.groupby(['Season', 'LTeamID']).agg(NumLosses=('DayNum','count'),\n",
    "                                             AvgLossesGap=('ScoreGap','mean'),\n",
    "                                             L_TotalPoints=('LScore', 'sum'),\n",
    "                                             L_MaxPoints=('LScore', 'max'),\n",
    "                                             L_MinPoints=('LScore', 'min')).reset_index().rename(columns={\"LTeamID\": \"TeamID\"})                                               "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# create team list containing all historical data\n",
    "df_features = win.merge(loss, on=['Season','TeamID'], how='outer').drop_duplicates().sort_values(['Season', 'TeamID']).reset_index(drop=True)\n",
    "df_features.fillna(0, inplace=True)  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Season</th>\n",
       "      <th>TeamID</th>\n",
       "      <th>NumWins</th>\n",
       "      <th>AvgWinsGap</th>\n",
       "      <th>W_TotalPoints</th>\n",
       "      <th>W_MaxPoints</th>\n",
       "      <th>W_MinPoints</th>\n",
       "      <th>NumLosses</th>\n",
       "      <th>AvgLossesGap</th>\n",
       "      <th>L_TotalPoints</th>\n",
       "      <th>L_MaxPoints</th>\n",
       "      <th>L_MinPoints</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1998</td>\n",
       "      <td>3102</td>\n",
       "      <td>4.0</td>\n",
       "      <td>15.500000</td>\n",
       "      <td>263.0</td>\n",
       "      <td>81.0</td>\n",
       "      <td>59.0</td>\n",
       "      <td>20.0</td>\n",
       "      <td>27.850000</td>\n",
       "      <td>1112.0</td>\n",
       "      <td>74.0</td>\n",
       "      <td>39.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1998</td>\n",
       "      <td>3103</td>\n",
       "      <td>11.0</td>\n",
       "      <td>15.000000</td>\n",
       "      <td>842.0</td>\n",
       "      <td>89.0</td>\n",
       "      <td>64.0</td>\n",
       "      <td>18.0</td>\n",
       "      <td>18.611111</td>\n",
       "      <td>1166.0</td>\n",
       "      <td>91.0</td>\n",
       "      <td>47.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1998</td>\n",
       "      <td>3104</td>\n",
       "      <td>21.0</td>\n",
       "      <td>21.619048</td>\n",
       "      <td>1657.0</td>\n",
       "      <td>104.0</td>\n",
       "      <td>62.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>5.666667</td>\n",
       "      <td>640.0</td>\n",
       "      <td>90.0</td>\n",
       "      <td>58.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1998</td>\n",
       "      <td>3106</td>\n",
       "      <td>6.0</td>\n",
       "      <td>8.333333</td>\n",
       "      <td>421.0</td>\n",
       "      <td>80.0</td>\n",
       "      <td>58.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>14.466667</td>\n",
       "      <td>865.0</td>\n",
       "      <td>73.0</td>\n",
       "      <td>36.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1998</td>\n",
       "      <td>3108</td>\n",
       "      <td>12.0</td>\n",
       "      <td>17.750000</td>\n",
       "      <td>898.0</td>\n",
       "      <td>84.0</td>\n",
       "      <td>66.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>16.636364</td>\n",
       "      <td>662.0</td>\n",
       "      <td>75.0</td>\n",
       "      <td>47.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Season  TeamID  NumWins  AvgWinsGap  W_TotalPoints  W_MaxPoints  \\\n",
       "0    1998    3102      4.0   15.500000          263.0         81.0   \n",
       "1    1998    3103     11.0   15.000000          842.0         89.0   \n",
       "2    1998    3104     21.0   21.619048         1657.0        104.0   \n",
       "3    1998    3106      6.0    8.333333          421.0         80.0   \n",
       "4    1998    3108     12.0   17.750000          898.0         84.0   \n",
       "\n",
       "   W_MinPoints  NumLosses  AvgLossesGap  L_TotalPoints  L_MaxPoints  \\\n",
       "0         59.0       20.0     27.850000         1112.0         74.0   \n",
       "1         64.0       18.0     18.611111         1166.0         91.0   \n",
       "2         62.0        9.0      5.666667          640.0         90.0   \n",
       "3         58.0       15.0     14.466667          865.0         73.0   \n",
       "4         66.0       11.0     16.636364          662.0         75.0   \n",
       "\n",
       "   L_MinPoints  \n",
       "0         39.0  \n",
       "1         47.0  \n",
       "2         58.0  \n",
       "3         36.0  \n",
       "4         47.0  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_features.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_features['WinRatio'] = df_features['NumWins'] / (df_features['NumWins'] + df_features['NumLosses'])\n",
    "\n",
    "df_features['AvgScoreGap'] = (\n",
    "    (df_features['NumWins']*df_features['AvgWinsGap']-df_features['NumLosses']*df_features['AvgLossesGap'])\n",
    "    /(df_features['NumWins'] + df_features['NumLosses']))\n",
    "\n",
    "df_features['PointsRatio'] = df_features['W_TotalPoints'] / (df_features['W_TotalPoints'] + df_features['L_TotalPoints'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Season</th>\n",
       "      <th>TeamID</th>\n",
       "      <th>WinRatio</th>\n",
       "      <th>AvgScoreGap</th>\n",
       "      <th>PointsRatio</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1998</td>\n",
       "      <td>3102</td>\n",
       "      <td>0.166667</td>\n",
       "      <td>-20.625000</td>\n",
       "      <td>0.191273</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1998</td>\n",
       "      <td>3103</td>\n",
       "      <td>0.379310</td>\n",
       "      <td>-5.862069</td>\n",
       "      <td>0.419323</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1998</td>\n",
       "      <td>3104</td>\n",
       "      <td>0.700000</td>\n",
       "      <td>13.433333</td>\n",
       "      <td>0.721376</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1998</td>\n",
       "      <td>3106</td>\n",
       "      <td>0.285714</td>\n",
       "      <td>-7.952381</td>\n",
       "      <td>0.327372</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1998</td>\n",
       "      <td>3108</td>\n",
       "      <td>0.521739</td>\n",
       "      <td>1.304348</td>\n",
       "      <td>0.575641</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8400</th>\n",
       "      <td>2022</td>\n",
       "      <td>3468</td>\n",
       "      <td>0.052632</td>\n",
       "      <td>-18.526316</td>\n",
       "      <td>0.062847</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8401</th>\n",
       "      <td>2022</td>\n",
       "      <td>3469</td>\n",
       "      <td>0.176471</td>\n",
       "      <td>-14.941176</td>\n",
       "      <td>0.209189</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8402</th>\n",
       "      <td>2022</td>\n",
       "      <td>3470</td>\n",
       "      <td>0.444444</td>\n",
       "      <td>-5.944444</td>\n",
       "      <td>0.491210</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8403</th>\n",
       "      <td>2022</td>\n",
       "      <td>3471</td>\n",
       "      <td>0.444444</td>\n",
       "      <td>-2.333333</td>\n",
       "      <td>0.513333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8404</th>\n",
       "      <td>2022</td>\n",
       "      <td>3472</td>\n",
       "      <td>0.250000</td>\n",
       "      <td>-11.900000</td>\n",
       "      <td>0.280244</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8405 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Season  TeamID  WinRatio  AvgScoreGap  PointsRatio\n",
       "0       1998    3102  0.166667   -20.625000     0.191273\n",
       "1       1998    3103  0.379310    -5.862069     0.419323\n",
       "2       1998    3104  0.700000    13.433333     0.721376\n",
       "3       1998    3106  0.285714    -7.952381     0.327372\n",
       "4       1998    3108  0.521739     1.304348     0.575641\n",
       "...      ...     ...       ...          ...          ...\n",
       "8400    2022    3468  0.052632   -18.526316     0.062847\n",
       "8401    2022    3469  0.176471   -14.941176     0.209189\n",
       "8402    2022    3470  0.444444    -5.944444     0.491210\n",
       "8403    2022    3471  0.444444    -2.333333     0.513333\n",
       "8404    2022    3472  0.250000   -11.900000     0.280244\n",
       "\n",
       "[8405 rows x 5 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cols = ['Season','TeamID','WinRatio', 'AvgScoreGap','PointsRatio']\n",
    "df_features_s = df_features[cols]\n",
    "df_features_s"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Tourney results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Data Section 1 file: WNCAATourneyCompactResults.csv\n",
    "\n",
    "This file identifies the game-by-game NCAA® tournament results for all seasons of historical data. The data is formatted exactly like the WRegularSeasonCompactResults data. Each season you will see 63 games listed, since there are no women's play-in games.\n",
    "\n",
    "Although the scheduling of the men's tournament rounds has been consistent for many years, there has been more variety in the scheduling of the women's rounds. There have been four different schedules over the course of the past 20+ years for the women's tournament, as follows:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Season</th>\n",
       "      <th>DayNum</th>\n",
       "      <th>WTeamID</th>\n",
       "      <th>WScore</th>\n",
       "      <th>LTeamID</th>\n",
       "      <th>LScore</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1998</td>\n",
       "      <td>137</td>\n",
       "      <td>3104</td>\n",
       "      <td>94</td>\n",
       "      <td>3422</td>\n",
       "      <td>46</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1998</td>\n",
       "      <td>137</td>\n",
       "      <td>3112</td>\n",
       "      <td>75</td>\n",
       "      <td>3365</td>\n",
       "      <td>63</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1998</td>\n",
       "      <td>137</td>\n",
       "      <td>3163</td>\n",
       "      <td>93</td>\n",
       "      <td>3193</td>\n",
       "      <td>52</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1998</td>\n",
       "      <td>137</td>\n",
       "      <td>3198</td>\n",
       "      <td>59</td>\n",
       "      <td>3266</td>\n",
       "      <td>45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1998</td>\n",
       "      <td>137</td>\n",
       "      <td>3203</td>\n",
       "      <td>74</td>\n",
       "      <td>3208</td>\n",
       "      <td>72</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1444</th>\n",
       "      <td>2021</td>\n",
       "      <td>148</td>\n",
       "      <td>3376</td>\n",
       "      <td>62</td>\n",
       "      <td>3400</td>\n",
       "      <td>34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1445</th>\n",
       "      <td>2021</td>\n",
       "      <td>148</td>\n",
       "      <td>3390</td>\n",
       "      <td>78</td>\n",
       "      <td>3257</td>\n",
       "      <td>63</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1446</th>\n",
       "      <td>2021</td>\n",
       "      <td>151</td>\n",
       "      <td>3112</td>\n",
       "      <td>69</td>\n",
       "      <td>3163</td>\n",
       "      <td>59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1447</th>\n",
       "      <td>2021</td>\n",
       "      <td>151</td>\n",
       "      <td>3390</td>\n",
       "      <td>66</td>\n",
       "      <td>3376</td>\n",
       "      <td>65</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1448</th>\n",
       "      <td>2021</td>\n",
       "      <td>153</td>\n",
       "      <td>3390</td>\n",
       "      <td>54</td>\n",
       "      <td>3112</td>\n",
       "      <td>53</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1449 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Season  DayNum  WTeamID  WScore  LTeamID  LScore\n",
       "0       1998     137     3104      94     3422      46\n",
       "1       1998     137     3112      75     3365      63\n",
       "2       1998     137     3163      93     3193      52\n",
       "3       1998     137     3198      59     3266      45\n",
       "4       1998     137     3203      74     3208      72\n",
       "...      ...     ...      ...     ...      ...     ...\n",
       "1444    2021     148     3376      62     3400      34\n",
       "1445    2021     148     3390      78     3257      63\n",
       "1446    2021     151     3112      69     3163      59\n",
       "1447    2021     151     3390      66     3376      65\n",
       "1448    2021     153     3390      54     3112      53\n",
       "\n",
       "[1449 rows x 6 columns]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_tourney = pd.read_csv(DATA_PATH + \"WNCAATourneyCompactResults.csv\")\n",
    "df_tourney.drop(['NumOT', 'WLoc'], axis=1, inplace=True)\n",
    "df_tourney"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_tourney = df_tourney.rename(columns = {'WTeamID' : 'W_TeamID', 'LTeamID' : 'L_TeamID', 'WScore' : 'W_Score', 'LScore' : 'L_Score'}) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## FiveThirtyEight Ratings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Season</th>\n",
       "      <th>TeamID</th>\n",
       "      <th>538rating</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2016</td>\n",
       "      <td>3163</td>\n",
       "      <td>113.09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2016</td>\n",
       "      <td>3124</td>\n",
       "      <td>101.59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2016</td>\n",
       "      <td>3323</td>\n",
       "      <td>101.07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2016</td>\n",
       "      <td>3376</td>\n",
       "      <td>101.07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2016</td>\n",
       "      <td>3268</td>\n",
       "      <td>99.01</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Season  TeamID  538rating\n",
       "0    2016    3163     113.09\n",
       "1    2016    3124     101.59\n",
       "2    2016    3323     101.07\n",
       "3    2016    3376     101.07\n",
       "4    2016    3268      99.01"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_538 = pd.read_csv(\"538ratingsWomen.csv\")\n",
    "df_538.drop('TeamName', axis=1, inplace=True)\n",
    "\n",
    "df_538.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Feature Engineering"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Season</th>\n",
       "      <th>DayNum</th>\n",
       "      <th>W_TeamID</th>\n",
       "      <th>W_Score</th>\n",
       "      <th>L_TeamID</th>\n",
       "      <th>L_Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2016</td>\n",
       "      <td>137</td>\n",
       "      <td>3107</td>\n",
       "      <td>61</td>\n",
       "      <td>3196</td>\n",
       "      <td>59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2016</td>\n",
       "      <td>137</td>\n",
       "      <td>3113</td>\n",
       "      <td>74</td>\n",
       "      <td>3308</td>\n",
       "      <td>52</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2016</td>\n",
       "      <td>137</td>\n",
       "      <td>3120</td>\n",
       "      <td>68</td>\n",
       "      <td>3385</td>\n",
       "      <td>57</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2016</td>\n",
       "      <td>137</td>\n",
       "      <td>3124</td>\n",
       "      <td>89</td>\n",
       "      <td>3225</td>\n",
       "      <td>59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2016</td>\n",
       "      <td>137</td>\n",
       "      <td>3177</td>\n",
       "      <td>97</td>\n",
       "      <td>3241</td>\n",
       "      <td>67</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Season  DayNum  W_TeamID  W_Score  L_TeamID  L_Score\n",
       "0    2016     137      3107       61      3196       59\n",
       "1    2016     137      3113       74      3308       52\n",
       "2    2016     137      3120       68      3385       57\n",
       "3    2016     137      3124       89      3225       59\n",
       "4    2016     137      3177       97      3241       67"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# train data\n",
    "df = df_tourney.copy()\n",
    "df = df[df['Season'] >= 2016].reset_index(drop=True)\n",
    "\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "# merge with seeds\n",
    "df = pd.merge(df, df_seeds, how='left', \n",
    "    left_on=['Season', 'W_TeamID'], right_on=['Season', 'TeamID']\n",
    "             ).drop('TeamID', axis=1).rename(columns={'Seed': 'W_Seed'})\n",
    "\n",
    "df = pd.merge(df, df_seeds, how='left', \n",
    "    left_on=['Season', 'L_TeamID'], right_on=['Season', 'TeamID']\n",
    "             ).drop('TeamID', axis=1).rename(columns={'Seed': 'L_Seed'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "# only get the number, remove the region letter\n",
    "def treat_seed(seed):\n",
    "    return int(re.sub(\"[^0-9]\", \"\", seed))\n",
    "\n",
    "df['W_Seed'] = df['W_Seed'].apply(treat_seed)\n",
    "df['L_Seed'] = df['L_Seed'].apply(treat_seed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 487,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Season</th>\n",
       "      <th>TeamID</th>\n",
       "      <th>WinRatio</th>\n",
       "      <th>AvgScoreGap</th>\n",
       "      <th>PointsRatio</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1998</td>\n",
       "      <td>3102</td>\n",
       "      <td>0.166667</td>\n",
       "      <td>-20.625000</td>\n",
       "      <td>0.191273</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1998</td>\n",
       "      <td>3103</td>\n",
       "      <td>0.379310</td>\n",
       "      <td>-5.862069</td>\n",
       "      <td>0.419323</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1998</td>\n",
       "      <td>3104</td>\n",
       "      <td>0.700000</td>\n",
       "      <td>13.433333</td>\n",
       "      <td>0.721376</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1998</td>\n",
       "      <td>3106</td>\n",
       "      <td>0.285714</td>\n",
       "      <td>-7.952381</td>\n",
       "      <td>0.327372</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1998</td>\n",
       "      <td>3108</td>\n",
       "      <td>0.521739</td>\n",
       "      <td>1.304348</td>\n",
       "      <td>0.575641</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8400</th>\n",
       "      <td>2022</td>\n",
       "      <td>3468</td>\n",
       "      <td>0.052632</td>\n",
       "      <td>-18.526316</td>\n",
       "      <td>0.062847</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8401</th>\n",
       "      <td>2022</td>\n",
       "      <td>3469</td>\n",
       "      <td>0.176471</td>\n",
       "      <td>-14.941176</td>\n",
       "      <td>0.209189</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8402</th>\n",
       "      <td>2022</td>\n",
       "      <td>3470</td>\n",
       "      <td>0.444444</td>\n",
       "      <td>-5.944444</td>\n",
       "      <td>0.491210</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8403</th>\n",
       "      <td>2022</td>\n",
       "      <td>3471</td>\n",
       "      <td>0.444444</td>\n",
       "      <td>-2.333333</td>\n",
       "      <td>0.513333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8404</th>\n",
       "      <td>2022</td>\n",
       "      <td>3472</td>\n",
       "      <td>0.250000</td>\n",
       "      <td>-11.900000</td>\n",
       "      <td>0.280244</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8405 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Season  TeamID  WinRatio  AvgScoreGap  PointsRatio\n",
       "0       1998    3102  0.166667   -20.625000     0.191273\n",
       "1       1998    3103  0.379310    -5.862069     0.419323\n",
       "2       1998    3104  0.700000    13.433333     0.721376\n",
       "3       1998    3106  0.285714    -7.952381     0.327372\n",
       "4       1998    3108  0.521739     1.304348     0.575641\n",
       "...      ...     ...       ...          ...          ...\n",
       "8400    2022    3468  0.052632   -18.526316     0.062847\n",
       "8401    2022    3469  0.176471   -14.941176     0.209189\n",
       "8402    2022    3470  0.444444    -5.944444     0.491210\n",
       "8403    2022    3471  0.444444    -2.333333     0.513333\n",
       "8404    2022    3472  0.250000   -11.900000     0.280244\n",
       "\n",
       "[8405 rows x 5 columns]"
      ]
     },
     "execution_count": 487,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_features_s"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "# merge with features\n",
    "df = pd.merge(df, df_features_s, how='left',\n",
    "    left_on=['Season', 'W_TeamID'], right_on=['Season', 'TeamID']\n",
    ").rename(columns={\n",
    "    'NumWins': 'W_NumWins',\n",
    "    'NumLosses': 'W_NumLosses',\n",
    "    'GapWins': 'W_GapWins',\n",
    "    'GapLosses': 'W_GapLosses',\n",
    "    'WinRatio': 'W_WinRatio',\n",
    "    'AvgScoreGap': 'W_AvgScoreGap',\n",
    "    'PointsRatio': 'W_PointsRatio',\n",
    "}).drop(columns='TeamID', axis=1)\n",
    "\n",
    "df = pd.merge(df, df_features_s, how='left',\n",
    "    left_on=['Season', 'L_TeamID'], right_on=['Season', 'TeamID']\n",
    ").rename(columns={\n",
    "    'NumWins': 'L_NumWins',\n",
    "    'NumLosses': 'L_NumLosses',\n",
    "    'GapWins': 'L_GapWins',\n",
    "    'GapLosses': 'L_GapLosses',\n",
    "    'WinRatio': 'L_WinRatio',\n",
    "    'AvgScoreGap': 'L_AvgScoreGap',\n",
    "    'PointsRatio': 'L_PointsRatio',\n",
    "}).drop(columns='TeamID', axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Season</th>\n",
       "      <th>DayNum</th>\n",
       "      <th>W_TeamID</th>\n",
       "      <th>W_Score</th>\n",
       "      <th>L_TeamID</th>\n",
       "      <th>L_Score</th>\n",
       "      <th>W_Seed</th>\n",
       "      <th>L_Seed</th>\n",
       "      <th>W_WinRatio</th>\n",
       "      <th>W_AvgScoreGap</th>\n",
       "      <th>W_PointsRatio</th>\n",
       "      <th>L_WinRatio</th>\n",
       "      <th>L_AvgScoreGap</th>\n",
       "      <th>L_PointsRatio</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2016</td>\n",
       "      <td>137</td>\n",
       "      <td>3107</td>\n",
       "      <td>61</td>\n",
       "      <td>3196</td>\n",
       "      <td>59</td>\n",
       "      <td>12</td>\n",
       "      <td>5</td>\n",
       "      <td>0.870968</td>\n",
       "      <td>18.677419</td>\n",
       "      <td>0.892371</td>\n",
       "      <td>0.733333</td>\n",
       "      <td>8.700000</td>\n",
       "      <td>0.766239</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2016</td>\n",
       "      <td>137</td>\n",
       "      <td>3113</td>\n",
       "      <td>74</td>\n",
       "      <td>3308</td>\n",
       "      <td>52</td>\n",
       "      <td>2</td>\n",
       "      <td>15</td>\n",
       "      <td>0.806452</td>\n",
       "      <td>9.258065</td>\n",
       "      <td>0.828093</td>\n",
       "      <td>0.857143</td>\n",
       "      <td>8.964286</td>\n",
       "      <td>0.881409</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2016</td>\n",
       "      <td>137</td>\n",
       "      <td>3120</td>\n",
       "      <td>68</td>\n",
       "      <td>3385</td>\n",
       "      <td>57</td>\n",
       "      <td>9</td>\n",
       "      <td>8</td>\n",
       "      <td>0.612903</td>\n",
       "      <td>3.032258</td>\n",
       "      <td>0.689315</td>\n",
       "      <td>0.718750</td>\n",
       "      <td>6.812500</td>\n",
       "      <td>0.748601</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2016</td>\n",
       "      <td>137</td>\n",
       "      <td>3124</td>\n",
       "      <td>89</td>\n",
       "      <td>3225</td>\n",
       "      <td>59</td>\n",
       "      <td>1</td>\n",
       "      <td>16</td>\n",
       "      <td>0.970588</td>\n",
       "      <td>24.617647</td>\n",
       "      <td>0.982974</td>\n",
       "      <td>0.709677</td>\n",
       "      <td>10.612903</td>\n",
       "      <td>0.758355</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2016</td>\n",
       "      <td>137</td>\n",
       "      <td>3177</td>\n",
       "      <td>97</td>\n",
       "      <td>3241</td>\n",
       "      <td>67</td>\n",
       "      <td>6</td>\n",
       "      <td>11</td>\n",
       "      <td>0.757576</td>\n",
       "      <td>16.272727</td>\n",
       "      <td>0.780689</td>\n",
       "      <td>0.843750</td>\n",
       "      <td>15.406250</td>\n",
       "      <td>0.869546</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Season  DayNum  W_TeamID  W_Score  L_TeamID  L_Score  W_Seed  L_Seed  \\\n",
       "0    2016     137      3107       61      3196       59      12       5   \n",
       "1    2016     137      3113       74      3308       52       2      15   \n",
       "2    2016     137      3120       68      3385       57       9       8   \n",
       "3    2016     137      3124       89      3225       59       1      16   \n",
       "4    2016     137      3177       97      3241       67       6      11   \n",
       "\n",
       "   W_WinRatio  W_AvgScoreGap  W_PointsRatio  L_WinRatio  L_AvgScoreGap  \\\n",
       "0    0.870968      18.677419       0.892371    0.733333       8.700000   \n",
       "1    0.806452       9.258065       0.828093    0.857143       8.964286   \n",
       "2    0.612903       3.032258       0.689315    0.718750       6.812500   \n",
       "3    0.970588      24.617647       0.982974    0.709677      10.612903   \n",
       "4    0.757576      16.272727       0.780689    0.843750      15.406250   \n",
       "\n",
       "   L_PointsRatio  \n",
       "0       0.766239  \n",
       "1       0.881409  \n",
       "2       0.748601  \n",
       "3       0.758355  \n",
       "4       0.869546  "
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "# merge with 538\n",
    "df = pd.merge(df, df_538, how='left',\n",
    "    left_on=['Season', 'W_TeamID'], right_on=['Season', 'TeamID']\n",
    "             ).drop('TeamID', axis=1).rename(columns={'538rating': 'W_538rating'})\n",
    "\n",
    "df = pd.merge(df, df_538, how='left', \n",
    "    left_on=['Season', 'L_TeamID'], right_on=['Season', 'TeamID']\n",
    "             ).drop('TeamID', axis=1).rename(columns={'538rating': 'L_538rating'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "def replace_win_loser(df):\n",
    "    team_a = df.copy()\n",
    "    team_b = df.copy()\n",
    "    \n",
    "    team_a_dict, team_b_dict = {}, {}\n",
    "    \n",
    "    for col in team_a.columns:\n",
    "        if col.find('W_') == 0:\n",
    "            new_col_name = str(col).replace('W_', 'A_')\n",
    "            team_a_dict[col] = new_col_name\n",
    "        if col.find('L_') == 0:\n",
    "            new_col_name = col.replace('L_', 'B_')    \n",
    "            team_a_dict[col] = new_col_name\n",
    "            \n",
    "    for col in team_b.columns:\n",
    "        if col.find('W_') == 0:\n",
    "            new_col_name = str(col).replace('W_', 'B_')\n",
    "            team_b_dict[col] = new_col_name\n",
    "        if col.find('L_') == 0:\n",
    "            new_col_name = col.replace('L_', 'A_')\n",
    "            team_b_dict[col] = new_col_name\n",
    "\n",
    "    team_a = team_a.rename(columns = team_a_dict)\n",
    "    team_b = team_b.rename(columns = team_b_dict)\n",
    "    \n",
    "    merged_df = pd.concat([team_a, team_b], axis = 0, sort = False)\n",
    "    return merged_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Season</th>\n",
       "      <th>DayNum</th>\n",
       "      <th>A_TeamID</th>\n",
       "      <th>A_Score</th>\n",
       "      <th>B_TeamID</th>\n",
       "      <th>B_Score</th>\n",
       "      <th>A_Seed</th>\n",
       "      <th>B_Seed</th>\n",
       "      <th>A_WinRatio</th>\n",
       "      <th>A_AvgScoreGap</th>\n",
       "      <th>A_PointsRatio</th>\n",
       "      <th>B_WinRatio</th>\n",
       "      <th>B_AvgScoreGap</th>\n",
       "      <th>B_PointsRatio</th>\n",
       "      <th>A_538rating</th>\n",
       "      <th>B_538rating</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2016</td>\n",
       "      <td>137</td>\n",
       "      <td>3107</td>\n",
       "      <td>61</td>\n",
       "      <td>3196</td>\n",
       "      <td>59</td>\n",
       "      <td>12</td>\n",
       "      <td>5</td>\n",
       "      <td>0.870968</td>\n",
       "      <td>18.677419</td>\n",
       "      <td>0.892371</td>\n",
       "      <td>0.733333</td>\n",
       "      <td>8.700000</td>\n",
       "      <td>0.766239</td>\n",
       "      <td>80.90</td>\n",
       "      <td>83.36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2016</td>\n",
       "      <td>137</td>\n",
       "      <td>3113</td>\n",
       "      <td>74</td>\n",
       "      <td>3308</td>\n",
       "      <td>52</td>\n",
       "      <td>2</td>\n",
       "      <td>15</td>\n",
       "      <td>0.806452</td>\n",
       "      <td>9.258065</td>\n",
       "      <td>0.828093</td>\n",
       "      <td>0.857143</td>\n",
       "      <td>8.964286</td>\n",
       "      <td>0.881409</td>\n",
       "      <td>91.32</td>\n",
       "      <td>71.42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2016</td>\n",
       "      <td>137</td>\n",
       "      <td>3120</td>\n",
       "      <td>68</td>\n",
       "      <td>3385</td>\n",
       "      <td>57</td>\n",
       "      <td>9</td>\n",
       "      <td>8</td>\n",
       "      <td>0.612903</td>\n",
       "      <td>3.032258</td>\n",
       "      <td>0.689315</td>\n",
       "      <td>0.718750</td>\n",
       "      <td>6.812500</td>\n",
       "      <td>0.748601</td>\n",
       "      <td>80.17</td>\n",
       "      <td>82.19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2016</td>\n",
       "      <td>137</td>\n",
       "      <td>3124</td>\n",
       "      <td>89</td>\n",
       "      <td>3225</td>\n",
       "      <td>59</td>\n",
       "      <td>1</td>\n",
       "      <td>16</td>\n",
       "      <td>0.970588</td>\n",
       "      <td>24.617647</td>\n",
       "      <td>0.982974</td>\n",
       "      <td>0.709677</td>\n",
       "      <td>10.612903</td>\n",
       "      <td>0.758355</td>\n",
       "      <td>101.59</td>\n",
       "      <td>72.69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2016</td>\n",
       "      <td>137</td>\n",
       "      <td>3177</td>\n",
       "      <td>97</td>\n",
       "      <td>3241</td>\n",
       "      <td>67</td>\n",
       "      <td>6</td>\n",
       "      <td>11</td>\n",
       "      <td>0.757576</td>\n",
       "      <td>16.272727</td>\n",
       "      <td>0.780689</td>\n",
       "      <td>0.843750</td>\n",
       "      <td>15.406250</td>\n",
       "      <td>0.869546</td>\n",
       "      <td>90.08</td>\n",
       "      <td>83.04</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Season  DayNum  A_TeamID  A_Score  B_TeamID  B_Score  A_Seed  B_Seed  \\\n",
       "0    2016     137      3107       61      3196       59      12       5   \n",
       "1    2016     137      3113       74      3308       52       2      15   \n",
       "2    2016     137      3120       68      3385       57       9       8   \n",
       "3    2016     137      3124       89      3225       59       1      16   \n",
       "4    2016     137      3177       97      3241       67       6      11   \n",
       "\n",
       "   A_WinRatio  A_AvgScoreGap  A_PointsRatio  B_WinRatio  B_AvgScoreGap  \\\n",
       "0    0.870968      18.677419       0.892371    0.733333       8.700000   \n",
       "1    0.806452       9.258065       0.828093    0.857143       8.964286   \n",
       "2    0.612903       3.032258       0.689315    0.718750       6.812500   \n",
       "3    0.970588      24.617647       0.982974    0.709677      10.612903   \n",
       "4    0.757576      16.272727       0.780689    0.843750      15.406250   \n",
       "\n",
       "   B_PointsRatio  A_538rating  B_538rating  \n",
       "0       0.766239        80.90        83.36  \n",
       "1       0.881409        91.32        71.42  \n",
       "2       0.748601        80.17        82.19  \n",
       "3       0.758355       101.59        72.69  \n",
       "4       0.869546        90.08        83.04  "
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = replace_win_loser(df).reset_index(drop=True)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols_to_diff = ['Seed', 'WinRatio', 'AvgScoreGap', 'PointsRatio', '538rating']\n",
    "\n",
    "for col in cols_to_diff:\n",
    "    df[col + 'Diff'] = df['A_'+col] - df['B_'+col]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create target variable\n",
    "\n",
    "df['ScoreDiff'] = df['A_Score'] - df['B_Score']\n",
    "df['A_Win'] = (df['ScoreDiff'] > 0).astype(int)\n",
    "df = df.drop(columns=['A_Score', 'B_Score'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 630 entries, 0 to 629\n",
      "Data columns (total 21 columns):\n",
      " #   Column           Non-Null Count  Dtype  \n",
      "---  ------           --------------  -----  \n",
      " 0   Season           630 non-null    int64  \n",
      " 1   DayNum           630 non-null    int64  \n",
      " 2   A_TeamID         630 non-null    int64  \n",
      " 3   B_TeamID         630 non-null    int64  \n",
      " 4   A_Seed           630 non-null    int64  \n",
      " 5   B_Seed           630 non-null    int64  \n",
      " 6   A_WinRatio       630 non-null    float64\n",
      " 7   A_AvgScoreGap    630 non-null    float64\n",
      " 8   A_PointsRatio    630 non-null    float64\n",
      " 9   B_WinRatio       630 non-null    float64\n",
      " 10  B_AvgScoreGap    630 non-null    float64\n",
      " 11  B_PointsRatio    630 non-null    float64\n",
      " 12  A_538rating      630 non-null    float64\n",
      " 13  B_538rating      630 non-null    float64\n",
      " 14  SeedDiff         630 non-null    int64  \n",
      " 15  WinRatioDiff     630 non-null    float64\n",
      " 16  AvgScoreGapDiff  630 non-null    float64\n",
      " 17  PointsRatioDiff  630 non-null    float64\n",
      " 18  538ratingDiff    630 non-null    float64\n",
      " 19  ScoreDiff        630 non-null    int64  \n",
      " 20  A_Win            630 non-null    int64  \n",
      "dtypes: float64(12), int64(9)\n",
      "memory usage: 103.5 KB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 481,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Season</th>\n",
       "      <th>DayNum</th>\n",
       "      <th>A_TeamID</th>\n",
       "      <th>B_TeamID</th>\n",
       "      <th>A_Seed</th>\n",
       "      <th>B_Seed</th>\n",
       "      <th>A_WinRatio</th>\n",
       "      <th>A_AvgScoreGap</th>\n",
       "      <th>A_PointsRatio</th>\n",
       "      <th>B_WinRatio</th>\n",
       "      <th>...</th>\n",
       "      <th>B_PointsRatio</th>\n",
       "      <th>A_538rating</th>\n",
       "      <th>B_538rating</th>\n",
       "      <th>SeedDiff</th>\n",
       "      <th>WinRatioDiff</th>\n",
       "      <th>AvgScoreGapDiff</th>\n",
       "      <th>PointsRatioDiff</th>\n",
       "      <th>538ratingDiff</th>\n",
       "      <th>ScoreDiff</th>\n",
       "      <th>A_Win</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2016</td>\n",
       "      <td>137</td>\n",
       "      <td>3107</td>\n",
       "      <td>3196</td>\n",
       "      <td>12</td>\n",
       "      <td>5</td>\n",
       "      <td>0.870968</td>\n",
       "      <td>18.677419</td>\n",
       "      <td>0.892371</td>\n",
       "      <td>0.733333</td>\n",
       "      <td>...</td>\n",
       "      <td>0.766239</td>\n",
       "      <td>80.90</td>\n",
       "      <td>83.36</td>\n",
       "      <td>7</td>\n",
       "      <td>0.137634</td>\n",
       "      <td>9.977419</td>\n",
       "      <td>0.126131</td>\n",
       "      <td>-2.46</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2016</td>\n",
       "      <td>137</td>\n",
       "      <td>3113</td>\n",
       "      <td>3308</td>\n",
       "      <td>2</td>\n",
       "      <td>15</td>\n",
       "      <td>0.806452</td>\n",
       "      <td>9.258065</td>\n",
       "      <td>0.828093</td>\n",
       "      <td>0.857143</td>\n",
       "      <td>...</td>\n",
       "      <td>0.881409</td>\n",
       "      <td>91.32</td>\n",
       "      <td>71.42</td>\n",
       "      <td>-13</td>\n",
       "      <td>-0.050691</td>\n",
       "      <td>0.293779</td>\n",
       "      <td>-0.053315</td>\n",
       "      <td>19.90</td>\n",
       "      <td>22</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2016</td>\n",
       "      <td>137</td>\n",
       "      <td>3120</td>\n",
       "      <td>3385</td>\n",
       "      <td>9</td>\n",
       "      <td>8</td>\n",
       "      <td>0.612903</td>\n",
       "      <td>3.032258</td>\n",
       "      <td>0.689315</td>\n",
       "      <td>0.718750</td>\n",
       "      <td>...</td>\n",
       "      <td>0.748601</td>\n",
       "      <td>80.17</td>\n",
       "      <td>82.19</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.105847</td>\n",
       "      <td>-3.780242</td>\n",
       "      <td>-0.059285</td>\n",
       "      <td>-2.02</td>\n",
       "      <td>11</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2016</td>\n",
       "      <td>137</td>\n",
       "      <td>3124</td>\n",
       "      <td>3225</td>\n",
       "      <td>1</td>\n",
       "      <td>16</td>\n",
       "      <td>0.970588</td>\n",
       "      <td>24.617647</td>\n",
       "      <td>0.982974</td>\n",
       "      <td>0.709677</td>\n",
       "      <td>...</td>\n",
       "      <td>0.758355</td>\n",
       "      <td>101.59</td>\n",
       "      <td>72.69</td>\n",
       "      <td>-15</td>\n",
       "      <td>0.260911</td>\n",
       "      <td>14.004744</td>\n",
       "      <td>0.224619</td>\n",
       "      <td>28.90</td>\n",
       "      <td>30</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2016</td>\n",
       "      <td>137</td>\n",
       "      <td>3177</td>\n",
       "      <td>3241</td>\n",
       "      <td>6</td>\n",
       "      <td>11</td>\n",
       "      <td>0.757576</td>\n",
       "      <td>16.272727</td>\n",
       "      <td>0.780689</td>\n",
       "      <td>0.843750</td>\n",
       "      <td>...</td>\n",
       "      <td>0.869546</td>\n",
       "      <td>90.08</td>\n",
       "      <td>83.04</td>\n",
       "      <td>-5</td>\n",
       "      <td>-0.086174</td>\n",
       "      <td>0.866477</td>\n",
       "      <td>-0.088857</td>\n",
       "      <td>7.04</td>\n",
       "      <td>30</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 21 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Season  DayNum  A_TeamID  B_TeamID  A_Seed  B_Seed  A_WinRatio  \\\n",
       "0    2016     137      3107      3196      12       5    0.870968   \n",
       "1    2016     137      3113      3308       2      15    0.806452   \n",
       "2    2016     137      3120      3385       9       8    0.612903   \n",
       "3    2016     137      3124      3225       1      16    0.970588   \n",
       "4    2016     137      3177      3241       6      11    0.757576   \n",
       "\n",
       "   A_AvgScoreGap  A_PointsRatio  B_WinRatio  ...  B_PointsRatio  A_538rating  \\\n",
       "0      18.677419       0.892371    0.733333  ...       0.766239        80.90   \n",
       "1       9.258065       0.828093    0.857143  ...       0.881409        91.32   \n",
       "2       3.032258       0.689315    0.718750  ...       0.748601        80.17   \n",
       "3      24.617647       0.982974    0.709677  ...       0.758355       101.59   \n",
       "4      16.272727       0.780689    0.843750  ...       0.869546        90.08   \n",
       "\n",
       "   B_538rating  SeedDiff  WinRatioDiff  AvgScoreGapDiff  PointsRatioDiff  \\\n",
       "0        83.36         7      0.137634         9.977419         0.126131   \n",
       "1        71.42       -13     -0.050691         0.293779        -0.053315   \n",
       "2        82.19         1     -0.105847        -3.780242        -0.059285   \n",
       "3        72.69       -15      0.260911        14.004744         0.224619   \n",
       "4        83.04        -5     -0.086174         0.866477        -0.088857   \n",
       "\n",
       "   538ratingDiff  ScoreDiff  A_Win  \n",
       "0          -2.46          2      1  \n",
       "1          19.90         22      1  \n",
       "2          -2.02         11      1  \n",
       "3          28.90         30      1  \n",
       "4           7.04         30      1  \n",
       "\n",
       "[5 rows x 21 columns]"
      ]
     },
     "execution_count": 481,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Test data (hold out data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_test = pd.read_csv(DATA_PATH + \"WSampleSubmissionStage1.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_test['Season'] = df_test['ID'].apply(lambda x: int(x.split('_')[0]))\n",
    "df_test['A_TeamID'] = df_test['ID'].apply(lambda x: int(x.split('_')[1]))\n",
    "df_test['B_TeamID'] = df_test['ID'].apply(lambda x: int(x.split('_')[2]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_test = pd.merge(\n",
    "    df_test,\n",
    "    df_seeds,\n",
    "    how='left',\n",
    "    left_on=['Season', 'A_TeamID'],\n",
    "    right_on=['Season', 'TeamID']\n",
    ").drop('TeamID', axis=1).rename(columns={'Seed': 'A_Seed'})\n",
    "\n",
    "df_test = pd.merge(\n",
    "    df_test, \n",
    "    df_seeds, \n",
    "    how='left', \n",
    "    left_on=['Season', 'B_TeamID'], \n",
    "    right_on=['Season', 'TeamID']\n",
    ").drop('TeamID', axis=1).rename(columns={'Seed': 'B_Seed'})\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_test['A_Seed'] = df_test['A_Seed'].apply(treat_seed)\n",
    "df_test['B_Seed'] = df_test['B_Seed'].apply(treat_seed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_test = pd.merge(df_test, df_features_s, how='left',\n",
    "    left_on=['Season', 'A_TeamID'], right_on=['Season', 'TeamID']\n",
    ").rename(columns={\n",
    "    'NumWins': 'A_NumWins',\n",
    "    'NumLosses': 'A_NumLosses',\n",
    "    'GapWins': 'A_GapWins',\n",
    "    'GapLosses': 'A_GapLosses',\n",
    "    'WinRatio': 'A_WinRatio',\n",
    "    'AvgScoreGap': 'A_AvgScoreGap',\n",
    "    'PointsRatio': 'A_PointsRatio',\n",
    "}).drop(columns='TeamID', axis=1)\n",
    "\n",
    "df_test = pd.merge(df_test, df_features_s, how='left',\n",
    "    left_on=['Season', 'B_TeamID'], right_on=['Season', 'TeamID']\n",
    ").rename(columns={\n",
    "    'NumWins': 'B_NumWins',\n",
    "    'NumLosses': 'B_NumLosses',\n",
    "    'GapWins': 'B_GapWins',\n",
    "    'GapLosses': 'B_GapLosses',\n",
    "    'WinRatio': 'B_WinRatio',\n",
    "    'AvgScoreGap': 'B_AvgScoreGap',\n",
    "    'PointsRatio': 'B_PointsRatio',\n",
    "}).drop(columns='TeamID', axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "# merge with 538\n",
    "df_test = pd.merge(df_test, df_538, how='left',\n",
    "    left_on=['Season', 'A_TeamID'], right_on=['Season', 'TeamID']\n",
    "             ).drop('TeamID', axis=1).rename(columns={'538rating': 'A_538rating'})\n",
    "\n",
    "df_test = pd.merge(df_test, df_538, how='left', \n",
    "    left_on=['Season', 'B_TeamID'], right_on=['Season', 'TeamID']\n",
    "             ).drop('TeamID', axis=1).rename(columns={'538rating': 'B_538rating'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols_to_diff = ['Seed', 'WinRatio', 'AvgScoreGap', 'PointsRatio', '538rating']\n",
    "\n",
    "for col in cols_to_diff:\n",
    "    df_test[col + 'Diff'] = df_test['A_'+col] - df_test['B_'+col]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Build model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import tree\n",
    "from sklearn.preprocessing import StandardScaler, MinMaxScaler\n",
    "from sklearn.metrics import log_loss\n",
    "from lightgbm import LGBMClassifier\n",
    "from xgboost import XGBClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['A_TeamID',\n",
       " 'B_TeamID',\n",
       " 'A_Seed',\n",
       " 'B_Seed',\n",
       " 'A_WinRatio',\n",
       " 'A_AvgScoreGap',\n",
       " 'A_PointsRatio',\n",
       " 'B_WinRatio',\n",
       " 'B_AvgScoreGap',\n",
       " 'B_PointsRatio',\n",
       " 'A_538rating',\n",
       " 'B_538rating',\n",
       " 'SeedDiff',\n",
       " 'WinRatioDiff',\n",
       " 'AvgScoreGapDiff',\n",
       " 'PointsRatioDiff',\n",
       " '538ratingDiff']"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "target_feature = 'A_Win'\n",
    "cols = ['ScoreDiff', 'Season', 'DayNum', 'A_Win']\n",
    "features = [col for col in df.columns if col not in cols]\n",
    "features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "features = ['A_TeamID',\n",
    " 'B_TeamID',\n",
    " 'A_Seed',\n",
    " 'B_Seed',\n",
    " 'A_WinRatio',\n",
    " 'A_AvgScoreGap',\n",
    " #'A_PointsRatio',\n",
    " 'B_WinRatio',\n",
    " 'B_AvgScoreGap',\n",
    " #'B_PointsRatio',\n",
    " 'A_538rating',\n",
    " 'B_538rating',\n",
    " 'SeedDiff',\n",
    " 'WinRatioDiff',\n",
    " 'AvgScoreGapDiff',\n",
    " #'PointsRatioDiff',\n",
    " '538ratingDiff']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "def xgb_model(train_df, test_df, features, target_feature):\n",
    "    print(features)\n",
    "    \n",
    "    i=1\n",
    "    cvs_train = []\n",
    "    cvs_val = []\n",
    "    pred_test = np.zeros(len(test_df))\n",
    "    preds_test = []\n",
    "    pred_val_df = pd.DataFrame(np.zeros(len(train_df)),columns=['pred_val'])\n",
    "    importancias=pd.DataFrame()\n",
    "    importancias['variable']=features\n",
    "    \n",
    "    seasons = df['Season'].unique()\n",
    "    for s in seasons[1:]:\n",
    "        print(f'\\nValidating on season {s}')\n",
    "        X_train = train_df[train_df['Season'] < s][features].reset_index(drop = True).copy()\n",
    "        X_val = train_df[train_df['Season'] == s][features].reset_index(drop = True).copy()\n",
    "        val_index = train_df[train_df['Season'] == s].index\n",
    "    \n",
    "        y_train = train_df[train_df['Season'] < s][target_feature].reset_index(drop = True).copy()\n",
    "        y_val = train_df[train_df['Season'] == s][target_feature].reset_index(drop = True).copy()\n",
    "    \n",
    "        X_test = test_df[features].copy()\n",
    "        \n",
    "        # scale\n",
    "        scaler = MinMaxScaler()\n",
    "        X_train = scaler.fit_transform(X_train)\n",
    "        X_val = scaler.transform(X_val)\n",
    "        X_test = scaler.transform(X_test)\n",
    "        \n",
    "        # model fit\n",
    "        params = {'learning_rate':0.5,\n",
    "                  'max_depth':4,\n",
    "                  'grow_policy':'lossguide',\n",
    "                  #'gamma':0.5\n",
    "        }\n",
    "\n",
    "        model = XGBClassifier(**params, n_estimators=1024, random_state = 85, use_label_encoder=False, eval_metric='logloss')\n",
    "        model.fit(X_train, y_train, eval_set = [(X_val, y_val)], verbose = 0, early_stopping_rounds = 128)\n",
    "        importancias['gain_'+ str(i)] = model.feature_importances_\n",
    "        i=i+1\n",
    "        \n",
    "        \n",
    "        # teain\n",
    "        pred_train = model.predict_proba(X_train)[:, 1]\n",
    "        loss_train = log_loss(y_train, pred_train)\n",
    "        \n",
    "        # val\n",
    "        pred_val = model.predict_proba(X_val)[:, 1]\n",
    "        pred_val_df.iloc[val_index,0] = pd.Series(pred_val).values\n",
    "        loss_val = log_loss(y_val, pred_val)\n",
    "        \n",
    "        cvs_train.append(loss_train)\n",
    "        cvs_val.append(loss_val)\n",
    "        print(f'\\t -> Scored {loss_val:.4f}')\n",
    "        \n",
    "        # test\n",
    "        pred_test = model.predict_proba(X_test)[:, 1]\n",
    "        preds_test.append(pred_test)\n",
    "\n",
    "      \n",
    "    importancias[\"gain_avg\"]=importancias[[\"gain_1\",\"gain_2\",\"gain_3\",\"gain_4\"]].mean(axis=1)\n",
    "    importancias=importancias.sort_values(\"gain_avg\",ascending=False).reset_index(drop=True)\n",
    "\n",
    "    predictions = np.mean(preds_test, 0)    \n",
    "    result = {}\n",
    "    result['cvs_train'] = cvs_train\n",
    "    result['cvs_val'] = cvs_val\n",
    "    result['importancias']=importancias\n",
    "    result['pred_val'] = pred_val_df\n",
    "    result['predictions'] = predictions\n",
    "\n",
    "    \n",
    "    return result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['A_TeamID', 'B_TeamID', 'A_Seed', 'B_Seed', 'A_WinRatio', 'A_AvgScoreGap', 'B_WinRatio', 'B_AvgScoreGap', 'A_538rating', 'B_538rating', 'SeedDiff', 'WinRatioDiff', 'AvgScoreGapDiff', '538ratingDiff']\n",
      "\n",
      "Validating on season 2017\n",
      "\t -> Scored 0.4754\n",
      "\n",
      "Validating on season 2018\n",
      "\t -> Scored 0.5332\n",
      "\n",
      "Validating on season 2019\n",
      "\t -> Scored 0.4251\n",
      "\n",
      "Validating on season 2021\n",
      "\t -> Scored 0.4095\n"
     ]
    }
   ],
   "source": [
    "result1= xgb_model(df, df_test, features, target_feature)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "training loss 0.2555996140747863\n",
      "validation loss 0.4608042297777646\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<BarContainer object of 14 artists>"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#print(result1['cvs_train'])\n",
    "print('training loss', sum(result1['cvs_train']) / len(result1['cvs_train']))\n",
    "#print( result1['cvs_val'])\n",
    "print('validation loss', sum(result1['cvs_val']) / len(result1['cvs_val']))\n",
    "data = result1['importancias'].sort_values('gain_avg', ascending=False)\n",
    "plt.barh('variable','gain_avg',data=data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['538ratingDiff', 'SeedDiff']"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# feature selection\n",
    "features_selected=data['variable'].tolist()[0:2]\n",
    "features_selected"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['538ratingDiff', 'SeedDiff']\n",
      "\n",
      "Validating on season 2017\n",
      "\t -> Scored 0.4533\n",
      "\n",
      "Validating on season 2018\n",
      "\t -> Scored 0.5140\n",
      "\n",
      "Validating on season 2019\n",
      "\t -> Scored 0.4046\n",
      "\n",
      "Validating on season 2021\n",
      "\t -> Scored 0.4106\n",
      "training loss 0.3065100818076023\n",
      "validation loss 0.44563807764401037\n"
     ]
    }
   ],
   "source": [
    "result2= xgb_model(df, df_test, features_selected, target_feature)\n",
    "#print(result2['cvs_train'])\n",
    "print('training loss', sum(result2['cvs_train']) / len(result2['cvs_train']))\n",
    "#print( result1['cvs_val'])\n",
    "print('validation loss', sum(result2['cvs_val']) / len(result2['cvs_val']))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Submission"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 582,
   "metadata": {},
   "outputs": [],
   "source": [
    "sub = df_test[['ID', 'Pred']].copy()\n",
    "sub['Pred'] = result2['predictions']\n",
    "sub.to_csv('submission.csv', index = False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Evaluate result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_result = pd.concat([df, result1['pred_val']], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Season</th>\n",
       "      <th>DayNum</th>\n",
       "      <th>A_TeamID</th>\n",
       "      <th>B_TeamID</th>\n",
       "      <th>A_Seed</th>\n",
       "      <th>B_Seed</th>\n",
       "      <th>A_WinRatio</th>\n",
       "      <th>A_AvgScoreGap</th>\n",
       "      <th>A_PointsRatio</th>\n",
       "      <th>B_WinRatio</th>\n",
       "      <th>B_AvgScoreGap</th>\n",
       "      <th>B_PointsRatio</th>\n",
       "      <th>A_538rating</th>\n",
       "      <th>B_538rating</th>\n",
       "      <th>SeedDiff</th>\n",
       "      <th>WinRatioDiff</th>\n",
       "      <th>AvgScoreGapDiff</th>\n",
       "      <th>PointsRatioDiff</th>\n",
       "      <th>538ratingDiff</th>\n",
       "      <th>ScoreDiff</th>\n",
       "      <th>A_Win</th>\n",
       "      <th>pred_val</th>\n",
       "      <th>Win_pred</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>63</th>\n",
       "      <td>2017</td>\n",
       "      <td>137</td>\n",
       "      <td>3113</td>\n",
       "      <td>3277</td>\n",
       "      <td>8</td>\n",
       "      <td>9</td>\n",
       "      <td>0.612903</td>\n",
       "      <td>6.161290</td>\n",
       "      <td>0.663276</td>\n",
       "      <td>0.656250</td>\n",
       "      <td>7.031250</td>\n",
       "      <td>0.686017</td>\n",
       "      <td>87.68</td>\n",
       "      <td>84.28</td>\n",
       "      <td>-1</td>\n",
       "      <td>-0.043347</td>\n",
       "      <td>-0.869960</td>\n",
       "      <td>-0.022741</td>\n",
       "      <td>3.40</td>\n",
       "      <td>12</td>\n",
       "      <td>1</td>\n",
       "      <td>0.760511</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64</th>\n",
       "      <td>2017</td>\n",
       "      <td>137</td>\n",
       "      <td>3166</td>\n",
       "      <td>3405</td>\n",
       "      <td>7</td>\n",
       "      <td>10</td>\n",
       "      <td>0.766667</td>\n",
       "      <td>8.066667</td>\n",
       "      <td>0.793835</td>\n",
       "      <td>0.750000</td>\n",
       "      <td>8.281250</td>\n",
       "      <td>0.782222</td>\n",
       "      <td>79.96</td>\n",
       "      <td>76.26</td>\n",
       "      <td>-3</td>\n",
       "      <td>0.016667</td>\n",
       "      <td>-0.214583</td>\n",
       "      <td>0.011613</td>\n",
       "      <td>3.70</td>\n",
       "      <td>27</td>\n",
       "      <td>1</td>\n",
       "      <td>0.190477</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65</th>\n",
       "      <td>2017</td>\n",
       "      <td>137</td>\n",
       "      <td>3177</td>\n",
       "      <td>3320</td>\n",
       "      <td>7</td>\n",
       "      <td>10</td>\n",
       "      <td>0.787879</td>\n",
       "      <td>13.727273</td>\n",
       "      <td>0.807505</td>\n",
       "      <td>0.741935</td>\n",
       "      <td>9.387097</td>\n",
       "      <td>0.768971</td>\n",
       "      <td>87.94</td>\n",
       "      <td>77.10</td>\n",
       "      <td>-3</td>\n",
       "      <td>0.045943</td>\n",
       "      <td>4.340176</td>\n",
       "      <td>0.038534</td>\n",
       "      <td>10.84</td>\n",
       "      <td>21</td>\n",
       "      <td>1</td>\n",
       "      <td>0.946857</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>66</th>\n",
       "      <td>2017</td>\n",
       "      <td>137</td>\n",
       "      <td>3199</td>\n",
       "      <td>3442</td>\n",
       "      <td>3</td>\n",
       "      <td>14</td>\n",
       "      <td>0.806452</td>\n",
       "      <td>21.806452</td>\n",
       "      <td>0.841700</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>9.800000</td>\n",
       "      <td>0.838242</td>\n",
       "      <td>94.36</td>\n",
       "      <td>71.83</td>\n",
       "      <td>-11</td>\n",
       "      <td>0.006452</td>\n",
       "      <td>12.006452</td>\n",
       "      <td>0.003459</td>\n",
       "      <td>22.53</td>\n",
       "      <td>21</td>\n",
       "      <td>1</td>\n",
       "      <td>0.902640</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>67</th>\n",
       "      <td>2017</td>\n",
       "      <td>137</td>\n",
       "      <td>3246</td>\n",
       "      <td>3125</td>\n",
       "      <td>4</td>\n",
       "      <td>13</td>\n",
       "      <td>0.677419</td>\n",
       "      <td>8.193548</td>\n",
       "      <td>0.686988</td>\n",
       "      <td>0.838710</td>\n",
       "      <td>12.096774</td>\n",
       "      <td>0.873491</td>\n",
       "      <td>89.12</td>\n",
       "      <td>73.32</td>\n",
       "      <td>-9</td>\n",
       "      <td>-0.161290</td>\n",
       "      <td>-3.903226</td>\n",
       "      <td>-0.186503</td>\n",
       "      <td>15.80</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0.969180</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\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>\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>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>625</th>\n",
       "      <td>2021</td>\n",
       "      <td>148</td>\n",
       "      <td>3400</td>\n",
       "      <td>3376</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>7.074074</td>\n",
       "      <td>0.723754</td>\n",
       "      <td>0.846154</td>\n",
       "      <td>17.500000</td>\n",
       "      <td>0.885328</td>\n",
       "      <td>86.82</td>\n",
       "      <td>101.83</td>\n",
       "      <td>5</td>\n",
       "      <td>-0.179487</td>\n",
       "      <td>-10.425926</td>\n",
       "      <td>-0.161574</td>\n",
       "      <td>-15.01</td>\n",
       "      <td>-28</td>\n",
       "      <td>0</td>\n",
       "      <td>0.055329</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>626</th>\n",
       "      <td>2021</td>\n",
       "      <td>148</td>\n",
       "      <td>3257</td>\n",
       "      <td>3390</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0.884615</td>\n",
       "      <td>16.884615</td>\n",
       "      <td>0.914425</td>\n",
       "      <td>0.925926</td>\n",
       "      <td>25.703704</td>\n",
       "      <td>0.934844</td>\n",
       "      <td>96.01</td>\n",
       "      <td>106.54</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.041311</td>\n",
       "      <td>-8.819088</td>\n",
       "      <td>-0.020419</td>\n",
       "      <td>-10.53</td>\n",
       "      <td>-15</td>\n",
       "      <td>0</td>\n",
       "      <td>0.342103</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>627</th>\n",
       "      <td>2021</td>\n",
       "      <td>151</td>\n",
       "      <td>3163</td>\n",
       "      <td>3112</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0.960000</td>\n",
       "      <td>31.520000</td>\n",
       "      <td>0.957582</td>\n",
       "      <td>0.761905</td>\n",
       "      <td>10.523810</td>\n",
       "      <td>0.796853</td>\n",
       "      <td>104.27</td>\n",
       "      <td>93.87</td>\n",
       "      <td>-2</td>\n",
       "      <td>0.198095</td>\n",
       "      <td>20.996190</td>\n",
       "      <td>0.160729</td>\n",
       "      <td>10.40</td>\n",
       "      <td>-10</td>\n",
       "      <td>0</td>\n",
       "      <td>0.680826</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>628</th>\n",
       "      <td>2021</td>\n",
       "      <td>151</td>\n",
       "      <td>3376</td>\n",
       "      <td>3390</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.846154</td>\n",
       "      <td>17.500000</td>\n",
       "      <td>0.885328</td>\n",
       "      <td>0.925926</td>\n",
       "      <td>25.703704</td>\n",
       "      <td>0.934844</td>\n",
       "      <td>101.83</td>\n",
       "      <td>106.54</td>\n",
       "      <td>0</td>\n",
       "      <td>-0.079772</td>\n",
       "      <td>-8.203704</td>\n",
       "      <td>-0.049516</td>\n",
       "      <td>-4.71</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.454524</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>629</th>\n",
       "      <td>2021</td>\n",
       "      <td>153</td>\n",
       "      <td>3112</td>\n",
       "      <td>3390</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0.761905</td>\n",
       "      <td>10.523810</td>\n",
       "      <td>0.796853</td>\n",
       "      <td>0.925926</td>\n",
       "      <td>25.703704</td>\n",
       "      <td>0.934844</td>\n",
       "      <td>93.87</td>\n",
       "      <td>106.54</td>\n",
       "      <td>2</td>\n",
       "      <td>-0.164021</td>\n",
       "      <td>-15.179894</td>\n",
       "      <td>-0.137992</td>\n",
       "      <td>-12.67</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.345556</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>504 rows × 23 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Season  DayNum  A_TeamID  B_TeamID  A_Seed  B_Seed  A_WinRatio  \\\n",
       "63     2017     137      3113      3277       8       9    0.612903   \n",
       "64     2017     137      3166      3405       7      10    0.766667   \n",
       "65     2017     137      3177      3320       7      10    0.787879   \n",
       "66     2017     137      3199      3442       3      14    0.806452   \n",
       "67     2017     137      3246      3125       4      13    0.677419   \n",
       "..      ...     ...       ...       ...     ...     ...         ...   \n",
       "625    2021     148      3400      3376       6       1    0.666667   \n",
       "626    2021     148      3257      3390       2       1    0.884615   \n",
       "627    2021     151      3163      3112       1       3    0.960000   \n",
       "628    2021     151      3376      3390       1       1    0.846154   \n",
       "629    2021     153      3112      3390       3       1    0.761905   \n",
       "\n",
       "     A_AvgScoreGap  A_PointsRatio  B_WinRatio  B_AvgScoreGap  B_PointsRatio  \\\n",
       "63        6.161290       0.663276    0.656250       7.031250       0.686017   \n",
       "64        8.066667       0.793835    0.750000       8.281250       0.782222   \n",
       "65       13.727273       0.807505    0.741935       9.387097       0.768971   \n",
       "66       21.806452       0.841700    0.800000       9.800000       0.838242   \n",
       "67        8.193548       0.686988    0.838710      12.096774       0.873491   \n",
       "..             ...            ...         ...            ...            ...   \n",
       "625       7.074074       0.723754    0.846154      17.500000       0.885328   \n",
       "626      16.884615       0.914425    0.925926      25.703704       0.934844   \n",
       "627      31.520000       0.957582    0.761905      10.523810       0.796853   \n",
       "628      17.500000       0.885328    0.925926      25.703704       0.934844   \n",
       "629      10.523810       0.796853    0.925926      25.703704       0.934844   \n",
       "\n",
       "     A_538rating  B_538rating  SeedDiff  WinRatioDiff  AvgScoreGapDiff  \\\n",
       "63         87.68        84.28        -1     -0.043347        -0.869960   \n",
       "64         79.96        76.26        -3      0.016667        -0.214583   \n",
       "65         87.94        77.10        -3      0.045943         4.340176   \n",
       "66         94.36        71.83       -11      0.006452        12.006452   \n",
       "67         89.12        73.32        -9     -0.161290        -3.903226   \n",
       "..           ...          ...       ...           ...              ...   \n",
       "625        86.82       101.83         5     -0.179487       -10.425926   \n",
       "626        96.01       106.54         1     -0.041311        -8.819088   \n",
       "627       104.27        93.87        -2      0.198095        20.996190   \n",
       "628       101.83       106.54         0     -0.079772        -8.203704   \n",
       "629        93.87       106.54         2     -0.164021       -15.179894   \n",
       "\n",
       "     PointsRatioDiff  538ratingDiff  ScoreDiff  A_Win  pred_val  Win_pred  \n",
       "63         -0.022741           3.40         12      1  0.760511         1  \n",
       "64          0.011613           3.70         27      1  0.190477         0  \n",
       "65          0.038534          10.84         21      1  0.946857         1  \n",
       "66          0.003459          22.53         21      1  0.902640         1  \n",
       "67         -0.186503          15.80          3      1  0.969180         1  \n",
       "..               ...            ...        ...    ...       ...       ...  \n",
       "625        -0.161574         -15.01        -28      0  0.055329         0  \n",
       "626        -0.020419         -10.53        -15      0  0.342103         0  \n",
       "627         0.160729          10.40        -10      0  0.680826         1  \n",
       "628        -0.049516          -4.71         -1      0  0.454524         0  \n",
       "629        -0.137992         -12.67         -1      0  0.345556         0  \n",
       "\n",
       "[504 rows x 23 columns]"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_result = df_result[df_result['pred_val'] > 0]\n",
    "\n",
    "df_result['Win_pred'] = df_result['pred_val'].apply(lambda x: 1 if x >0.5 else 0)\n",
    "df_result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.7876984126984127\n",
      "[[198  54]\n",
      " [ 53 199]]\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.79      0.79      0.79       252\n",
      "           1       0.79      0.79      0.79       252\n",
      "\n",
      "    accuracy                           0.79       504\n",
      "   macro avg       0.79      0.79      0.79       504\n",
      "weighted avg       0.79      0.79      0.79       504\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import accuracy_score, confusion_matrix, classification_report\n",
    "\n",
    "print(accuracy_score(df_result['A_Win'], df_result['Win_pred']))\n",
    "print(confusion_matrix(df_result['A_Win'], df_result['Win_pred']))\n",
    "print(classification_report(df_result['A_Win'], df_result['Win_pred']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x7fc464e59100>"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter('pred_val','A_Win', s=1, data=df_result)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.FacetGrid at 0x7fc4649f5bb0>"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.displot(df_result['pred_val'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import math\n",
    "  \n",
    "x = np.linspace(-10, 10, 100)\n",
    "z = 1/(1 + np.exp(-x))\n",
    "  \n",
    "plt.plot(x, z)\n",
    "plt.xlabel(\"x\")\n",
    "plt.ylabel(\"Sigmoid(X)\")\n",
    "  \n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.8.12"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {
    "height": "calc(100% - 180px)",
    "left": "10px",
    "top": "150px",
    "width": "165px"
   },
   "toc_section_display": true,
   "toc_window_display": true
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}