FeatureAnalysis.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import sys\n",
    "#sys.path.append('C:\\\\Jacob\\\\pip')\n",
    "import time\n",
    "import os\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "import matplotlib\n",
    "matplotlib.use('TKAgg')\n",
    "from matplotlib import pyplot\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.metrics import auc, roc_curve\n",
    "from sklearn import metrics, svm\n",
    "from sklearn.metrics import precision_recall_fscore_support\n",
    "import pandas as pd\n",
    "from sklearn import preprocessing\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "import ModelsFinal as m\n",
    "from sklearn.utils import shuffle\n",
    "from tensorflow.keras.callbacks import TensorBoard\n",
    "from sklearn.metrics import f1_score, precision_score, recall_score, confusion_matrix\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def dataProc(X1,y1):\n",
    "    X1 = pd.DataFrame(X1)\n",
    "    X1 = X1.fillna(X1.mean())\n",
    "    X1=X1.to_numpy()\n",
    "    y1 = y1.to_numpy()\n",
    "    \n",
    "    \n",
    "    min_max_scaler = preprocessing.MinMaxScaler()\n",
    "    X1 = min_max_scaler.fit_transform(X1)\n",
    "    \n",
    "    X1, y1 = shuffle(X1, y1, random_state=0)\n",
    "    return X1,y1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1\n",
      "2\n",
      "3\n",
      "4\n",
      "5\n",
      "6\n",
      "7\n",
      "8\n",
      "9\n",
      "10\n"
     ]
    }
   ],
   "source": [
    "sysName = \"IEEE30\"\n",
    "\n",
    "X=[]\n",
    "Xt = []\n",
    "y = []\n",
    "yt = []\n",
    "\n",
    "for i in range(1,11,1):\n",
    "    print (i)\n",
    "    X.append(pd.read_csv(\"Data/{}/{}Data_10k_{}Sparsity.csv\".format(sysName, sysName, i/10 ), header=None).sample(100,random_state=1))\n",
    "    Xt.append(pd.read_csv(\"Data/{}/{}Data_2k_{}Sparsity.csv\".format(sysName, sysName, i/10 ), header=None).sample(100,random_state=1))\n",
    "    y.append(pd.read_csv(\"Data/{}/{}Labels_10k_{}Sparsity.csv\".format(sysName, sysName, i/10 ), header=None).sample(100,random_state=1))\n",
    "    yt.append(pd.read_csv(\"Data/{}/{}Labels_2k_{}Sparsity.csv\".format(sysName, sysName, i/10 ), header=None).sample(100,random_state=1))\n",
    "    \n",
    "    X[i-1],y[i-1] = dataProc(X[i-1],y[i-1])\n",
    "    Xt[i-1],yt[i-1] = dataProc(Xt[i-1],yt[i-1])\n",
    "\n",
    "    \n",
    "X_train = np.concatenate((X[0],X[1]), axis=0)\n",
    "X_test = np.concatenate((Xt[0],Xt[1]), axis=0)\n",
    "\n",
    "y_train = np.concatenate((y[0],y[1]), axis=0)\n",
    "y_test = np.concatenate((yt[0],yt[1]), axis=0)\n",
    "\n",
    "numFeatures = len(X_train[0])\n",
    "\n",
    "for i in range(2,10,1):\n",
    "    X_train = np.concatenate((X_train,X[i]), axis=0) \n",
    "    X_test = np.concatenate((X_test,Xt[i]), axis=0) \n",
    "    y_train = np.concatenate((y_train,y[i]), axis=0)\n",
    "    y_test = np.concatenate((y_test,yt[i]), axis=0)\n",
    "\n",
    "X_train, y_train = shuffle(X_train, y_train, random_state=0)\n",
    "X_test, y_test = shuffle(X_test, y_test, random_state=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2000, 170)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X = np.concatenate( (X_train,X_test), axis = 0)\n",
    "y = np.concatenate( (y_train,y_test), axis = 0)\n",
    "\n",
    "X.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2000,)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y = y.ravel()\n",
    "y.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "import featureProcessing as fp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "Info gain ratio done 0.8249704837799072   0.05941755156774004\n",
      "1\n",
      "Info gain ratio done 0.8387618064880371   0.06202551664506646\n",
      "2\n",
      "Info gain ratio done 0.8489055633544922   0.06197532544968528\n",
      "3\n",
      "Info gain ratio done 0.8367798328399658   0.06208882016525825\n",
      "4\n",
      "Info gain ratio done 0.8484301567077637   0.06250843371122025\n",
      "5\n",
      "Info gain ratio done 0.856680154800415   0.06212547875429241\n",
      "6\n",
      "Info gain ratio done 0.833655595779419   0.06209757199529054\n",
      "7\n",
      "Info gain ratio done 0.8497302532196045   0.06241078865789197\n",
      "8\n",
      "Info gain ratio done 0.836796760559082   0.06230573943788927\n",
      "9\n",
      "Info gain ratio done 0.8507628440856934   0.062409619208295165\n",
      "10\n",
      "Info gain ratio done 0.8527231216430664   0.06223401857538061\n",
      "11\n",
      "Info gain ratio done 0.8504483699798584   0.06217953028811748\n",
      "12\n",
      "Info gain ratio done 0.8374569416046143   0.06296078798657077\n",
      "13\n",
      "Info gain ratio done 0.848876953125   0.06236376358825435\n",
      "14\n",
      "Info gain ratio done 0.8227839469909668   0.062185182827378206\n",
      "15\n",
      "Info gain ratio done 0.8307802677154541   0.06308107023519421\n",
      "16\n",
      "Info gain ratio done 0.8497700691223145   0.06232609486271263\n",
      "17\n",
      "Info gain ratio done 0.8378438949584961   0.06218185441403613\n",
      "18\n",
      "Info gain ratio done 0.8547153472900391   0.06242118622924568\n",
      "19\n",
      "Info gain ratio done 0.8507285118103027   0.06235631473313453\n",
      "20\n",
      "Info gain ratio done 0.8663928508758545   0.06253197064677178\n",
      "21\n",
      "Info gain ratio done 0.8503696918487549   0.0625614573100104\n",
      "22\n",
      "Info gain ratio done 0.9165525436401367   0.062373660253348465\n",
      "23\n",
      "Info gain ratio done 0.8612706661224365   0.06266841572329841\n",
      "24\n",
      "Info gain ratio done 0.8645288944244385   0.06229885839218302\n",
      "25\n",
      "Info gain ratio done 0.8483278751373291   0.06240552326592135\n",
      "26\n",
      "Info gain ratio done 0.8653173446655273   0.06243053142248086\n",
      "27\n",
      "Info gain ratio done 0.838062047958374   0.06246320212435747\n",
      "28\n",
      "Info gain ratio done 0.8490328788757324   0.06245053331687161\n",
      "29\n",
      "Info gain ratio done 0.8344745635986328   0.06237472471637547\n",
      "30\n",
      "Info gain ratio done 0.8529238700866699   0.06234101350612325\n",
      "31\n",
      "Info gain ratio done 0.8521966934204102   0.06254051969214373\n",
      "32\n",
      "Info gain ratio done 0.8519351482391357   0.06246602016710841\n",
      "33\n",
      "Info gain ratio done 0.8507556915283203   0.06233007201921887\n",
      "34\n",
      "Info gain ratio done 0.8497323989868164   0.06244638555425945\n",
      "35\n",
      "Info gain ratio done 0.8297812938690186   0.062275257229283884\n",
      "36\n",
      "Info gain ratio done 0.8363327980041504   0.06248950606362851\n",
      "37\n",
      "Info gain ratio done 0.8557479381561279   0.0623614205475416\n",
      "38\n",
      "Info gain ratio done 0.8511054515838623   0.06250152119310337\n",
      "39\n",
      "Info gain ratio done 0.8676583766937256   0.06237554012184759\n",
      "40\n",
      "Info gain ratio done 0.8693459033966064   0.062229655506209784\n",
      "41\n",
      "Info gain ratio done 0.7960929870605469   0.05929051776765834\n",
      "42\n",
      "Info gain ratio done 0.834913969039917   0.06199474767978568\n",
      "43\n",
      "Info gain ratio done 0.8420579433441162   0.06223079169071386\n",
      "44\n",
      "Info gain ratio done 0.8415937423706055   0.062140646650717085\n",
      "45\n",
      "Info gain ratio done 0.8428821563720703   0.06246290881572932\n",
      "46\n",
      "Info gain ratio done 0.844876766204834   0.06215310181502787\n",
      "47\n",
      "Info gain ratio done 0.8478710651397705   0.06234878066871598\n",
      "48\n",
      "Info gain ratio done 0.8286929130554199   0.06244742935726327\n",
      "49\n",
      "Info gain ratio done 0.8437469005584717   0.06248762459914823\n",
      "50\n",
      "Info gain ratio done 0.8437762260437012   0.062159000380090944\n",
      "51\n",
      "Info gain ratio done 0.8417518138885498   0.06227364658606776\n",
      "52\n",
      "Info gain ratio done 0.8477358818054199   0.062082734120202675\n",
      "53\n",
      "Info gain ratio done 0.8349006175994873   0.06298533134640058\n",
      "54\n",
      "Info gain ratio done 0.8518569469451904   0.06245538773081647\n",
      "55\n",
      "Info gain ratio done 0.8397846221923828   0.062311401274115726\n",
      "56\n",
      "Info gain ratio done 0.815300703048706   0.06314124467722583\n",
      "57\n",
      "Info gain ratio done 0.8534173965454102   0.06242095249727762\n",
      "58\n",
      "Info gain ratio done 0.8360123634338379   0.06209337227525513\n",
      "59\n",
      "Info gain ratio done 0.8796496391296387   0.06241630336208818\n",
      "60\n",
      "Info gain ratio done 0.842653751373291   0.0623420140602199\n",
      "61\n",
      "Info gain ratio done 0.829512357711792   0.06233678534192647\n",
      "62\n",
      "Info gain ratio done 0.827667236328125   0.062284383835969764\n",
      "63\n",
      "Info gain ratio done 0.8409264087677002   0.06229840450677243\n",
      "64\n",
      "Info gain ratio done 0.8418900966644287   0.06239413325334326\n",
      "65\n",
      "Info gain ratio done 0.8237998485565186   0.06245532730130028\n",
      "66\n",
      "Info gain ratio done 0.841933012008667   0.06240885193843271\n",
      "67\n",
      "Info gain ratio done 0.853553056716919   0.06221530074781349\n",
      "68\n",
      "Info gain ratio done 0.8267955780029297   0.06217379469304513\n",
      "69\n",
      "Info gain ratio done 0.8417859077453613   0.062391760328879775\n",
      "70\n",
      "Info gain ratio done 0.8441388607025146   0.062390588385112224\n",
      "71\n",
      "Info gain ratio done 0.8487646579742432   0.06243543524929755\n",
      "72\n",
      "Info gain ratio done 0.8539283275604248   0.06244305183184775\n",
      "73\n",
      "Info gain ratio done 0.8273570537567139   0.06242461521247661\n",
      "74\n",
      "Info gain ratio done 0.839759111404419   0.0627127414547864\n",
      "75\n",
      "Info gain ratio done 0.8178131580352783   0.062493810155539266\n",
      "76\n",
      "Info gain ratio done 0.8507301807403564   0.06220175021610732\n",
      "77\n",
      "Info gain ratio done 0.8367645740509033   0.06244888336502284\n",
      "78\n",
      "Info gain ratio done 0.8786213397979736   0.06254965403001198\n",
      "79\n",
      "Info gain ratio done 0.8467376232147217   0.06240069136664205\n",
      "80\n",
      "Info gain ratio done 0.8417510986328125   0.06236302179659693\n",
      "81\n",
      "Info gain ratio done 0.8517255783081055   0.06198314374357441\n",
      "82\n",
      "Info gain ratio done 0.8090007305145264   0.05926114639326856\n",
      "83\n",
      "Info gain ratio done 0.811556339263916   0.06071799201674497\n",
      "84\n",
      "Info gain ratio done 0.8598911762237549   0.06263098158211956\n",
      "85\n",
      "Info gain ratio done 0.8384754657745361   0.06302204156862581\n",
      "86\n",
      "Info gain ratio done 0.8228302001953125   0.06299288606833857\n",
      "87\n",
      "Info gain ratio done 0.8567171096801758   0.06323417764584219\n",
      "88\n",
      "Info gain ratio done 0.8377954959869385   0.06212899028313265\n",
      "89\n",
      "Info gain ratio done 0.831780195236206   0.062349464909561905\n",
      "90\n",
      "Info gain ratio done 0.8544220924377441   0.06231587072626143\n",
      "91\n",
      "Info gain ratio done 0.8349366188049316   0.06321918002110793\n",
      "92\n",
      "Info gain ratio done 0.8478412628173828   0.06251935198923081\n",
      "93\n",
      "Info gain ratio done 0.8203451633453369   0.062418768614682096\n",
      "94\n",
      "Info gain ratio done 0.8147914409637451   0.0616119729172405\n",
      "95\n",
      "Info gain ratio done 0.8297834396362305   0.06263208162688144\n",
      "96\n",
      "Info gain ratio done 0.8268129825592041   0.06286909049114471\n",
      "97\n",
      "Info gain ratio done 0.7973752021789551   0.0605189327887895\n",
      "98\n",
      "Info gain ratio done 0.8347804546356201   0.06256762941099756\n",
      "99\n",
      "Info gain ratio done 0.8248059749603271   0.06212707128159802\n",
      "100\n",
      "Info gain ratio done 0.8407540321350098   0.06232815203886873\n",
      "101\n",
      "Info gain ratio done 0.8198287487030029   0.06265988388130236\n",
      "102\n",
      "Info gain ratio done 0.7948954105377197   0.06106294997563333\n",
      "103\n",
      "Info gain ratio done 0.8382768630981445   0.06237863622233326\n",
      "104\n",
      "Info gain ratio done 0.8367650508880615   0.06256275616462217\n",
      "105\n",
      "Info gain ratio done 0.8238377571105957   0.06270294867845201\n",
      "106\n",
      "Info gain ratio done 0.8417515754699707   0.06282973584931771\n",
      "107\n",
      "Info gain ratio done 0.8475339412689209   0.06285086441010758\n",
      "108\n",
      "Info gain ratio done 0.8295490741729736   0.06295378323197764\n",
      "109\n",
      "Info gain ratio done 0.8364498615264893   0.06262000402526856\n",
      "110\n",
      "Info gain ratio done 0.8493573665618896   0.06298645096197998\n",
      "111\n",
      "Info gain ratio done 0.8646330833435059   0.06347824097698676\n",
      "112\n",
      "Info gain ratio done 0.8509492874145508   0.06271123503311858\n",
      "113\n",
      "Info gain ratio done 0.8454024791717529   0.06299683536672865\n",
      "114\n",
      "Info gain ratio done 0.8625123500823975   0.06285096646708986\n",
      "115\n",
      "Info gain ratio done 0.8137023448944092   0.06287535090377389\n",
      "116\n",
      "Info gain ratio done 0.8457410335540771   0.06297347055788005\n",
      "117\n",
      "Info gain ratio done 0.8364615440368652   0.06260964538912646\n",
      "118\n",
      "Info gain ratio done 0.8404178619384766   0.06229406205204901\n",
      "119\n",
      "Info gain ratio done 0.8604061603546143   0.06230796222419054\n",
      "120\n",
      "Info gain ratio done 0.8267617225646973   0.06191340310073491\n",
      "121\n",
      "Info gain ratio done 0.8135786056518555   0.062082069154369646\n",
      "122\n",
      "Info gain ratio done 0.8397886753082275   0.062254717787475995\n",
      "123\n",
      "Info gain ratio done 0.8377625942230225   0.06224892288587074\n",
      "124\n",
      "Info gain ratio done 0.8257935047149658   0.06282647880466949\n",
      "125\n",
      "Info gain ratio done 0.8228051662445068   0.06222246014535371\n",
      "126\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Info gain ratio done 0.84415602684021   0.06320403800307318\n",
      "127\n",
      "Info gain ratio done 0.8773238658905029   0.06298025371535818\n",
      "128\n",
      "Info gain ratio done 0.8798768520355225   0.0630683377086471\n",
      "129\n",
      "Info gain ratio done 0.8108618259429932   0.06082528596796552\n",
      "130\n",
      "Info gain ratio done 0.8437178134918213   0.06275376754604917\n",
      "131\n",
      "Info gain ratio done 0.8373074531555176   0.06215512505298601\n",
      "132\n",
      "Info gain ratio done 0.8529665470123291   0.06363786120616154\n",
      "133\n",
      "Info gain ratio done 0.8606960773468018   0.06247637872588596\n",
      "134\n",
      "Info gain ratio done 0.8224778175354004   0.0624547160485229\n",
      "135\n",
      "Info gain ratio done 0.810537576675415   0.061738174020527045\n",
      "136\n",
      "Info gain ratio done 0.8259620666503906   0.06148092019630816\n",
      "137\n",
      "Info gain ratio done 0.8287878036499023   0.06264571342693247\n",
      "138\n",
      "Info gain ratio done 0.8317780494689941   0.0628712917132507\n",
      "139\n",
      "Info gain ratio done 0.8577473163604736   0.06277084177527248\n",
      "140\n",
      "Info gain ratio done 0.8537192344665527   0.06283319817400829\n",
      "141\n",
      "Info gain ratio done 0.8487353324890137   0.06254933861635452\n",
      "142\n",
      "Info gain ratio done 0.8534133434295654   0.06257165879563015\n",
      "143\n",
      "Info gain ratio done 0.8098363876342773   0.06066958868828324\n",
      "144\n",
      "Info gain ratio done 0.8633816242218018   0.06254820228701213\n",
      "145\n",
      "Info gain ratio done 0.834768533706665   0.06299059351189137\n",
      "146\n",
      "Info gain ratio done 0.8408889770507812   0.06294365784593313\n",
      "147\n",
      "Info gain ratio done 0.8335506916046143   0.06293200552420619\n",
      "148\n",
      "Info gain ratio done 0.8257954120635986   0.06258836866404065\n",
      "149\n",
      "Info gain ratio done 0.8536684513092041   0.06275979783080027\n",
      "150\n",
      "Info gain ratio done 0.8469152450561523   0.06261915298669415\n",
      "151\n",
      "Info gain ratio done 0.8319933414459229   0.06266135925388759\n",
      "152\n",
      "Info gain ratio done 0.8465731143951416   0.06274539269093038\n",
      "153\n",
      "Info gain ratio done 0.8407518863677979   0.062330687915142104\n",
      "154\n",
      "Info gain ratio done 0.8409600257873535   0.0627667993602708\n",
      "155\n",
      "Info gain ratio done 0.8306055068969727   0.06298055126709025\n",
      "156\n",
      "Info gain ratio done 0.8389999866485596   0.06280735938371317\n",
      "157\n",
      "Info gain ratio done 0.856400728225708   0.06325444454478707\n",
      "158\n",
      "Info gain ratio done 0.8337728977203369   0.06249700742104343\n",
      "159\n",
      "Info gain ratio done 0.8407542705535889   0.06197577036196048\n",
      "160\n",
      "Info gain ratio done 0.8307924270629883   0.06219044663346567\n",
      "161\n",
      "Info gain ratio done 0.8357675075531006   0.06205465957509934\n",
      "162\n",
      "Info gain ratio done 0.3271353244781494   0.06498336146517739\n",
      "163\n",
      "Info gain ratio done 0.846747636795044   0.06259433590975018\n",
      "164\n",
      "Info gain ratio done 0.8367745876312256   0.06237448695419405\n",
      "165\n",
      "Info gain ratio done 0.8043563365936279   0.06160642926764145\n",
      "166\n",
      "Info gain ratio done 0.7639596462249756   0.05962843559880685\n",
      "167\n",
      "Info gain ratio done 0.8327839374542236   0.06318552920457271\n",
      "168\n",
      "Info gain ratio done 0.8468222618103027   0.06248659423482028\n",
      "169\n",
      "Info gain ratio done 0.8574354648590088   0.06271832904554148\n"
     ]
    }
   ],
   "source": [
    "infoGainRatio = []\n",
    "\n",
    "for i in range(len(X[0])):\n",
    "    print(i)\n",
    "\n",
    "    time1= time.time()\n",
    "    g2 = fp.info_gain_ratio(y,X[:,i])\n",
    "    time2=time.time()\n",
    "    print(\"Info gain ratio done\", time2-time1, \" \", g2)\n",
    "\n",
    "    infoGainRatio.append(g2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "IGdf = pd.DataFrame(infoGainRatio)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0.059418</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>0.062026</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>0.061975</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>0.062089</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>0.062508</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          0\n",
       "0  0.059418\n",
       "1  0.062026\n",
       "2  0.061975\n",
       "3  0.062089\n",
       "4  0.062508"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "IGdf.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "export_csv = IGdf.to_csv(r'IEEE30_IGratio.csv', index = None, header=True)"
   ]
  },
  {
   "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
}