plots.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "import pickle\n",
    "%matplotlib inline\n",
    "import re"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "y_train = pd.read_csv('https://s3-us-west-2.amazonaws.com/pphilip-usp-inhibition/'\n",
    "                      'data/y_train_postprocessing.csv')\n",
    "y_test = pd.read_csv('https://s3-us-west-2.amazonaws.com/pphilip-usp-inhibition/'\n",
    "                     'data/y_test_postprocessing.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "y_train.columns = ['ID', 'Activity_Score']\n",
    "df_train = y_train.groupby('Activity_Score')['ID'].nunique().to_frame()\n",
    "df_train['score'] = df_train.index\n",
    "df_train.columns = ['score_counts', 'score']\n",
    "df_train = df_train.reset_index(drop=True)\n",
    "\n",
    "y_test.columns = ['ID', 'Activity_Score']\n",
    "df_test = y_test.groupby('Activity_Score')['ID'].nunique().to_frame()\n",
    "df_test['score'] = df_test.index\n",
    "df_test.columns = ['score_counts', 'score']\n",
    "df_test = df_test.reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAg8AAAEKCAYAAACVALaMAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nUOtiXAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f2434049dd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(df_train['score'], df_train['score_counts'], label='Training data')\n",
    "plt.plot(df_test['score'], df_test['score_counts'], label='Test data')\n",
    "plt.xlabel('Inhibition score')\n",
    "plt.ylabel('Number of molecule samples')\n",
    "plt.axvspan(0, 40, facecolor='orange', alpha=0.2)\n",
    "plt.axvspan(40, 100, facecolor='violet', alpha=0.2)\n",
    "plt.legend(bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0.)\n",
    "plt.savefig('./plots/all_score_counts.png', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfwAAAEKCAYAAAD3mecXAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nREQkA/x/QMdP9NsbK/8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f23e5506450>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(df_train['score'], df_train['score_counts'], label='Training data')\n",
    "plt.plot(df_test['score'], df_test['score_counts'], label='Test data')\n",
    "plt.xlabel('Inhibition score')\n",
    "plt.ylabel('Number of molecule samples')\n",
    "axes = plt.gca()\n",
    "axes.set_xlim([40, 100])\n",
    "axes.set_ylim([0, 250])\n",
    "plt.legend(bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0.)\n",
    "plt.savefig('./plots/high_score_counts.png', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "668.0\n",
      "232.0\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWQAAADxCAYAAAD8x81kAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fa157a78390>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWQAAADxCAYAAAD8x81kAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAHmtJREFUeJzt3XmcFNW99/HPbwYUMFIogoILGBfAqCCgz6OIS2KMGo3X\nNW6X4G4gcbtmM0t5ksdEr5rFuMRoriYuuaKoF9e4g4qKghI31Cdo4gLGtVBBGKbP/aNqzDAMMNjL\nqe76vl+vfjEzvX17KL6cPlV9yrz3iIhIeE2hA4iISEqFLCKSEypkEZGcUCGLiOSECllEJCdUyCIi\nOaFCloZiZh914Tavmtl6nfz8a2b2/ezrq8zs4E5uM9rMLsy+3s3Mdmp33UlmNq68VyBF1i10AJG8\n8N5PAaas4jZPAk9m3+4GfARMz677XTXzSePTCFkaUjZ6fdDMbjSzOWZ2rZlZu5t828xmmdkzZjY0\nu894M7uo3W32MLMnzewlM9u33ePeZmaDgZOA08zsaTMba2ZnmdkZ2e0eNLNzzWxGdv+x2c97mdkk\nM3vezG42s8fNbHQtfieSfxohSyPbDvgC8CbwCDAGeDi77h3v/UgzmwCcARzXyf0HAzsAmwEPmNnm\nbVd47181s98BH3nvzwcwsy91uH837/0OZrYPEAN7ABOA9733W5nZ1sDTlXmp0gg0QpZGNsN7/7r3\nvkRafIPbXXdT9ufMDj9vb5L3vuS9fxmYCwxdzefv7Dl2Bv4bwHv/LPDX1XxMaWAqZGlki9t93cqy\n7wgXr+Dn7XVc6GV1F37pynOIfEqFLLJih5hZk5ltBnweeLHD9R8Ca6/mYz4CHApgZlsB25SdUhqG\nCllkxf4BzADuBE7y3n/S4fpbgQPadup18TEvAfqZ2fPA/wOeA5JKBZb6Zlp+U6R2zKwZ6O69/yQb\ned8LDPHeLwkcTXJA81oitdWL9IiN7oABE1TG0kYjZBGRnNAcsohITqiQRURyQoUsIpIT2qknuZU4\n1x3YABjQ4dIX6E66/XYHuvU+5KctLX/fZv6iGQf0Az5pd/kQeB14Lbu8HsXx4o7PJZIHKmQJKnGu\nBzAcGAWMADYhLd2BpMVrK753O1b60LovngXsuopb+sS5t/lXQbddngcej+L4nc/wMkQqQoUsNZM4\n15O0dEe1uwyjQtuhLzV35WYG9M8uozrJOJf0wyCPZ5enojju+IEQkapQIUvVZFMOuwD7AV8iLd8u\nteZnUmqqxDGcn88uh2XftyTOzSYt50eAO6M4/qACzyOyHBWyVFTi3DrAPsDXgK8AUc2e3Dd3bXpj\n9XQHRmeXiaQF/QDpSm63RHH8VhWeUwpKhSxlS5zbgrSA9yNdczjMdlWqyUFD3YE9s8sliXOPkpbz\nzVEcv1KLANK4VMjymSTO9QGOJF3YfUTgOAD4UlVGyCvTRPof0BjggsS5p4HJwJVRHL9R4yzSAFTI\nsloS53YGjgcOAXoGjrOsru3Uq6YR2SVOnLsVuBS4N4pjrU8gXaJCllVKnFsTOAI4mZyMhjtV+xHy\ninQDDsguLyfOXQT8VxTHqzwjthSbCllWKHGuP/Bt4ESgX+A4q9TFw95qbQvgN4BLnLsMuDCK4zcD\nZ5KcUiHLchLnegPfAU4D1gocp+t8U15GyJ3pA3wPOD1x7logjuL4H4EzSc6okOVT2dTEROAHwHqB\n46y+/ExZrEx3YDxwWOLcb4BfRHGsM4YIoEIWIHGuGRgHnEX60eW6FOAoi3L0IB0xH5s491Pgd1Ec\ntwTOJIFptbeCS5z7N9JT0f8XdVzGAJSa63F7Xg+4EHguce7A0GEkLI2QCypxbhhwOekxtI2hvkbI\nHW0BTE6cewT4jyiOHw8dSGpPhVww2fTEGYAD1gwcp6LqbMpiRcYAjyXO/R44PYrjj0MHktqpx7d4\n8hllo+LpwDk0WBkDUGpqpO35BGB24lzjvIORVdIIuQCyUfF3SHfaNV4Rt6nPOeSV2QyYljh3PvDj\nKI51duoG12gbsHSQOLcV6aj4FzRyGQO+8QoZ0n+j3wWeSJzbNnQYqS6NkBtU4pyRjop/SoMX8aca\ns5DbbEtayjFwXhTHraEDSeU18gZcWIlzawE3AudSlDIGKOX6k3qVsAbpO51piXMbhQ4jladCbjCJ\nc4NJpygKd0yr9w21U29ldgJmJM5tHzqIVFZRNuBCSJzbFXiC9O1t8ZSac7m6UJUMAKYmzh0SOohU\njgq5QSTOfRO4h3pcg6JSGnsOuTM9gesT534cOohUhnbq1bnsRKIXAieFzhKaL9YIuY0BP02c2xI4\nLorjxaEDyWdXtBFFQ0mc60s6Ki58GQNQnDnkzhwF3J84l/t1q2XFirwB17XEuQ2AqcCuobPkRvGm\nLDpq29m3Vegg8tkUfQOuS9khT1OBL4TOkiu+SVNwMJh0Z18xd+zWORVynUmc2xSYBmwZOkve+FJT\nEeeQO7MecF/i3PDQQWT1qJDrSHaM8VRg07BJckqF3F5bKef3pLSyHBVynUic2xi4H9g4dJbc8oU8\nymJl+gL3Js5tHTqIdI0KuQ4kzg0A7kMj45XyJc0hd6IvcE/i3Gahg8iqqZBzLju07T7SM0rIyhT7\nsLeV2YB0pKz1L3JOG3COZR/6uBEYFjpLXdBRFiszmHSkXNxPctYBFXK+XQjsFjpE3dCUxaoMBSZl\nJyyQHFIh51Ti3AT0CbzV4fE6yqILdgfOCx1COmfe+9AZpIPEuS8Cf6Hga43c+/LLfP+uu2gtlRg3\nciSnjR27zPUXTZ/O1bNm0dzURL9B/1z62zGndtukTx9efucdjps8mZbWVn69337ssPHGLG1t5aBr\nruHPhx9OrzXWCPSKcuXfozi+JnQIWZZGyDmT7Q2/gYKXcWupxBl33MGNRx7J4xMncuOzzzLnn/9c\n5jbbDhjAAyecwPQJEzhgDxbH99wDwJVPPsk5e+3FDUceyW+nTwfgD08+yaHbbqsy/pffJ85tFzqE\nLEuFnCOJc72BKcC6obOENvONN/j8uusyeN11WaNbNw7aemvuePHFZW6zy6abflqwO45g8ZsLFgDQ\nvbmZRS0tLGppoXtTEx8sWsRdL77I4cP1wbV2egI3aydfvqiQcyJxrgm4DtDCMMC8BQvYsHfvT78f\n2Ls387LC7czlN9Bjj803B+C47bfngoce4pu33MLpY8dy3rRpnD52LE06Kq6jQaTrKWvuPScK/bY4\nZ74LfDV0iHp0/ezZzHqRbnceOAaAjfv04fajjwZg7rvv8uaCBQzp148TbrqJltZWfrj77my+ngaG\nmS+S7uQ7PXQQ0Qg5F7KPtrrQOfJkQO/evNFuRPzmggUMaDdibvPg3/7GBQ89xK2X2D/X7Lb8+OJn\n99/Pj774RS57/HHGjRyJ+/KXOXfq1Kpmr0OnJc4dEDqEqJCDS5zrBlxFekZhyYwcOJC/vfsur77/\nPkuWLmXys8+y95Ahy9xm9rx5nHrbbfz58MMZsF5TS8fHePjVV9lg7bXZrG9fFra00GRGkxkLW5a7\nqcAliXPrhA5RdJqyCO8HwKjQIfKmW3Mz5+2zDwddfTWt3nPUdtsxrH9/zr7/frYbOJB9hg7lJ3ff\nzcdLlvCNSZNovb11w0HdruO/jzgCAO8950+bxpUHHwzA+FGjOH7yZJaWSvxy331DvrS82gD4JXB0\n6CBFpuOQA8rWq30C6B46S73rfdDZLy6Y/MMhq76lrMJXoji+O3SIotKURSDZOhVXoTKuDG8aWVTG\n7xPnPhc6RFGpkMP5EaDFwyvGWkMnaBCDgJ+HDlFUKuQAsk9InRk6R0PxVgodoYFMTJwbEzpEEamQ\nayxxzoDL0A7VivIq5EpqAq5InFszdJCiUSHX3sHA9qFDNBwVcqUNBb4fOkTRqJBrKDvm+OzQORqS\ndupVw39orYvaUiHX1vHoVEzV4Zs0Qq68tYHvhQ5RJCrkGkmcWwv4SegcDUuFXC0Ts5PsSg2okGvn\nNNJPQ0k1aMqiWnqSHqIpNaBCroFsHu47oXM0Mq8RcjUdlzg3OHSIIlAh18aPgOWXKpPK8U0aIVfP\nGmi6rSZUyFWWOLcJ8M3QORqepiyqbVzinNYKqTIVcvWdgpbWrL5Sswq5uprRmt1Vp0KuomyRlmND\n5ygCrymLWjgkcW7T0CEamQq5usYDUegQhVDSplwDTWj6raq0FVdJtmbFyaFzFIZGyLVybOJcz9Ah\nGpUKuXr2RJ/Kq52SCrlG1gUOCx2iUamQq+fE0AGKxHttyjU0IXSARqWtuAoS5zYA9gudo1BKzaET\nFMnoxLltQ4doRCrk6jgGrXdcWyrkWtPJUKtAhVwdx4QOUDiasqi1o7LzQkoFaSuusMS5EcBmoXMU\njdcIudbWQ9NyFadCrryvhQ5QSDrKIoRDQwdoNCrkyts/dIBCKjVb6AgFtGfinN6aVJAKuYIS5zYC\nRobOUUSasghiHWDH0CEaiQq5sjRdEYpv0gg5jH1CB2gkKuTKUiGHoimLUL4aOkAjUSFXSOLc2sDu\noXMUlaYsgtk2cW7D0CEahQq5cvZC6x6HU9KURUCatqgQFXLlaLoiJE1ZhKRCrhAVcuWMDR2gyLwK\nOaQ9Euf07rACVMgVkJ1VelDoHIWmQg7pc8D/CR2iEaiQK2N06ACFV2rWthzWdqEDNAJtxJUxKnSA\notOURXDDQwdoBCrkytAIOTR9MCS0EaEDNAIVcmWokEPTlEVoX0ic0xrgZdJGXKbEufWBjULnKDqv\nQg5tTWBY6BD1Thtx+TQ6zgMVch5o2qJM2ojLpx16eaBCzgPt2CuTNuLybRM6gIAvNWlbDk8j5DJp\nIy6f5o/zwKuQc0Aj5DJpIy7fwNABBCg1a7m38NZLnOsZOkQ9UyGXIXHOgAGhcwiaQ86P/qED1DNt\nxOXpB+hU6DngS00aIefD+qED1DMVcnk0XZEXmrLIC42Qy6BCLo8KOS+0Uy8vVMhl0EZcHhVyTmjK\nIjdUyGVQIZdHhZwXXlMWOaE55DKokMujQs4LjZDzQiPkMqiQy9M7dADJ+CatNJYPKuQyqJDLo0Pe\nckKrveVG39AB6pk24vKokPPCm0bI+aBOKYN+eeXRmXbzQlMWeaFOKYN+eeVRCeRFSYWcEzqVVhlU\nyNIgdE69nNDfQxlUyOVZGjqASM740AHqmQq5PK2hA4jkzOLQAeqZCrk8KmSRZamQy6BCLk9L6AAi\nOaNCLoMKuTzvhA4gkjMq5DKokMszP3QAkZxJQgeoZyrk8qiQRZb1eugA9UyFXJ55oQOI5IwKuQwq\n5PJohCyyLBVyGVTI5VEhiyxLhVwGFXJ5VMgiy1Ihl0GFXIYojpcA74fOIZITrWi/SllUyOXTBiiS\nmhfFsT69WgYVcvleDR1AJCc0XVEmFXL5ng4dQCQnXgsdoN6pkMs3K3QAkZyYHTpAvVMhl0+FLJKa\nETpAvVMhlymK41fQkRYiHngidIh6p0KuDM0jS9H9/yiOPwgdot6pkCtD0xZSdJquqAAVcmWokKXo\nVMgVoEKuDBWyFJ0KuQJUyJXxEvBR6BAigbSg/SgVoUKugCiOS8B9oXOIBPJMFMefhA7RCFTIlXNb\n6AAigUwNHaBRqJAr5zbSYzFFiubm0AEahQq5QqI4ng/MDJ1DpMbeBh4JHaJRqJArS9MWUjRTsn0o\nUgEq5Mq6NXQAkRrTdEUFqZArKIrjWcCboXOI1MhHwL2hQzQSFXLladpCiuLOKI4Xhw7RSFTIladp\nCykKTVdUmAq58u4B3gsdQqTKlgC3hw7RaFTIFZa9hbsmdA6RKrsniuMFoUM0GhVydVwROoBIlV0W\nOkAjUiFXQRTHz6DVr6Rx/QNNV1SFCrl6Lg8dQKRKrtCHQapDhVw916Fz7UnjWYqm5KpGhVwlURwv\nBP4QOodIhd0QxfG80CEalQq5ui4G9NZOGskFoQM0MhVyFUVx/CowJXQOkQp5KIpjrWhYRSrk6vt5\n6AAiFfLL0AEanQq5yqI4fgK4KXQOkTI9j97tVZ0KuTZ+BLSGDiFShu/pULfqUyHXQBTHLwBXh84h\n8hk9GMWxVjGsARVy7ZwFaKlCqTce+G7oEEWhQq6RKI7/DvwudA6R1TQp2w8iNaBCrq2zSc+yIFIP\nlgBnhg5RJCrkGori+G3gV6FziHTRJVEczw0dokhUyLV3PvBW6BAiq/AB8LPQIYpGhVxj2aLeE0Ln\nEFmFX0RxrDPf1JgKOYAojm8CJoXOIbICs9DUWhAq5HC+BbwdOoRIB4uBcVEct4QOUkQq5ECyHXzf\nCp1DpIMfR3H8XOgQRWXe+9AZCi1xbjJwYOgceXDpY4/xp5kz8cC4kSOZsOOOPDN/PqffdhsfL1nC\nxn36cPmBB9K7R4/l7nvxo49y9axZGLDV+utz8f7706N7d46fPJnn3nqLvbbckp/ssQcA502dyrD+\n/dl32LDavsD8exjYVR+RDkcj5PAmAO+GDhHa82+9xZ9mzuS+44/n4ZNO4i8vvcTcd9/l5ClTiPfY\ng+kTJrDv0KFcOH36cvd9c8ECLnv8cR444QQenTiR1lKJyc8+y7Pz59OjWzemT5jArDffJPnkE+Z/\n+CFPvvGGynh5HwPjVcZhqZADi+L4LeCU0DlCe+mddxi10Ub0WmMNujU3M2bwYG594QX+9u67jBk0\nCIDdN9uMW59/vtP7t5ZKfNLSwtLWVha1tDBg7bXp3tzMJ0uXUiqVaGltpdmMnz/wAD/YbbcavrK6\n8Z0ojv8WOkTRqZBzIIrja4GbQ+cIaVj//jz697/z3sKFLFyyhHtefpnXFyxgaL9+3D5nDgC3PPcc\nbyxYsNx9B/buzbd22omtf/UrhlxwAb179OCLm2/OkH796NurF7tcdhl7DRnC3Pfeo+Q9IwYOrPXL\ny7u7ozi+NHQIgW6hA8inxgPDgKGBcwQxpF8/Ttl5Zw64+mp6de/ONhtsQLMZF+2/P9+7807OmzaN\nvYcMoXtz83L3/WDRIu6YM4fZp55K1KMH35g0ietnz+brw4dzzt57f3q7r193Hb/ed1/OnzaNZ+fP\nZ/fNNuMbo0bV8mXm0fvAsaFDSEoj5JzIPjCyP+knpApp3MiRTD3xRO485hj69OjB5n37smW/ftw8\nbhxTTzyRg7femk3XWWe5+z04dy6D1lmH9dZai+7Nzew3bBgzXnttmdvcPmcOIwYM4OMlS3jlvfe4\n6tBD+Z/nn2fhkiW1enl5tBQ4NIrj10MHkZQKOUeiOH4JOJyCnhj17Y/SdZde++ADbn3hBQ7eZptP\nf1YqlThv2jSOHj16ufttFEU8+frrLFyyBO89U195hS379fv0+pbWVi597DFOGTOGRUuXYmZAOu+8\npLXQ5w04OYrje0OHkH/RlEXORHF8V+LcD4BzQ2eptXGTJvHewoV0a27m/K9+lT49e3LpY49xxYwZ\nAOw3bBhHbbcdAPMWLODkKVO44aijGL3RRnxtq63Y9bLL6NbUxDYDBjC+3VTE5TNmcPjw4fRaYw22\nXn99FrW0sNMll/DlLbagT8+eQV5rDlyseeP80XHIOZU4dy1wROgc0pDuBvaJ4rjQbw/ySFMW+XUc\n6ZoCIpU0h3TeWGWcQxoh51ji3MbAk0D/0FmkIbwH7KDjjfNLI+Qci+L4NWBf4MPQWaTutQAHqYzz\nTYWcc9n5zL4KLAydRepWCTg6iuMHQweRlVMh14Eojh8C/g2dtVpWX1sZXxs6iKyaCrlORHF8D3AI\n6YknRbrCA8dFcfyn0EGka1TIdSSK41uBg9BIWVbNAydEcXxl6CDSdTrKog4lzu1FuhjR8gsDi0Ar\n6VKa14QOIqtHhVynEue+BEwBeoXOIrmyBDgsiuNCrx5Yr1TIdSxxbgfgFmBA6CySCwuBA6M4/kvo\nIPLZaA65jkVxPAPYHpgZOosE9xqwi8q4vqmQ61wUx28AY4FJobNIMA8Bo6M41n/MdU6F3ACiOF4E\nHAacRbp3XYrjEuBLURz/M3QQKZ/mkBtM4twhwB+Bwq4rWRBLgAlRHP8hdBCpHBVyA0qcGwX8D7Bh\n6CxSFfNId949FjqIVJamLBpQNpc4Crg1dBapuMeAUSrjxqQRcoNLnBsP/BqIAkeR8rQAvwDOjuJY\nH59vUCrkAkic2wj4A7Bn6CzymcwEjoni+K+hg0h1qZALJHHuROB84HOhs0iXLCY9cuY8neGjGFTI\nBZM4tylwJbBr6CyyUtOBY6M4nhM6iNSOduoVTBTHrwC7A6cAHwSOI8tbCJwKjFUZF49GyAWWOLcu\ncCbwLWDNwHGKzgM3At+P4nhu6DAShgpZSJwbBPwMOBK9awrhTuCHURw/FTqIhKVClk8lzg0HzgW+\nEjpLQTwEnBnF8cOhg0g+qJBlOdlay/8JjAydpUHNIh0R3xU6iOSLClk6lThnpCdWPR3YOXCcRvEC\n8BNgchTH+ocny1Ehyyolzo0GTiM9yWr3wHHqTQm4HbgIuEdFLCujQpYuS5zbEDgWOA7YOHCcvHsH\nuAq4JDvUUGSVVMiy2hLnmoG9gROzP5vDJsqNEnA36cfUp2jNCVldKmQpS+JcP2A/YH/gyxRvHeYS\n8DjpynpXR3H8euA8UsdUyFIxiXO9SA+Z2x/YF+gbNlHVfAD8hXRu+M4ojt8JnEcahApZqiKb1tiZ\n9EiNfYAtwyYq2/OkBXw78EgUx0sD55EGpEKWmkicWwcYnV22zy4bBQ21Ym8DT7W7PB7F8atBE0kh\nqJAlmMS5DVi2oDcnPe1UrxrGeIVly/fp7EzeIjWnQpbcSZzrQ1rMG5KOojdsd9mAdMfhGu0u3Ul3\nrrVmf5ZIV017q91lfmffR3G8uFavS2RVVMgiIjmhlb1ERHJChSwikhMqZBGRnFAhi4jkhApZRCQn\n6qaQzcyb2QXtvj/DzM6q0GOfZGbjVnGb8WZ20Qquu8PM+pjZYDN7dgW3ucLMtsq+PrPDddM/a/bs\n/t3NbFb29UflPFa5VvZ76sJ97zCzPpXOJFIv6qaQgcXAgWa2XqUf2Hv/O+/9n8q4/z7e+5Wewdl7\nf5z3/vns2zM7XLfTZ33uzM7AI2U+RnBd+T2KNLJ6KuSlwO9JF0pfRjYyvd/M/mpm95nZJtnPrzKz\nC81supnNNbODO3tgMzvLzM7Ivn7QzM41sxlm9pKZjW1304FmdpeZvWxm/9nu/q+2+4+im5lda2Yv\nmNmNZtar3eOONrNzgJ5m9rSZXZtd91H2527Z7W40sznZ41h23T7Zz2Zmr+m2drn2Ij1RZvvXtLLH\n2j77nczOXufaZtbDzK40s2fM7Ckz2z277Xgzu2kFr3svM5uVPc59nfxer2r/O2/3OgeY2bTsd/Bs\n2++47fdoZueY2cQV/P18x8yeyP6uXWd/nyL1qp4KGeBi4Egzizr8/LfAH7332wLXAhe2u24A6Qhy\nX+CcLj5PN+/9DsCpQNzu5yOArwPbAF83s84WaR8CXOK9HwYsACa0v9J7/31gkfd+hPf+yE7uv132\nvFsBnwfGmFkP4DJgb+/9KKBfh/vsDjzYxcdaA7geOMV7PxzYA1gETEzj+W2Aw4E/Zs/b6es2s37A\n5cBB2eMc0snzr8gRwF+89yOA4cDTHa6/Hji03feHAteb2Z7AFsAOWaZRZrbLajyvSK7VVSF77xcA\nfwJO7nDVjsB12ddXs+w54G7x3pey6YL1u/hUN2V/zgQGt/v5fd77xHv/CenqX4M6ue9r3vu26YNr\nWP3z0c3w3r/uvS+RFtVgYCgw13vfduaJP7fd2Mw2BN7z3i/s4mMNAeZ575+A9HfqvV+a5bwm+9kc\n4O/8a4W2zl73/wWmtWXy3r+3Gq/xCeDobB/ANt77D9tf6b1/CuhvZgPNbDjwvvf+NWDP7PIU6YlC\nh5IWtEhDqKtCzvya9DRCa3Xx9u3XKmh7y3529na548is431agW4reKyO17Xp+Fn01f1seleeo729\nSNfmrcRjVSpTm6Vk25iZNZGuO4H3fhqwC/AGcNUKdqjeABxMOjK/PvuZAb/I3l2M8N5v7r3/w+q+\nGJG8qrtCzkZik0hLuc104LDs6yOBh1bxGD9s+0ddhYibmNmO2ddHAA93cpsWM1udk4W+CHzezAZn\n33+93XXLzR934bEGmNn2ANn8cTfS39mR2c+2BDbJbrsijwG7mNmm2X3W7eQ2rwKjsq+/RnaCVDMb\nBLzlvb8cuAIY2cl9ryf9Oz2YtJwh/Y/nGDP7XPY4G5pZ/y68ZpG6UHeFnLkAaH+0xbdJ3wL/Ffh3\n4JQgqVIvAhPN7AVgHeDSTm7ze+CvbTv1VsV7v4h0LvouM5sJfAgkZtYMbJ5NMXSJ934JaaH/1sxm\nA/cAPYBLgCYze4a0DMd771e4Epr3/m3gBOCm7HGu7+RmlwO7ZtfvCHyc/Xw3YLaZPZVl+U0nj/8c\nsDbwhvd+Xvazu0mnph7Nct6Y3UakIWi1tzphZp/z3n+UHSlxMfAy6VzsUd77k8KmE5FKUCHXCTM7\nDfgG6TzsU8DxK9iRJyJ1SoUsIpIT9TqHLCLScFTIIiI5oUIWEckJFbKISE6okEVEcuJ/AT/3h9Jo\nOARCAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fa157219110>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "yes = 0\n",
    "no = 0\n",
    "for index, row in df_train.iterrows():\n",
    "    if row['score'] >39:\n",
    "        yes += row['score_counts']\n",
    "    else:\n",
    "        no += row['score_counts']\n",
    "df_train_yes = yes/(yes+no)*100\n",
    "print yes\n",
    "df_train_no = no/(yes+no)*100\n",
    "\n",
    "yes = 0\n",
    "no = 0\n",
    "for index, row in df_test.iterrows():\n",
    "    if row['score'] >39:\n",
    "        yes += row['score_counts']\n",
    "    else:\n",
    "        no += row['score_counts']\n",
    "df_test_yes = yes/(yes+no)*100\n",
    "print yes\n",
    "df_test_no = no/(yes+no)*100\n",
    "\n",
    "labels = 'Non-inhibiting/Inconclusive', 'Inhibiting'\n",
    "sizes = [df_train_no, df_train_yes]\n",
    "colors=['lightcoral', 'gold']\n",
    "fig1, ax1 = plt.subplots()\n",
    "ax1.pie(sizes, labels=labels, autopct='%1.1f%%', colors=colors, startangle=90)\n",
    "ax1.axis('equal')  # Equal aspect ratio ensures that pie is drawn as a circle.\n",
    "plt.savefig('./plots/train_score_counts_pie.png', bbox_inches='tight')\n",
    "\n",
    "labels = 'Non-inhibiting/Inconclusive', 'Inhibiting'\n",
    "sizes = [df_test_no, df_test_yes]\n",
    "explode = (0, 0)  # only \"explode\" the 2nd slice (i.e. 'Hogs')\n",
    "colors=['lightcoral', 'gold']\n",
    "fig1, ax1 = plt.subplots()\n",
    "ax1.pie(sizes,labels=labels, autopct='%1.1f%%', colors=colors,startangle=90)\n",
    "ax1.axis('equal')  # Equal aspect ratio ensures that pie is drawn as a circle.\n",
    "plt.savefig('./plots/test_score_counts_pie.png', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "n_features = 50\n",
    "x_train = pd.read_csv('https://s3-us-west-2.amazonaws.com/pphilip-usp-inhibition/'\n",
    "                      'data/x_train_postprocessing_rfr_%d.csv' % n_features)\n",
    "x_test = pd.read_csv('https://s3-us-west-2.amazonaws.com/pphilip-usp-inhibition/'\n",
    "                     'data/x_test_postprocessing_rfr_%d.csv' % n_features)\n",
    "y_train = pd.read_csv('https://s3-us-west-2.amazonaws.com/pphilip-usp-inhibition/'\n",
    "                      'data/y_train_postprocessing.csv')\n",
    "y_test = pd.read_csv('https://s3-us-west-2.amazonaws.com/pphilip-usp-inhibition/'\n",
    "                     'data/y_test_postprocessing.csv')\n",
    "x_train.drop(x_train.columns[0], axis=1, inplace=True)\n",
    "x_test.drop(x_test.columns[0], axis=1, inplace=True)\n",
    "y_train.drop(y_train.columns[0], axis=1, inplace=True)\n",
    "y_test.drop(y_test.columns[0], axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "for column in x_train:\n",
    "    plt.scatter(x_train[column], y_train, label='Training set', color='blue')\n",
    "    plt.scatter(x_test[column], y_test, label='Test set', color='red')\n",
    "    plt.axhspan(0, 40, facecolor='orange', alpha=0.2)\n",
    "    plt.axhspan(40, 100, facecolor='violet', alpha=0.2)\n",
    "    plt.title('%s effect on inhibition activity score' % x_train[column].name)\n",
    "    plt.xlabel('%s' % x_train[column].name)\n",
    "    plt.ylabel('Activity score')\n",
    "    plt.savefig('./plots/feature_plots/%s.png' % x_train[column].name, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def plot_results():\n",
    "    \"\"\"\n",
    "    Plots the performance metrics, comparing the different methods over the number of features used.\n",
    "    :return: \n",
    "    \"\"\"\n",
    "    n_features = [25, 50, 75, 100, 150, 200, 250]\n",
    "    mean_abs, mean_sq, median_abs, r2, exp_var_score = ({} for i in range(5))\n",
    "    for n in n_features:\n",
    "        with open('./trained_networks/lr_%d_data.pkl' % n, 'rb') as result:\n",
    "            lr_clf = pickle.load(result)\n",
    "            mean_abs[\"lr{0}\".format(n)] = pickle.load(result)\n",
    "            mean_sq[\"lr{0}\".format(n)] = pickle.load(result)\n",
    "            median_abs[\"lr{0}\".format(n)] = pickle.load(result)\n",
    "            r2[\"lr{0}\".format(n)] = pickle.load(result)\n",
    "            exp_var_score[\"lr{0}\".format(n)] = pickle.load(result)\n",
    "            lr_y_pred = pickle.load(result)\n",
    "\n",
    "        with open('./trained_networks/nn_%d_data.pkl' % n, 'rb') as result:\n",
    "            nn_clf = pickle.load(result)\n",
    "            nn_net = pickle.load(result)\n",
    "            mean_abs[\"nn{0}\".format(n)] = pickle.load(result)\n",
    "            mean_sq[\"nn{0}\".format(n)] = pickle.load(result)\n",
    "            median_abs[\"nn{0}\".format(n)] = pickle.load(result)\n",
    "            r2[\"nn{0}\".format(n)] = pickle.load(result)\n",
    "            exp_var_score[\"nn{0}\".format(n)] = pickle.load(result)\n",
    "            nn_y_pred = pickle.load(result)\n",
    "\n",
    "        with open('./trained_networks/dt_%d_data.pkl' % n, 'rb') as result:\n",
    "            dt_clf = pickle.load(result)\n",
    "            mean_abs[\"dt{0}\".format(n)] = pickle.load(result)\n",
    "            mean_sq[\"dt{0}\".format(n)] = pickle.load(result)\n",
    "            median_abs[\"dt{0}\".format(n)] = pickle.load(result)\n",
    "            r2[\"dt{0}\".format(n)] = pickle.load(result)\n",
    "            exp_var_score[\"dt{0}\".format(n)] = pickle.load(result)\n",
    "            dt_y_pred = pickle.load(result)\n",
    "\n",
    "        with open('./trained_networks/rr_%d_data.pkl' % n, 'rb') as result:\n",
    "            rr_clf = pickle.load(result)\n",
    "            mean_abs[\"rr{0}\".format(n)] = pickle.load(result)\n",
    "            mean_sq[\"rr{0}\".format(n)] = pickle.load(result)\n",
    "            median_abs[\"rr{0}\".format(n)] = pickle.load(result)\n",
    "            r2[\"rr{0}\".format(n)] = pickle.load(result)\n",
    "            exp_var_score[\"rr{0}\".format(n)] = pickle.load(result)\n",
    "            rr_ridge_alpha = pickle.load(result)\n",
    "            rr_y_pred = pickle.load(result)\n",
    "\n",
    "        with open('./trained_networks/brr_%d_data.pkl' % n, 'rb') as result:\n",
    "            brr_clf = pickle.load(result)\n",
    "            mean_abs[\"brr{0}\".format(n)] = pickle.load(result)\n",
    "            mean_sq[\"brr{0}\".format(n)] = pickle.load(result)\n",
    "            median_abs[\"brr{0}\".format(n)] = pickle.load(result)\n",
    "            r2[\"brr{0}\".format(n)] = pickle.load(result)\n",
    "            exp_var_score[\"brr{0}\".format(n)] = pickle.load(result)\n",
    "            brr_y_pred = pickle.load(result)\n",
    "\n",
    "        with open('./trained_networks/lasso_%d_data.pkl' % n, 'rb') as result:\n",
    "            lasso_clf = pickle.load(result)\n",
    "            mean_abs[\"lasso{0}\".format(n)] = pickle.load(result)\n",
    "            mean_sq[\"lasso{0}\".format(n)] = pickle.load(result)\n",
    "            median_abs[\"lasso{0}\".format(n)] = pickle.load(result)\n",
    "            r2[\"lasso{0}\".format(n)] = pickle.load(result)\n",
    "            exp_var_score[\"lasso{0}\".format(n)] = pickle.load(result)\n",
    "            lasso_y_pred = pickle.load(result)\n",
    "    \n",
    "    return mean_abs, mean_sq, median_abs, r2, exp_var_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "ename": "IOError",
     "evalue": "[Errno 2] No such file or directory: './trained_networks/nn_25_data.pkl'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mIOError\u001b[0m                                   Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-7-89963447c0e0>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mmean_abs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmean_sq\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmedian_abs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mr2\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mexp_var_score\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mplot_results\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      2\u001b[0m \u001b[0mscore_list\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0mmean_abs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmean_sq\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmedian_abs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mr2\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mexp_var_score\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      3\u001b[0m \u001b[0mlol\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mlambda\u001b[0m \u001b[0mlst\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msz\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0mlst\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m+\u001b[0m\u001b[0msz\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mi\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlst\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msz\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      4\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mi\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mscore_list\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m     \u001b[0mlists\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0msorted\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mitems\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;31m# sorted by key, return a list of tuples\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m<ipython-input-6-8d00f1456095>\u001b[0m in \u001b[0;36mplot_results\u001b[0;34m()\u001b[0m\n\u001b[1;32m     16\u001b[0m             \u001b[0mlr_y_pred\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mpickle\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mload\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mresult\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     17\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 18\u001b[0;31m         \u001b[0;32mwith\u001b[0m \u001b[0mopen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'./trained_networks/nn_%d_data.pkl'\u001b[0m \u001b[0;34m%\u001b[0m \u001b[0mn\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'rb'\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0mresult\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     19\u001b[0m             \u001b[0mnn_clf\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mpickle\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mload\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mresult\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     20\u001b[0m             \u001b[0mnn_net\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mpickle\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mload\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mresult\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mIOError\u001b[0m: [Errno 2] No such file or directory: './trained_networks/nn_25_data.pkl'"
     ]
    }
   ],
   "source": [
    "mean_abs, mean_sq, median_abs, r2, exp_var_score = plot_results()\n",
    "score_list = [mean_abs, mean_sq, median_abs, r2, exp_var_score]\n",
    "lol = lambda lst, sz: [lst[i:i+sz] for i in range(0, len(lst), sz)]\n",
    "for i in score_list:\n",
    "    lists = sorted(i.items()) # sorted by key, return a list of tuples\n",
    "    for j in lol(lists, 7):\n",
    "        y = [k[1] for k in j[:][:]]\n",
    "        x = [100, 150, 200, 25, 250, 50, 75]\n",
    "        plt.scatter(x, y)\n",
    "    \n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "all_scores = pd.read_csv('./data/all_feature_scores_rfr.csv')\n",
    "all_scores.columns = ['feature_ranking', 'feature_score']\n",
    "all_scores['cumulative_score'] = list(np.cumsum(all_scores['feature_score']))\n",
    "select_scores = pd.read_csv('./data/select_feature_scores_rfr.csv')\n",
    "select_scores.columns = ['feature_ranking', 'feature_score']\n",
    "select_scores['cumulative_score'] = list(np.cumsum(select_scores['feature_score']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgkAAAELCAYAAABTQhVuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f23d7bc6990>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure()\n",
    "ax = plt.subplot(111)\n",
    "ax.plot(all_scores['feature_ranking'], all_scores['feature_score'], label='All samples')\n",
    "ax.plot(select_scores['feature_ranking'], select_scores['feature_score'], label='Active samples')\n",
    "plt.xlabel('Feature rankings in decreasing order')\n",
    "plt.ylabel('Feature importance')\n",
    "chartBox = ax.get_position()\n",
    "ax.set_position([chartBox.x0, chartBox.y0, chartBox.width, chartBox.height])\n",
    "ax.legend(loc='upper center', bbox_to_anchor=(1.20, 1.02), ncol=1)\n",
    "plt.savefig('./plots/rfr_feature_importance_ranking.png', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfwAAAEKCAYAAAD3mecXAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f23d80c9c90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure()\n",
    "ax = plt.subplot(111)\n",
    "ax.plot(all_scores['feature_ranking'], all_scores['cumulative_score'], label='All samples')\n",
    "ax.plot(select_scores['feature_ranking'], select_scores['cumulative_score'], label='Active samples')\n",
    "plt.xlabel('Number of features')\n",
    "chartBox = ax.get_position()\n",
    "ax.set_position([chartBox.x0, chartBox.y0, chartBox.width, chartBox.height])\n",
    "ax.legend(loc='upper center', bbox_to_anchor=(1.20, 1.02), ncol=1)\n",
    "plt.ylabel('Cumulative feature importance')\n",
    "plt.savefig('./plots/rfr_feature_cumulative_importance.png', bbox_inches='tight')"
   ]
  },
  {
   "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": "2.7.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}