Validation3_Random forests.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "import pickle\n",
    "import copy\n",
    "import numpy as np\n",
    "import random\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib\n",
    "import tensorflow as tf\n",
    "import pdb\n",
    "import scipy.io as sio\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from pylab import rcParams\n",
    "import math\n",
    "import time\n",
    "\n",
    "##Input Data\n",
    "\n",
    "F1=open(\"E:/holiday_2_train_2016\",'rb')\n",
    "#F1=open(\"E:/revisiondata/trainnorm_2016\",'rb')\n",
    "train=pickle.load(F1)\n",
    "F1.close()\n",
    "\n",
    "F2=open(\"E:/holiday_2_test_2016\",'rb')\n",
    "#F2=open(\"E:/revisiondata/testnorm_2016\",'rb')\n",
    "test=pickle.load(F2)\n",
    "F2.close()\n",
    "\n",
    "F4=open(\"E:/holiday_2_test_2017\",'rb')\n",
    "#F4=open(\"E:/revisiondata/testnorm_2017\",'rb')\n",
    "data2017=pickle.load(F4)\n",
    "F4.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "##Data reshape\n",
    "\n",
    "X1=[]\n",
    "Y1=[]\n",
    "for i in range(len(train)):\n",
    "    if(train[i][1][0]==1):\n",
    "        Y1.append(0)\n",
    "    else:\n",
    "        Y1.append(1)\n",
    "    #tmp=copy.deepcopy(np.reshape(np.array(train[i][0]),(30*30*48)))\n",
    "    tmp=copy.deepcopy(np.sum(np.reshape(np.array(train[i][0]),(900,48)),axis=0))\n",
    "    #tmp=copy.deepcopy(np.sum(np.reshape(np.array(train[i][0]),(900,48)),axis=1))\n",
    "    X1.append(tmp)\n",
    "\n",
    "X2=[]\n",
    "Y2=[]\n",
    "for i in range(len(test)):\n",
    "    if(test[i][1][0]==1):\n",
    "        Y2.append(0)\n",
    "    else:\n",
    "        Y2.append(1)\n",
    "    #tmp=copy.deepcopy(np.reshape(np.array(test[i][0]),(30*30*48)))\n",
    "    tmp=copy.deepcopy(np.sum(np.reshape(np.array(test[i][0]),(900,48)),axis=0))\n",
    "    #tmp=copy.deepcopy(np.sum(np.reshape(np.array(test[i][0]),(900,48)),axis=1))\n",
    "    X2.append(tmp)\n",
    "\n",
    "\n",
    "\n",
    "X4=[]\n",
    "Y4=[]\n",
    "for i in range(len(data2017)):\n",
    "    if(data2017[i][1][0]==1):\n",
    "        Y4.append(0)\n",
    "    else:\n",
    "        Y4.append(1)\n",
    "    #tmp=copy.deepcopy(np.reshape(np.array(data2017[i][0]),(30*30*48)))\n",
    "    tmp=copy.deepcopy(np.sum(np.reshape(np.array(data2017[i][0]),(900,48)),axis=0))\n",
    "    #tmp=copy.deepcopy(np.sum(np.reshape(np.array(data2017[i][0]),(900,48)),axis=1))\n",
    "    X4.append(tmp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "##Validation3, Random forests\n",
    "\n",
    "importance=[]\n",
    "accuracy=[]\n",
    "repeat=50\n",
    "for n in range(repeat):\n",
    "    clf=RandomForestClassifier(n_estimators=10).fit(X1,Y1)\n",
    "    tmp_i=clf.feature_importances_\n",
    "    importance.append(tmp_i)\n",
    "    \n",
    "    tmp_p=[]\n",
    "    result=clf.predict(X2)\n",
    "    p=0.0\n",
    "    for i in range(len(result)):\n",
    "        if(result[i]==Y2[i]):\n",
    "            p+=1\n",
    "    p=p/len(result)\n",
    "    tmp_p.append(p)\n",
    "\n",
    "    result=clf.predict(X1)\n",
    "    p=0.0\n",
    "    for i in range(len(result)):\n",
    "        if(result[i]==Y1[i]):\n",
    "            p+=1\n",
    "    p=p/len(result)\n",
    "    tmp_p.append(p)\n",
    "\n",
    "\n",
    "    result=clf.predict(X4)\n",
    "    p=0.0\n",
    "    for i in range(len(result)):\n",
    "        if(result[i]==Y4[i]):\n",
    "            p+=1\n",
    "    p=p/len(result)\n",
    "    tmp_p.append(p)\n",
    "    \n",
    "    accuracy.append(tmp_p)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9292\n",
      "0.99796875\n",
      "0.9472\n",
      "-------------------\n",
      "0.0457210761384,\n",
      "0.0283650445457,\n",
      "0.0107311458064,\n",
      "0.00656178634432,\n",
      "0.00434454608758,\n",
      "0.0087033297948,\n",
      "0.0412202893111,\n",
      "0.0344351643627,\n",
      "0.064571300583,\n",
      "0.0605586091072,\n",
      "0.0262654577376,\n",
      "0.0135749231015,\n",
      "0.0212477827929,\n",
      "0.112319978273,\n",
      "0.0304626087842,\n",
      "0.0163282765813,\n",
      "0.0708848285152,\n",
      "0.117258184074,\n",
      "0.00662649509099,\n",
      "0.00407362730181,\n",
      "0.00815799101187,\n",
      "0.0107991074409,\n",
      "0.0133637756667,\n",
      "0.0334670422939,\n",
      "0.0255956340755,\n",
      "0.0162663895975,\n",
      "0.00407138091215,\n",
      "0.0061969705118,\n",
      "0.0171585503056,\n",
      "0.0103016985317,\n",
      "0.00442433692366,\n",
      "0.00394349795112,\n",
      "0.0028396309038,\n",
      "0.00795873837872,\n",
      "0.0266272265042,\n",
      "0.017954019676,\n",
      "0.00640700105777,\n",
      "0.00385167756043,\n",
      "0.00407761155852,\n",
      "0.00613219244644,\n",
      "0.00521134234334,\n",
      "0.00320499796612,\n",
      "0.00425262565566,\n",
      "0.00361190569291,\n",
      "0.00331704879887,\n",
      "0.00491818880152,\n",
      "0.00665889959944,\n",
      "0.0149760634998,\n"
     ]
    }
   ],
   "source": [
    "##Get the classification accuracy and element importance\n",
    "\n",
    "result1=np.mean(np.array(accuracy),axis=0)\n",
    "for i in result1:\n",
    "    print(str(i))\n",
    "print('-------------------')\n",
    "result2=np.mean(np.array(importance),axis=0)\n",
    "for i in result2:\n",
    "    print(str(i)+',')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}