weight_viz.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import os\n",
    "import json\n",
    "from pyhocon import ConfigFactory as hocon\n",
    "import random\n",
    "import numpy as np\n",
    "from tuna import *\n",
    "from collections import defaultdict\n",
    "from glob import glob\n",
    "from sgd_lsl import plot_matrix\n",
    "from matplotlib import pyplot as plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "with open('test_baseline/params.json', 'r') as infile:\n",
    "    weights = defaultdict(float, json.load(infile))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "words, attrs = zip(*[f.split('|') for f in weights.keys()])\n",
    "words = sorted(set(words))\n",
    "attrs = sorted(set(attrs))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "158 24\n"
     ]
    }
   ],
   "source": [
    "print len(words), len(attrs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def l2_norm_sq(pair):\n",
    "    s, vec = pair\n",
    "    return np.sum(vec ** 2)\n",
    "\n",
    "word_vecs = sorted([(w, np.array([weights['%s|%s' % (w, a)] for a in attrs]))\n",
    "                    for w in words], key=l2_norm_sq, reverse=True)\n",
    "words_sorted, word_vecs = zip(*word_vecs)\n",
    "attr_vecs = sorted([(a, np.array([weights['%s|%s' % (w, a)] for w in words_sorted]))\n",
    "                    for a in attrs], key=l2_norm_sq, reverse=True)\n",
    "attrs_sorted, attr_vecs = zip(*attr_vecs)\n",
    "\n",
    "attrs_mat = np.array(attr_vecs)\n",
    "words_mat = np.array(word_vecs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAvgAABHvCAYAAABGxzeTAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nyP8HHXUH3CgAXKEAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fdfed40b750>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot.figure(figsize=(15, 80))\n",
    "plot_matrix(attrs_mat.T, xlabels=attrs_sorted, ylabels=words_sorted)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "with open('runs/39/predictions.0.jsons', 'r') as infile:\n",
    "    preds = [json.loads(line) for line in infile]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.702331 +/- 0.248448\n"
     ]
    }
   ],
   "source": [
    "import metrics\n",
    "md = np.array([metrics.multiset_dice(y, prediction)\n",
    "               for (i, x, y, prediction) in preds])\n",
    "print '%f +/- %f' % (md.mean(), md.std())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 189,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def load_json(filename):\n",
    "    false = False\n",
    "    true = True\n",
    "    null = None\n",
    "    with open(filename, 'r') as infile:\n",
    "        return defaultdict(lambda: None, eval(infile.read()))\n",
    "\n",
    "def extract_accuracy(stdout_file):\n",
    "    accuracy = None\n",
    "    acc_std = None\n",
    "    error = float('nan')\n",
    "    try:\n",
    "        with open(stdout_file, 'r') as infile:\n",
    "            for line in infile:\n",
    "                if line.startswith('Error:'):\n",
    "                    error = float(line.split()[1])\n",
    "                \n",
    "                if line.startswith('accuracy mean:'):\n",
    "                    accuracy = float(line.split()[2])\n",
    "                elif line.startswith('accuracy std:'):\n",
    "                    acc_std = float(line.split()[2])\n",
    "                elif '--' in line:\n",
    "                    tokens = line.split(':')[-1].split()\n",
    "                    acc_str = tokens[0]\n",
    "                    try:\n",
    "                        ci_lower, ci_upper = tokens[1][1:-1].split('--')\n",
    "                    except IndexError:\n",
    "                        print line\n",
    "                        ci_lower = ci_upper = 0.0\n",
    "                    return float(acc_str), (float(ci_lower), float(ci_upper)), error\n",
    "                elif line.startswith('Mean accuracy:'):\n",
    "                    tokens = line.split()\n",
    "                    acc_str = tokens[2]\n",
    "                    try:\n",
    "                        std_str = tokens[4]\n",
    "                        if std_str.endswith(')'): std_str = std_str[:-1]\n",
    "                    except IndexError:\n",
    "                        print line\n",
    "                        std_str = '0.0'\n",
    "                    return float(acc_str), float(std_str), error\n",
    "    except IOError, e:\n",
    "        #print '%s: %s' % (stdout_file, e)\n",
    "        return None, None, None\n",
    "    #if accuracy is None:\n",
    "        #print '%s: accuracy mean not found' % stdout_file\n",
    "    return accuracy, acc_std, error\n",
    "\n",
    "def summary(config):\n",
    "    cpbaseline = ('trials' in config)\n",
    "    mat_size = ('(%s%s)' % (config['mat_size'],\n",
    "                            'noamb' if config['ambiguities'] is False else '')\n",
    "                if config['artificial'] or not config['data_dir'] else '')\n",
    "    \n",
    "    dataset = '%s%s' % (config['data_dir'],\n",
    "                        '(generation)' if config['generation'] else '')\n",
    "    \n",
    "    return dataset, '%s%s%s%s%s%s%s%s%s eta=%s l2=%s feats=%s' % (\n",
    "        dataset,\n",
    "        ' literal' if config['literal'] else '',\n",
    "        ' cpbaseline' if cpbaseline else '',\n",
    "        ' sampled' if config['samples_x'] else '',\n",
    "        ' relevant' if config['only_relevant_alts'] else '',\n",
    "        ' null' if config['null_message'] else '',\n",
    "        ' artificial' if config['artificial'] else '',\n",
    "        ' loc=%s' % config['filter_loc'] if config['filter_loc'] is not None else '',\n",
    "        mat_size,\n",
    "        config['sgd_eta'],\n",
    "        config['l2_coeff'],\n",
    "        str(config['features']) if config['features'] else 'default',\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.890 ~0.000           18.035 singular/furniture eta=0.1 l2=None feats=default runs/11\n",
      "0.880 ~0.000           87.944 singular/furniture eta=0.1 l2=0.05 feats=default runs/12\n",
      "0.890 ~0.000           31.509 singular/furniture eta=0.1 l2=0.01 feats=default runs/13\n",
      "0.610 ~0.070           29.842 singular/furniture literal eta=0.1 l2=0.01 feats=default runs/141\n",
      "0.240 ~0.030          774.511 singular/furniture(generation) eta=0.1 l2=0.01 feats=default runs/142\n",
      "0.110 ~0.020          -15.448 plural/furniture(generation) eta=0.1 l2=0.01 feats=default runs/144\n",
      "0.820 ~0.050          288.864 singular/furniture literal eta=0.1 l2=0.01 feats=default runs/146\n",
      "0.410 ~0.070          166.939 singular/people literal eta=0.1 l2=0.01 feats=default runs/147\n",
      "0.520 ~0.030          490.835 plural/furniture literal eta=0.1 l2=0.01 feats=default runs/148\n",
      "0.290 ~0.050          211.388 plural/people literal eta=0.1 l2=0.01 feats=default runs/149\n",
      "0.240 ~0.030          553.425 singular/furniture(generation) eta=0.01 l2=0.01 feats=default runs/150\n",
      "0.240 ~0.020          795.621 singular/furniture(generation) eta=1.0 l2=0.01 feats=default runs/151\n",
      "0.260 ~0.040          899.609 singular/furniture(generation) eta=0.1 l2=0.0 feats=default runs/152\n",
      "0.240 ~0.030          484.054 singular/furniture(generation) eta=0.1 l2=0.1 feats=default runs/153\n",
      "0.230 ~0.030          361.430 singular/furniture(generation) eta=0.1 l2=1.0 feats=default runs/154\n",
      "0.910 ~0.030           12.355 singular/furniture eta=0.1 l2=0.01 feats=default runs/155\n",
      "0.750 ~0.040            4.349 singular/people eta=0.1 l2=0.01 feats=default runs/156\n",
      "0.910 ~0.020           30.194 singular/furniture eta=0.01 l2=0.01 feats=default runs/159\n",
      "0.910 ~0.030           23.001 singular/furniture eta=1.0 l2=0.01 feats=default runs/160\n",
      "0.770 ~0.040            0.000 singular/furniture eta=0.1 l2=0.0 feats=default runs/161\n",
      "0.910 ~0.020           88.992 singular/furniture eta=0.1 l2=0.1 feats=default runs/162\n",
      "0.880 ~0.030          293.519 singular/furniture eta=0.1 l2=1.0 feats=default runs/163\n",
      "0.110 ~0.020          343.322 singular/furniture(generation) literal eta=0.1 l2=0.01 feats=default runs/164\n",
      "0.030 ~0.010         -298.356 plural/furniture(generation) literal eta=0.1 l2=0.01 feats=default runs/166\n",
      "0.120 ~0.040          336.715 singular/furniture(generation) literal eta=0.01 l2=0.01 feats=default runs/168\n",
      "0.070 ~0.020          401.978 singular/furniture(generation) literal eta=1.0 l2=0.01 feats=default runs/169\n",
      "0.080 ~0.040          343.072 singular/furniture(generation) literal eta=0.1 l2=0.0 feats=default runs/170\n",
      "0.070 ~0.010          342.598 singular/furniture(generation) literal eta=0.1 l2=0.1 feats=default runs/171\n",
      "0.070 ~0.040          341.831 singular/furniture(generation) literal eta=0.1 l2=1.0 feats=default runs/172\n",
      "0.255 (0.203--0.306)      nan singular/furniture(generation) eta=0.01 l2=0.01 feats=default runs/173\n",
      "0.108 (0.089--0.126)      nan plural/furniture(generation) eta=0.01 l2=0.01 feats=default runs/175\n",
      "0.912 (0.873--0.951)      nan singular/furniture eta=0.01 l2=0.01 feats=default runs/177\n",
      "0.567 (0.533--0.600)      nan singular/people eta=0.01 l2=0.01 feats=default runs/178\n",
      "0.890 ~0.000           31.509 singular/furniture eta=0.1 l2=0.01 feats=default runs/18\n",
      "0.248 (0.208--0.288)  563.663 singular/furniture(generation) sampled eta=0.01 l2=0.01 feats=default runs/181\n",
      "0.089 (0.041--0.137)  590.339 singular/people(generation) sampled eta=0.01 l2=0.01 feats=default runs/182\n",
      "0.096 (0.075--0.118)  505.460 plural/furniture(generation) sampled eta=0.01 l2=0.01 feats=default runs/183\n",
      "0.910 (0.874--0.945)   40.766 singular/furniture sampled eta=0.01 l2=0.01 feats=default runs/185\n",
      "0.578 (0.519--0.636)   41.502 singular/people sampled eta=0.01 l2=0.01 feats=default runs/186\n",
      "0.523 (0.437--0.609)  365.813 plural/furniture sampled eta=0.01 l2=0.01 feats=default runs/187\n",
      "0.389 (0.331--0.446)  177.624 plural/people sampled eta=0.01 l2=0.01 feats=default runs/188\n",
      "0.510 (0.370--0.660)      nan None(None) eta=None l2=None feats=default runs/190\n",
      "0.600 (0.510--0.690)    2.000 singular/furniture artificial(None) eta=0.1 l2=0.0 feats=default runs/191\n",
      "0.860 (0.860--0.860)    8.112 singular/furniture artificial(None) eta=0.1 l2=0.0 feats=default runs/193\n",
      "0.510 (0.370--0.660)      nan None(None) eta=None l2=None feats=default runs/194\n",
      "0.720 (0.640--0.800)      nan None(None) eta=None l2=None feats=default runs/195\n",
      "0.720 (0.640--0.800)      nan None(2) eta=None l2=None feats=default runs/200\n",
      "0.500 (0.490--0.510)      nan None(3) eta=None l2=None feats=default runs/201\n",
      "0.350 (0.350--0.350)      nan None(4) eta=None l2=None feats=default runs/202\n",
      "0.670 (0.610--0.730)      nan None(2noamb) eta=None l2=None feats=default runs/203\n",
      "0.480 (0.470--0.490)      nan None(3noamb) eta=None l2=None feats=default runs/204\n",
      "0.330 (0.330--0.330)      nan None(4noamb) eta=None l2=None feats=default runs/205\n",
      "0.860 (0.860--0.860)    8.112 singular/furniture artificial(2) eta=0.1 l2=0.0 feats=default runs/206\n",
      "0.830 (0.830--0.830)    8.105 singular/furniture artificial(2noamb) eta=0.1 l2=0.0 feats=default runs/209\n",
      "0.860 (0.860--0.860)    8.115 singular/furniture sampled artificial(2) eta=0.1 l2=0.0 feats=default runs/212\n",
      "0.830 (0.830--0.830)    8.123 singular/furniture sampled artificial(2noamb) eta=0.1 l2=0.0 feats=default runs/215\n",
      "0.590 (0.590--0.600)  643.613 singular/furniture sampled artificial(3noamb) eta=0.1 l2=0.0 feats=default runs/216\n",
      "0.240 (0.200--0.270)  579.203 singular/furniture(generation) sampled eta=0.01 l2=0.01 feats=default runs/218\n",
      "0.100 (0.070--0.130)  561.346 plural/furniture(generation) sampled eta=0.01 l2=0.01 feats=default runs/220\n",
      "0.900 (0.860--0.940)   41.728 singular/furniture sampled eta=0.01 l2=0.01 feats=default runs/222\n",
      "0.550 (0.460--0.640)   47.304 singular/people sampled eta=0.01 l2=0.01 feats=default runs/223\n",
      "0.440 (0.370--0.510)  932.049 plural/furniture sampled eta=0.01 l2=0.01 feats=default runs/224\n",
      "0.350 (0.290--0.400)  226.258 plural/people sampled eta=0.01 l2=0.01 feats=default runs/225\n",
      "0.240 (0.190--0.280)  599.022 singular/furniture(generation) sampled null eta=0.01 l2=0.01 feats=default runs/226\n",
      "0.100 (0.070--0.120)  573.640 plural/furniture(generation) sampled null eta=0.01 l2=0.01 feats=default runs/228\n",
      "0.910 (0.870--0.950)   33.627 singular/furniture sampled null eta=0.01 l2=0.01 feats=default runs/230\n",
      "0.650 (0.590--0.700)   66.199 singular/people sampled null eta=0.01 l2=0.01 feats=default runs/231\n",
      "0.470 (0.410--0.540)  902.129 plural/furniture sampled null eta=0.01 l2=0.01 feats=default runs/232\n",
      "0.470 (0.400--0.540)  222.378 plural/people sampled null eta=0.01 l2=0.01 feats=default runs/233\n",
      "0.195 (0.153--0.238)  591.519 singular/furniture(generation) sampled relevant null eta=0.01 l2=0.01 feats=default runs/235\n",
      "0.022 (0.008--0.036) 1313.995 singular/people(generation) sampled relevant null eta=0.01 l2=0.01 feats=default runs/236\n",
      "0.069 (0.050--0.088)  744.254 plural/furniture(generation) sampled relevant null eta=0.01 l2=0.01 feats=default runs/237\n",
      "0.003 (-0.001--0.007) 1466.339 plural/people(generation) sampled relevant null eta=0.01 l2=0.01 feats=default runs/238\n",
      "0.771 (0.717--0.826)  126.176 singular/furniture sampled relevant null eta=0.01 l2=0.01 feats=default runs/239\n",
      "0.475 (0.398--0.552)  160.754 singular/people sampled relevant null eta=0.01 l2=0.01 feats=default runs/240\n",
      "0.153 (0.114--0.191) 1035.163 plural/furniture sampled relevant null eta=0.01 l2=0.01 feats=default runs/241\n",
      "0.350 (0.284--0.416)  321.908 plural/people sampled relevant null eta=0.01 l2=0.01 feats=default runs/242\n",
      "0.017 (0.009--0.024)  336.000 singular/furniture(generation) sampled relevant eta=0.01 l2=0.01 feats=default runs/243\n",
      "0.003 (-0.004--0.010)  288.000 singular/people(generation) sampled relevant eta=0.01 l2=0.01 feats=default runs/244\n",
      "0.006 (-0.001--0.014)  624.000 plural/furniture(generation) sampled relevant eta=0.01 l2=0.01 feats=default runs/245\n",
      "0.001 (-0.002--0.005)  576.000 plural/people(generation) sampled relevant eta=0.01 l2=0.01 feats=default runs/246\n",
      "0.724 (0.655--0.792)  152.167 singular/furniture sampled relevant eta=0.01 l2=0.01 feats=default runs/247\n",
      "0.408 (0.296--0.521)  180.160 singular/people sampled relevant eta=0.01 l2=0.01 feats=default runs/248\n",
      "0.160 (0.120--0.200) 1061.041 plural/furniture sampled relevant eta=0.01 l2=0.01 feats=default runs/249\n",
      "0.301 (0.255--0.348)  388.395 plural/people sampled relevant eta=0.01 l2=0.01 feats=default runs/250\n",
      "0.231 (0.161--0.301)  461.831 singular/furniture(generation) sampled null eta=0.01 l2=0.01 feats=default runs/251\n",
      "0.119 (0.099--0.140)  937.106 singular/people(generation) sampled null eta=0.01 l2=0.01 feats=default runs/252\n",
      "0.090 (0.076--0.104)  576.829 plural/furniture(generation) sampled null eta=0.01 l2=0.01 feats=default runs/253\n",
      "0.024 (0.006--0.041) 1390.280 plural/people(generation) sampled null eta=0.01 l2=0.01 feats=default runs/254\n",
      "0.890 (0.865--0.916)   70.561 singular/furniture sampled null eta=0.01 l2=0.01 feats=default runs/255\n",
      "0.544 (0.505--0.584)   96.527 singular/people sampled null eta=0.01 l2=0.01 feats=default runs/256\n",
      "0.449 (0.372--0.525) 1061.848 plural/furniture sampled null eta=0.01 l2=0.01 feats=default runs/257\n",
      "0.307 (0.257--0.356)  411.346 plural/people sampled null eta=0.01 l2=0.01 feats=default runs/258\n",
      "0.226 (0.163--0.289)  450.662 singular/furniture(generation) sampled eta=0.01 l2=0.01 feats=default runs/259\n",
      "0.800 ~0.000            2.101 singular/furniture eta=0.1 l2=0.01 feats=default runs/26\n",
      "0.072 (0.035--0.109)  788.363 singular/people(generation) sampled eta=0.01 l2=0.01 feats=default runs/260\n",
      "0.087 (0.075--0.099)  581.336 plural/furniture(generation) sampled eta=0.01 l2=0.01 feats=default runs/261\n",
      "0.008 (0.002--0.015) 1503.991 plural/people(generation) sampled eta=0.01 l2=0.01 feats=default runs/262\n",
      "0.881 (0.855--0.907)   78.715 singular/furniture sampled eta=0.01 l2=0.01 feats=default runs/263\n",
      "0.525 (0.485--0.565)  163.505 singular/people sampled eta=0.01 l2=0.01 feats=default runs/264\n",
      "0.423 (0.332--0.514)  873.478 plural/furniture sampled eta=0.01 l2=0.01 feats=default runs/265\n",
      "0.275 (0.262--0.288)  522.413 plural/people sampled eta=0.01 l2=0.01 feats=default runs/266\n",
      "0.160 (0.091--0.228)      nan singular/furniture(generation) eta=0.01 l2=0.01 feats=cross_product runs/267\n",
      "0.067 (0.038--0.095)      nan plural/furniture(generation) eta=0.01 l2=0.01 feats=cross_product runs/269\n",
      "0.800 ~0.000            2.101 singular/furniture eta=0.1 l2=0.01 feats=default runs/27\n",
      "0.044 (0.017--0.071)      nan plural/people(generation) eta=0.01 l2=0.01 feats=cross_product runs/270\n",
      "0.945 (0.915--0.976)      nan singular/furniture eta=0.01 l2=0.01 feats=cross_product runs/271\n",
      "0.958 (0.937--0.980)      nan singular/people eta=0.01 l2=0.01 feats=cross_product runs/272\n",
      "0.642 (0.581--0.704)      nan plural/furniture eta=0.01 l2=0.01 feats=cross_product runs/273\n",
      "0.818 (0.802--0.834)      nan plural/people eta=0.01 l2=0.01 feats=cross_product runs/274\n",
      "0.160 (0.091--0.228)  355.389 singular/furniture(generation) eta=0.01 l2=0.01 feats=cross_product runs/275\n",
      "0.094 (0.052--0.137)  328.031 singular/people(generation) eta=0.01 l2=0.01 feats=cross_product runs/276\n",
      "0.067 (0.038--0.095) -253.093 plural/furniture(generation) eta=0.01 l2=0.01 feats=cross_product runs/277\n",
      "0.044 (0.017--0.071) -210.101 plural/people(generation) eta=0.01 l2=0.01 feats=cross_product runs/278\n",
      "0.945 (0.915--0.976)    4.761 singular/furniture eta=0.01 l2=0.01 feats=cross_product runs/279\n",
      "0.958 (0.937--0.980)    0.607 singular/people eta=0.01 l2=0.01 feats=cross_product runs/280\n",
      "0.642 (0.581--0.704)  239.819 plural/furniture eta=0.01 l2=0.01 feats=cross_product runs/281\n",
      "0.818 (0.802--0.834)   41.754 plural/people eta=0.01 l2=0.01 feats=cross_product runs/282\n",
      "0.852 (0.510--1.195)   22.552 singular/furniture sampled null eta=0.1 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count'] runs/283\n",
      "0.193 (0.086--0.300)  746.663 singular/furniture(generation) sampled null eta=0.1 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count'] runs/284\n",
      "0.014 (0.008--0.020)  336.000 singular/furniture(generation) sampled null eta=0.01 l2=0.01 feats=['attr_type', 'attr_count'] runs/285\n",
      "0.003 (-0.004--0.010)  288.000 singular/people(generation) sampled null eta=0.01 l2=0.01 feats=['attr_type', 'attr_count'] runs/286\n",
      "0.233 (0.161--0.306)  630.590 singular/furniture(generation) sampled null eta=0.01 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count'] runs/287\n",
      "0.150 (0.113--0.187) 1036.821 singular/people(generation) sampled null eta=0.01 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count'] runs/288\n",
      "0.010 (0.004--0.015)  354.874 singular/furniture(generation) sampled null eta=0.01 l2=0.01 feats=['falsehood', 'attr_type', 'attr_count'] runs/289\n",
      "0.006 (-0.003--0.014)  394.703 singular/people(generation) sampled null eta=0.01 l2=0.01 feats=['falsehood', 'attr_type', 'attr_count'] runs/290\n",
      "0.272 (0.215--0.329)  196.241 singular/furniture(generation) sampled null loc=True eta=0.1 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count'] runs/291\n",
      "0.298 (0.234--0.361)  198.674 singular/furniture(generation) sampled null loc=True eta=0.01 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count'] runs/292\n",
      "0.134 (0.066--0.203)  294.861 singular/people(generation) sampled null loc=True eta=0.01 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count'] runs/293\n",
      "0.615 (0.522--0.707)  259.224 singular/furniture(generation) sampled null loc=False eta=0.01 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count'] runs/294\n",
      "0.415 (0.324--0.506)  535.635 singular/people(generation) sampled null loc=False eta=0.01 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count'] runs/295\n",
      "0.272 (0.215--0.329)  196.241 singular/furniture(generation) sampled null loc=True eta=0.1 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count'] runs/296\n",
      "0.157 (0.029--0.285)  281.481 singular/furniture(generation) sampled null loc=True eta=0.1 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count'] runs/297\n",
      "0.219 (-0.038--0.476)  265.776 singular/furniture(generation) sampled null loc=False eta=0.1 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count'] runs/298\n",
      "0.193 (0.086--0.300)  746.663 singular/furniture(generation) sampled null eta=0.1 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count'] runs/299\n",
      "0.200 (0.148--0.252)  281.664 singular/furniture(generation) sampled null loc=True eta=0.01 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count'] runs/300\n",
      "0.089 (-0.005--0.182)  425.759 singular/people(generation) sampled null loc=True eta=0.01 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count'] runs/301\n",
      "0.267 (0.184--0.349)  260.944 singular/furniture(generation) sampled null loc=False eta=0.01 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count'] runs/302\n",
      "0.206 (0.170--0.241)  376.494 singular/people(generation) sampled null loc=False eta=0.01 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count'] runs/303\n",
      "0.169 (0.081--0.257)  360.971 singular/furniture(generation) eta=0.01 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count'] runs/304\n",
      "0.108 (0.068--0.148)  333.679 singular/people(generation) eta=0.01 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count'] runs/305\n",
      "0.281 (0.204--0.358)  379.621 singular/furniture(generation) eta=0.01 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count', 'attr_pair'] runs/306\n",
      "0.178 (0.108--0.248)  356.878 singular/people(generation) eta=0.01 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count', 'attr_pair'] runs/307\n",
      "0.276 (0.222--0.330) 1951.247 singular/furniture(generation) sampled null eta=0.01 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count', 'attr_pair'] runs/308\n",
      "0.225 (0.170--0.280) 2202.787 singular/people(generation) sampled null eta=0.01 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count', 'attr_pair'] runs/309\n",
      "0.287 (0.257--0.304) 1671.392 singular/furniture(generation) sampled null eta=0.01 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count', 'attr_pair'] runs/310\n",
      "0.287 (0.258--0.304) 1671.392 singular/furniture(generation) sampled null eta=0.01 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count', 'attr_pair'] runs/312\n",
      "0.149 (0.131--0.174)  355.557 singular/furniture(generation) eta=0.01 l2=0.01 feats=['cross_product'] runs/314\n",
      "0.090 (0.061--0.117)  330.458 singular/people(generation) eta=0.01 l2=0.01 feats=['cross_product'] runs/315\n",
      "0.254 (0.227--0.282)  378.817 singular/furniture(generation) eta=0.01 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count', 'attr_pair'] runs/316\n",
      "1.000 (1.000--1.000)      nan None(2) eta=None l2=None feats=default runs/318\n",
      "0.926 (0.924--0.927)      nan None(3) eta=None l2=None feats=default runs/319\n",
      "1.000 (1.000--1.000)      nan None(2noamb) eta=None l2=None feats=default runs/321\n",
      "1.000 (1.000--1.000)      nan None(3noamb) eta=None l2=None feats=default runs/322\n",
      "1.000 (1.000--1.000)    0.000 singular/furniture sampled artificial(2) eta=0.1 l2=0.0 feats=['cross_product'] runs/324\n",
      "1.000 (1.000--1.000)    0.000 singular/furniture sampled artificial(2noamb) eta=0.1 l2=0.0 feats=['cross_product'] runs/327\n",
      "0.987 (0.986--0.988)    0.000 singular/furniture sampled artificial(3noamb) eta=0.1 l2=0.0 feats=['cross_product'] runs/328\n",
      "1.000 (1.000--1.000)    0.000 singular/furniture sampled artificial(2) eta=0.1 l2=0.0 feats=['cross_product'] runs/330\n",
      "0.911 (0.903--0.915)  139.455 singular/furniture sampled artificial(3) eta=0.1 l2=0.0 feats=['cross_product'] runs/331\n",
      "1.000 (1.000--1.000)    0.000 singular/furniture sampled artificial(2noamb) eta=0.1 l2=0.0 feats=['cross_product'] runs/333\n",
      "0.987 (0.986--0.988)    0.000 singular/furniture sampled artificial(3noamb) eta=0.1 l2=0.0 feats=['cross_product'] runs/334\n",
      "0.210 ~0.000           53.830 singular/furniture(generation) eta=0.1 l2=0.01 feats=default runs/36\n",
      "0.210 ~0.000           53.830 singular/furniture(generation) eta=0.1 l2=0.01 feats=default runs/37\n",
      "0.210 ~0.000           53.830 singular/furniture(generation) eta=0.1 l2=0.01 feats=default runs/39\n",
      "0.150 ~0.060              nan singular/furniture cpbaseline eta=0.1 l2=None feats=null runs/40\n",
      "0.170 ~0.080              nan singular/people cpbaseline eta=0.1 l2=None feats=null runs/41\n",
      "0.040 ~0.010              nan plural/furniture cpbaseline eta=0.1 l2=None feats=null runs/42\n",
      "0.030 ~0.040              nan plural/people cpbaseline eta=0.1 l2=None feats=null runs/43\n",
      "0.940 ~0.050              nan singular/furniture cpbaseline eta=0.1 l2=None feats=cross_product runs/44\n",
      "0.960 ~0.050              nan singular/people cpbaseline eta=0.1 l2=None feats=cross_product runs/45\n",
      "0.580 ~0.100              nan plural/furniture cpbaseline eta=0.1 l2=None feats=cross_product runs/46\n",
      "0.820 ~0.050              nan plural/people cpbaseline eta=0.1 l2=None feats=cross_product runs/47\n",
      "0.910 ~0.030           17.818 singular/furniture eta=0.1 l2=0.01 feats=default runs/66\n",
      "0.760 ~0.030            4.020 singular/people eta=0.1 l2=0.01 feats=default runs/67\n",
      "0.910 ~0.030           13.095 singular/furniture eta=0.01 l2=0.01 feats=default runs/70\n",
      "0.700 ~0.070          427.147 singular/furniture eta=1.0 l2=0.01 feats=default runs/71\n",
      "0.910 ~0.020            0.000 singular/furniture eta=0.1 l2=0.0 feats=default runs/72\n",
      "0.860 ~0.020          139.237 singular/furniture eta=0.1 l2=0.1 feats=default runs/73\n",
      "0.590 ~0.050          338.076 singular/furniture eta=0.1 l2=1.0 feats=default runs/74\n",
      "0.140 ~0.050              nan singular/furniture cpbaseline eta=0.1 l2=None feats=null runs/84\n",
      "0.130 ~0.070              nan singular/people cpbaseline eta=0.1 l2=None feats=null runs/85\n",
      "0.940 ~0.050              nan singular/furniture cpbaseline eta=0.1 l2=None feats=cross_product runs/86\n",
      "0.950 ~0.020              nan singular/people cpbaseline eta=0.1 l2=None feats=cross_product runs/87\n",
      "\n",
      "plural/furniture(generation): 0.11 ~0.02 (runs/144 plural/furniture(generation) eta=0.1 l2=0.01 feats=default)\n",
      "plural/people: 0.82 ~0.05 (runs/47 plural/people cpbaseline eta=0.1 l2=None feats=cross_product)\n",
      "plural/people(generation): 0.044 ~(0.017, 0.071) (runs/270 plural/people(generation) eta=0.01 l2=0.01 feats=cross_product)\n",
      "singular/people: 0.96 ~0.05 (runs/45 singular/people cpbaseline eta=0.1 l2=None feats=cross_product)\n",
      "None: 1.0 ~(1.0, 1.0) (runs/318 None(2) eta=None l2=None feats=default)\n",
      "singular/people(generation): 0.415 ~(0.324, 0.506) (runs/295 singular/people(generation) sampled null loc=False eta=0.01 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count'])\n",
      "singular/furniture: 1.0 ~(1.0, 1.0) (runs/324 singular/furniture sampled artificial(2) eta=0.1 l2=0.0 feats=['cross_product'])\n",
      "singular/furniture(generation): 0.615 ~(0.522, 0.707) (runs/294 singular/furniture(generation) sampled null loc=False eta=0.01 l2=0.01 feats=['cross_product', 'attr_type', 'attr_count'])\n",
      "plural/furniture: 0.642 ~(0.581, 0.704) (runs/273 plural/furniture eta=0.01 l2=0.01 feats=cross_product)\n"
     ]
    }
   ],
   "source": [
    "import math\n",
    "\n",
    "best = defaultdict(lambda: (None, {}, 0.0, 0.0))\n",
    "\n",
    "def accept(config):\n",
    "    return True\n",
    "\n",
    "error_corr = defaultdict(list)\n",
    "\n",
    "for dirname in sorted(glob('runs/*')):\n",
    "    statsfile = os.path.join(dirname, '_qry.json')\n",
    "    configfile = os.path.join(dirname, 'config.json')\n",
    "    try:\n",
    "        stats = load_json(statsfile)\n",
    "    except Exception, e:\n",
    "        #print '%s: %s' % (statsfile, e)\n",
    "        continue\n",
    "    try:\n",
    "        config = load_json(configfile)\n",
    "    except Exception, e:\n",
    "        #print '%s: %s' % (configfile, e)\n",
    "        config = defaultdict(lambda: None)\n",
    "\n",
    "    data_dir = config['data_dir']\n",
    "    accuracy = None\n",
    "    error = float(stats['Error']) if 'Error' in stats else float('nan')\n",
    "    if 'accuracy mean' in stats.keys():\n",
    "        accuracy = float(stats['accuracy mean'])\n",
    "        acc_std = float(stats['accuracy std'])\n",
    "    else:\n",
    "        for key, value in stats.iteritems():\n",
    "            if key.startswith('Mean accuracy'):\n",
    "                accuracy = float(key.split()[2])\n",
    "                acc_std = float(value[:-1])\n",
    "                break\n",
    "    \n",
    "    if accuracy is None:\n",
    "        accuracy, acc_std, error = extract_accuracy(os.path.join(dirname, 'stdout.log'))\n",
    "\n",
    "    if accuracy and accept(config):\n",
    "        settings = ''.join((str(config['data_dir']),\n",
    "                            '(generation)' if config['generation'] else ''))\n",
    "        _, _, best_acc, _ = best[settings]\n",
    "        if hasattr(acc_std, '__len__'):\n",
    "            metrics = '%5.3f (%5.3f--%5.3f) %8.3f' % ((accuracy,) + acc_std + (error,))\n",
    "        else:\n",
    "            metrics = '%5.3f ~%5.3f         %8.3f' % (accuracy, acc_std, error)\n",
    "        print metrics, summary(config)[1], dirname\n",
    "        if not math.isnan(error):\n",
    "            error_corr[summary(config)[0]].append((accuracy, error))\n",
    "        if accuracy > best_acc:\n",
    "            best[settings] = (dirname, config, accuracy, acc_std)\n",
    "\n",
    "\n",
    "print\n",
    "for settings, (run, config, accuracy, acc_std) in best.iteritems():\n",
    "    if run is not None:\n",
    "        print '%s: %s ~%s (%s %s)' % (settings, accuracy, acc_std, run, summary(config)[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "plural/furniture(generation)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYYAAAEACAYAAAC3adEgAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f20dcc22c50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "plural/people\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX4AAAEACAYAAAC08h1NAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nG2f+hyIEAmoAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f20dc110690>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "plural/people(generation)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYkAAAEACAYAAABGYoqtAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f20dcb24f10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "singular/people\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXgAAAEACAYAAAC57G0KAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nqVEGvCQ1yoCXpEb9P0JzCAGJPL74AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f20dcac9cd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "singular/people(generation)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYkAAAEACAYAAABGYoqtAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f20dca8d290>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "singular/furniture\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX4AAAEACAYAAAC08h1NAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f20dc89ee90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "singular/furniture(generation)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYkAAAEACAYAAABGYoqtAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n/wcebaJHlgC0BAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f20dc64ef90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "plural/furniture\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXsAAAEACAYAAABS29YJAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nkhpg2EtSA/4fYfik+dewL1sAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f20dc504f90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for dataset, pairs in error_corr.iteritems():\n",
    "    print dataset\n",
    "    y, x = zip(*pairs)\n",
    "    plot.scatter(x, y)\n",
    "    plot.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "c = hocon.parse_file('runs/42/_qry.json')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['begin_time',\n",
       " 'end_time',\n",
       " 'elapsed_time',\n",
       " 'host',\n",
       " 'working_dir',\n",
       " 'run_dir',\n",
       " 'git_rev',\n",
       " 'Trials',\n",
       " 'Train percentage',\n",
       " '\"Mean iterations to convergence:  99.00 (+/\"',\n",
       " '\"Mean accuracy: 0.04 (+/\"',\n",
       " 'Learning rate']"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "c.keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "420\n"
     ]
    }
   ],
   "source": [
    "from training_instances import get_singular_instances\n",
    "\n",
    "insts = get_singular_instances(glob('../TUNA/corpus/singular/furniture/*.xml'))\n",
    "print len(insts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(['grey', 'frontal', 'table'],\n",
       " ['colour:grey',\n",
       "  'orientation:front',\n",
       "  'type:desk',\n",
       "  'size:large',\n",
       "  'x-dimension:4',\n",
       "  'y-dimension:1'],\n",
       " [['colour:grey',\n",
       "   'orientation:front',\n",
       "   'type:desk',\n",
       "   'size:large',\n",
       "   'x-dimension:4',\n",
       "   'y-dimension:1'],\n",
       "  ['colour:blue',\n",
       "   'orientation:front',\n",
       "   'type:desk',\n",
       "   'size:large',\n",
       "   'x-dimension:4',\n",
       "   'y-dimension:3'],\n",
       "  ['colour:red',\n",
       "   'orientation:back',\n",
       "   'type:desk',\n",
       "   'size:large',\n",
       "   'x-dimension:3',\n",
       "   'y-dimension:3'],\n",
       "  ['colour:green',\n",
       "   'orientation:left',\n",
       "   'type:desk',\n",
       "   'size:small',\n",
       "   'x-dimension:2',\n",
       "   'y-dimension:2'],\n",
       "  ['colour:blue',\n",
       "   'orientation:front',\n",
       "   'type:fan',\n",
       "   'size:large',\n",
       "   'x-dimension:5',\n",
       "   'y-dimension:1'],\n",
       "  ['colour:red',\n",
       "   'orientation:back',\n",
       "   'type:fan',\n",
       "   'size:large',\n",
       "   'x-dimension:4',\n",
       "   'y-dimension:2'],\n",
       "  ['colour:green',\n",
       "   'orientation:left',\n",
       "   'type:fan',\n",
       "   'size:small',\n",
       "   'x-dimension:5',\n",
       "   'y-dimension:3']])"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "insts[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "410"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(set((tuple(x), tuple(y)) for x, y, d in insts))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "720\n"
     ]
    }
   ],
   "source": [
    "from training_instances import get_generation_instances\n",
    "\n",
    "insts = get_generation_instances(glob('../TUNA/corpus/plural/people/*.xml'))\n",
    "print len(insts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXgAAAEACAYAAAC57G0KAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f20ddd70550>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dist, _, _ = plot.hist([len(i[2]) for i in insts], bins=np.arange(-0.5, 12.5, 1.0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAEACAYAAABfxaZOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAEFJJREFUeJzt3X2MZXV9x/H3BxdRRIOwC1FmEeJDbHnyISolNqytJnRX\nsYaGVsGnQtrUEMFYNNYKSxPTWKK0aOIfgiQuiitIqRaQ2LQTn4jVsJFlV2s1qe6gsA9IlT98qPvt\nH3O2O12H3bn3zr1n7v7er2SSc86cc+539d43Z+7DTKoKSVIbjuh7AEnS5Bh9SWqI0Zekhhh9SWqI\n0Zekhhh9SWrIQaOf5ONJHk6ydcG245J8Mcn9Se5JcuyC712fZFuS+5K8cMH2N3fbtyV503j+KZKk\nQznUlf5NwHkHbLsGuLOqzgTu7tZJcgFwclWdBlzSHUuSZwDvA17WfV2V5MRl+xdIkpbsoNGvqi8D\nPzlg83pgU7d8M7ChW96wb3tVbQFWJZkBXgXcXVWPVdVjwBe6bZKkCRvmOf01VbUHoKp2Ayd0208C\ndizYbw6Y6bbPLbJdkjRhy/1Cbpb5fJKkZbRqiGN2JVldVbuTrAF2dtvngLXA17v1Geav/OeYfy5/\nn7XA1xY7cRJ/EZAkDaiqln7BXVUH/QJOAbYuWP8wcEW3/A7g+m75AuAfu+UXAd/qlp8JfA94avf1\nfeDEx7mtOtQ8K/UL2Nj3DM7f/xzOP51f0zz/oN086JV+kluAc4HVSXYAVwFXA5uT/CnwEHBhd6uf\nTfKKJNuAXwBv7bb/KMn72f8TwN9U1cNL/q+SJGnZHDT6VfX6x/nWou++qarLHmf7TXRv4ZQk9cdP\n5C6f2b4HGNFs3wOMaLbvAUY02/cAI5rte4ARzfY9wKSke05oRUhSNcgLEpLUuEG76ZW+JDXE6EtS\nQ4y+JDXE6EtSQ4y+JDXE6EtSQ4y+JDXE6EtSQ4y+JDXE6EtSQ4y+JDVkmD+iIukQJvEHgfw9VRqG\n0ZfGZpzdt/cajk/vSFJDjL4kNcToS1JDjL4kNcToS1JDjL4kNcToS1JDjL4kNcToS1JDjL4kNcTo\nS1JDjL4kNcToS1JDjL4kNcToS1JDjL4kNcToS1JDjL4kNcToS1JDjL4kNcToS1JDho5+kmuSfDfJ\nd5LcluToJKcmuTfJ1iSfTnJkt+9RSTZ327+a5FnL90+QJC3VUNFP8hzgjcDpVfV84NfA64HrgQ9U\n1RnAQ8Bl3SGXAT/utl/b7SdJmrBhr/QfAX4FPCXJKuBo4IfA2VV1R7fPzcCGbnk9sKlb/hxwTpIM\neduSpCENFf2qegT4IPOh/xHwKPAAsHvBbg8CM93yDLCjO3YvsAc4YbiRJUnDWjXMQUmeDVwBnAL8\nN3Ar8KrlGCjJxgWrs1U1uxznlaTDQZJ1wLphjx8q+sBLga9V1Z5uiNuBc4HVC/aZAea65TngZGBn\nkiOA44Fdi524qjYOOZMkHfa6C+HZfetJrh7k+GGj/z3gvUmeDPwceCWwFTgxyR92z+tfDNzV7X9X\nt/5N4LXAvd3TPFIvklTfM4xq3P+GqvJ1t8NQqoa733RPw1wE7AW2AG8BngF8CjgG2Aa8sap+leQo\n5l/I/S3gZ8Abquq/FjlneUfTJMwHc5zNDNN+fh+L02HQbg4d/XEw+poUo3/o8/tYnA6DdtNP5EpS\nQ4y+JDXE6EtSQ4y+JDXE6EtSQ4y+JDXE6EtSQ4y+JDXE6EtSQ4y+JDXE6EtSQ4y+JDXE6EtSQ4y+\nJDXE6EtSQ4y+JDXE6EtSQ4y+JDXE6EtSQ4y+JDXE6EtSQ4y+JDXE6EtSQ4y+JDXE6EtSQ4y+JDXE\n6EtSQ4y+JDXE6EtSQ4y+JDXE6EtSQ4y+JDXE6EtSQ4y+JDXE6EtSQ4aOfpJjk9ya5FtJvp3k7CTH\nJflikvuT3JPk2AX7X59kW5L7krxwecaXJA1ilCv9jwG3V9VZwGnAduAa4M6qOhO4u1snyQXAyVV1\nGnAJcNNIU0uShjJU9JMcD7ygqm4BqKq9VfVTYD2wqdvtZmBDt7xh3/aq2gKsSjIzyuCSpMENe6X/\nXGBXks8keSDJJ5I8FVhTVXsAqmo3cEK3/0nAjgXHzwFGX5ImbNjoHwG8BLi2qk4HHgHed4hjcsB6\nDXnbkqQhrRryuB3Ag1X1jW79NuAqYGeS1VW1O8kaYGf3/TlgLfD1bn2m2/YbkmxcsDpbVbNDzihJ\nh50k64B1Qx9fNdwFd5JvAm+oqu92oX468z8BfL+q/j7JO4BTq+rt3Qu5F1fV65K8CLipewH4wHNW\nVR34E4G07JLUeH/YDNN+fh+L02HQbo4S/bOAG4CjgR8AFzF/T9wMnAg8BFxYVY92+38EeAXwC+DS\nqrpv1OGlYRn9Q5/fx+J0mFj0x8Hoa1KM/qHP72NxOgzaTT+RK0kNMfqS1BCjL0kNMfqS1BCjL0kN\nMfqS1BCjL0kNMfqS1BCjL0kNMfqS1BCjL0kNMfqS1BCjL0kNMfqS1BCjL0kNMfqS1BCjL0kNMfqS\n1BCjL0kNWdX3ANJi5v+Grfo07v8P/Bu8/TD6WsHG/YfFdXD+73848ukdSWqI0Zekhhh9SWqI0Zek\nhhh9SWqI0Zekhhh9SWqI0Zekhhh9SWqI0Zekhhh9SWqI0Zekhhh9SWqI0Zekhhh9SWrISNFP8oQk\nW5J8vls/Ncm9SbYm+XSSI7vtRyXZ3G3/apJnLcfwkqTBjHqlfzmwnf1/beF64ANVdQbwEHBZt/0y\n4Mfd9mu7/SRJEzZ09JPMAOuBG+ZX8wTg7Kq6o9vlZmBDt7we2NQtfw44J4l/OkeSJmyUK/3rgCuB\nvd36CcDuBd9/EJjplmeAHQBVtRfY0+0vSZqgoaKf5NXAzqrawv4/dumVuyStcMP+YfRzgPOTrAee\nBDwN+ACwesE+M8BctzwHnAzsTHIEcDywa7ETJ9m4YHW2qmaHnFGSDjtJ1gHrhj6+arS/eJ/kXOAv\nq+o13bt4bqyqO5L8A/CDqvpQkncCa6vqiiSvA95aVecvcq6qKn9iEElq//sDxnILeP5+z+9jfXkM\n2s3lep/+vnvH24F3J9kKnAh8uNv+EeCZ3fYru/0kSRM28pX+cvJKX/t4pX/4n9/H+vLo60pfkjQF\njL4kNcToS1JDjL4kNcToS1JDjL4kNcToS1JDjL4kNcToS1JDjL4kNcToS1JDjL4kNcToS1JDjL4k\nNcToS1JDjL4kNcToS1JDjL4kNcToS1JDjL4kNcToS1JDjL4kNcToS1JDjL4kNcToS1JDjL4kNcTo\nS1JDjL4kNcToS1JDjL4kNcToS1JDjL4kNcToS1JDjL4kNcToS1JDjL4kNWTVMAclWQt8Eng68ETg\nxqr6uyTHAZuBE4EfA39cVY92x1wP/D7wC+CSqtqyDPOrJ0mq7xk03cZ9H6qqjPP802rYK/1fAm+r\nqjOAFwOXJjkLuAa4s6rOBO7u1klyAXByVZ0GXALcNPLkWgFqjF86/Hn/6cNQ0a+qh6vqgW75MeB+\n4CRgPbCp2+1mYEO3vGHf9u4Kf1WSmRHmliQNYeTn9JOcArwE+Aqwpqr2AFTVbuCEbreTgB0LDpsD\njL4kTdhQz+nvk+QY4Dbg8qr6aXLQp9AO/OaiP4Ml2bhgdbaqZkeZUZIOJ0nWAeuGPX7o6Cc5Evgs\n8MmquqPbvCvJ6qranWQNsLPbPgesBb7erc90235DVW0cdiZJOtx1F8Kz+9aTXD3I8UM9vZP5S/ob\nge1Vdd2Cb90FXNwtX9yt79t+UXfsi4BfV9WDw9y2JGl4qRr8le4kLwe+xPwLuPtO8B7g39n/ls2H\ngAsXvGXzI8ArmH/L5qVVdd8i5y3fZjUd5t9uN853SQTP7/lHOX8rLRm0m0NFf1yM/vQw+p5/pZ+/\nlZYM2k0/kStJDTH6ktQQoy9JDTH6ktQQoy9JDTH6ktQQoy9JDTH6ktQQoy9JDTH6ktQQoy9JDTH6\nktQQoy9JDTH6ktQQoy9JDTH6ktQQoy9JDTH6ktQQoy9JDTH6ktQQoy9JDTH6ktQQoy9JDTH6ktQQ\noy9JDVnV9wAajyTV9wxSn8b9GKiqjPP842L0D2vjvM9P5f1dTfH+vxif3pGkhhh9SWqI0Zekhhh9\nSWqI0Zekhhh9SWqI0Zekhhh9SWrIRKOf5LwkW5NsT/LuSd62JGmC0U9yFPBR4DzgTOCPkrxwUrc/\nbknW9T3DaGb7HmBEs30PMKLZvgcY0WzfA4xotu8BJmaSV/ovA7ZV1YNV9T/AZmDDBG9/3NYNsnOS\nGufX4OPPDn7IijLb9wAjmu17gBHN9j3AiGYHPmJlPX6XbpK/e2cG2LFgfY4BQ3n48XeDSNNrXI/f\n8T52Jxn9qfqtj0kuBV47wCHPS/Licc0jScshVZNpcZLfBd5dVa/u1q8EnlhV71+wz1T9h0GSVoJB\nfs3zJK/0vwGcnuQkYCdwIfDnC3eY1t9PLUnTYmLRr6qfJ/kL4B7mX0DeVFX3Ter2JUkTfHpHktS/\nFfWJ3CQf6j64tT3JPyc5vu+ZlmKaP3SWZG2SL3Xz/0eSd/U906CSPCHJliSf73uWQSU5NsmtSb6V\n5NtJfqfvmQaR5Jok303ynSS3JTm675kOJsnHkzycZOuCbccl+WKS+5Pck+TYPmc8mMeZf6Burqjo\nA58HTq+q3wYeAP6653kO6TD40NkvgbdV1RnAi4FLk5zV80yDuhzYzpS9Q6zzMeD2qjoLOA3Y1vM8\nS5bkOcAbmX/MPh/4NfD6fqc6pJuYf6wudA1wZ1WdCdzdra9Ui80/UDdXVPSr6t+qam+3+lXgpD7n\nWaKp/tBZVT1cVQ90y48B9wPP7HeqpUsyA6wHbmDKPpzQXZG9oKpuAaiqvVX1057HGsQjwK+ApyRZ\nBRwN/KDfkQ6uqr4M/OSAzeuBTd3yzazgx+9i8w/azRUV/QP8GfBPfQ+xBIt96Gymp1lGkuQU4CXA\nV/qdZCDXAVcCew+14wr0XGBXks8keSDJJ5Ic0/dQS1VVjwAfBH4I/Ah4tKr+pd+phrKmqvYAVNVu\n4ISe5xnFIbs58eh3z51tXeTrNQv2eS/wy6r65KTnG8I0PqXwG7rY3ApcXlU/63uepUjyamBnVW1h\nyq7yO0cw/x/Za6vqdOavnN/X70hLl+TZwBXAKcz/dHhMkot6HaphS+3mJN+nD0BVvepg30/yZuZ/\nvPq9yUw0sjlg7YL1tfz/K/8VL8mRwGeBT1XVHX3PM4BzgPOTrAeeBDwtySeq6k09z7VUO4AHq+ob\n3fptTFH0gZcCX9t3lZzkduDlwDRcrC20K8nqqtqdZA3znyOaKoN0c0U9vZPkPOBdwPlV9fO+51mi\n//vQWRfPC5l/MWgqJAlwI7C9qq7re55BVNVfVdXaqjoV+BPgX6co+FTVDmB3kud1m14JfLvHkQb1\nPeDsJE/u7kev7LZNm7uAi7vli7v1qTFoN1fU+/ST/CfwROZ/zAW4t6re1uNIS5LkD4Br2f+hs7/t\neaQlS/Jy4EvMv4C7787wnqr6Qn9TDS7JucA7q+r8vmcZRPdOqRvY/yLoRVV14AuNK1aSjcBFzL+m\nsgV4y0q+YEtyC3AusBp4GLiK+efANwMnAg8BF1bVo70NeRCLzH818B4G6OaKir4kabxW1NM7kqTx\nMvqS1BCjL0kNMfqS1BCjL0kNMfqS1BCjL0kNMfqS1JD/BWisEd4d4V2oAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f20ddd35650>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dist_alts, _, _ = plot.hist([len(a) for a in insts[0][-1]], bins=np.arange(-0.5, 12.5, 1.0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7f20dd588c10>,\n",
       " <matplotlib.lines.Line2D at 0x7f20dd588e90>,\n",
       " <matplotlib.lines.Line2D at 0x7f20dd596590>]"
      ]
     },
     "execution_count": 146,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXUAAAEACAYAAABMEua6AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nHzAxZGW7AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f20dd5f88d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "true_cum = np.cumsum(dist) / sum(dist)\n",
    "alts_cum = np.cumsum(dist_alts) / sum(dist_alts)\n",
    "\n",
    "plot.plot(range(len(true_cum)), true_cum,\n",
    "          range(len(alts_cum)), alts_cum,\n",
    "          range(len(alts_cum)), true_cum / (alts_cum + 1.0 - true_cum) / 2.0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4.2032486565705911"
      ]
     },
     "execution_count": 150,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "alts_cum[4] * 21.7"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([    0.,    12.,    78.,   298.,   793.,  1585.,  2509.,  3301.,\n",
       "        3796.,  4016.,  4082.,  4094.])"
      ]
     },
     "execution_count": 149,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.cumsum(dist_alts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}