DF_RF.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "0067e469-cd60-4e11-ab93-e37643635d66",
   "metadata": {},
   "source": [
    "# Decision tree and Random forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "0dcc7eb4-c17f-4601-aa1c-169062ae7858",
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import time\n",
    "from sklearn.preprocessing import RobustScaler\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.datasets import make_classification\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier, AdaBoostClassifier\n",
    "from sklearn.metrics import confusion_matrix, ConfusionMatrixDisplay, classification_report, balanced_accuracy_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "6ff7cbf2-3af0-4267-869d-cf30213df1e2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<bound method NDFrame.head of                                                    red   green    blue  \\\n",
      "data/12910/0/12910_idx5_x3001_y701_class0.png   2.8792  2.7112  2.8076   \n",
      "data/12910/0/12910_idx5_x1501_y401_class0.png   2.7392  2.7464  2.0908   \n",
      "data/12910/0/12910_idx5_x2251_y1101_class0.png  1.1296  2.6620  2.2704   \n",
      "data/12910/0/12910_idx5_x2551_y1151_class0.png  1.9812  2.8940  2.2452   \n",
      "data/12910/0/12910_idx5_x2001_y101_class0.png   2.8660  3.0032  2.5404   \n",
      "...                                                ...     ...     ...   \n",
      "data/13666/1/13666_idx5_x901_y1101_class1.png   2.1980  2.3572  2.7020   \n",
      "data/13666/1/13666_idx5_x1051_y1001_class1.png  2.6968  2.4552  2.5280   \n",
      "data/13666/1/13666_idx5_x1051_y1151_class1.png  2.4272  2.8244  2.5152   \n",
      "data/13666/1/13666_idx5_x1051_y1051_class1.png  2.5724  2.3860  2.5444   \n",
      "data/13666/1/13666_idx5_x901_y1151_class1.png   2.7212  2.7648  3.0000   \n",
      "\n",
      "                                                    gray  gray_var  edges  \\\n",
      "data/12910/0/12910_idx5_x3001_y701_class0.png   0.755730  0.003960     83   \n",
      "data/12910/0/12910_idx5_x1501_y401_class0.png   0.483340  0.009377    350   \n",
      "data/12910/0/12910_idx5_x2251_y1101_class0.png  0.799321  0.001401     14   \n",
      "data/12910/0/12910_idx5_x2551_y1151_class0.png  0.824825  0.001107      0   \n",
      "data/12910/0/12910_idx5_x2001_y101_class0.png   0.675241  0.005805    187   \n",
      "...                                                  ...       ...    ...   \n",
      "data/13666/1/13666_idx5_x901_y1101_class1.png   0.659095  0.002654     31   \n",
      "data/13666/1/13666_idx5_x1051_y1001_class1.png  0.623201  0.004284    136   \n",
      "data/13666/1/13666_idx5_x1051_y1151_class1.png  0.648590  0.002005     28   \n",
      "data/13666/1/13666_idx5_x1051_y1051_class1.png  0.632355  0.003022     70   \n",
      "data/13666/1/13666_idx5_x901_y1151_class1.png   0.655819  0.004799     97   \n",
      "\n",
      "                                                ridges  group  \n",
      "data/12910/0/12910_idx5_x3001_y701_class0.png     1292      0  \n",
      "data/12910/0/12910_idx5_x1501_y401_class0.png     1297      0  \n",
      "data/12910/0/12910_idx5_x2251_y1101_class0.png    1315      0  \n",
      "data/12910/0/12910_idx5_x2551_y1151_class0.png    1316      0  \n",
      "data/12910/0/12910_idx5_x2001_y101_class0.png     1225      0  \n",
      "...                                                ...    ...  \n",
      "data/13666/1/13666_idx5_x901_y1101_class1.png     1333      1  \n",
      "data/13666/1/13666_idx5_x1051_y1001_class1.png    1263      1  \n",
      "data/13666/1/13666_idx5_x1051_y1151_class1.png    1397      1  \n",
      "data/13666/1/13666_idx5_x1051_y1051_class1.png    1361      1  \n",
      "data/13666/1/13666_idx5_x901_y1151_class1.png     1308      1  \n",
      "\n",
      "[277523 rows x 8 columns]>\n"
     ]
    }
   ],
   "source": [
    "#Import dataframe from JSON file\n",
    "jsonStr = \"parameters_df.json\"\n",
    "df = pd.read_json(jsonStr, orient ='index')\n",
    "print(df.head)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b0df9e26-c19b-4b76-8432-b9c228365a05",
   "metadata": {},
   "source": [
    "## Splitting in train and test set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "249485ec-21db-4663-be45-fd314785887e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_train : \n",
      "                                                   red   green    blue  \\\n",
      "data/12910/0/12910_idx5_x3001_y701_class0.png   2.8792  2.7112  2.8076   \n",
      "data/12910/0/12910_idx5_x1501_y401_class0.png   2.7392  2.7464  2.0908   \n",
      "data/12910/0/12910_idx5_x2251_y1101_class0.png  1.1296  2.6620  2.2704   \n",
      "data/12910/0/12910_idx5_x2551_y1151_class0.png  1.9812  2.8940  2.2452   \n",
      "data/12910/0/12910_idx5_x2001_y101_class0.png   2.8660  3.0032  2.5404   \n",
      "\n",
      "                                                    gray  gray_var  edges  \\\n",
      "data/12910/0/12910_idx5_x3001_y701_class0.png   0.755730  0.003960     83   \n",
      "data/12910/0/12910_idx5_x1501_y401_class0.png   0.483340  0.009377    350   \n",
      "data/12910/0/12910_idx5_x2251_y1101_class0.png  0.799321  0.001401     14   \n",
      "data/12910/0/12910_idx5_x2551_y1151_class0.png  0.824825  0.001107      0   \n",
      "data/12910/0/12910_idx5_x2001_y101_class0.png   0.675241  0.005805    187   \n",
      "\n",
      "                                                ridges  \n",
      "data/12910/0/12910_idx5_x3001_y701_class0.png     1292  \n",
      "data/12910/0/12910_idx5_x1501_y401_class0.png     1297  \n",
      "data/12910/0/12910_idx5_x2251_y1101_class0.png    1315  \n",
      "data/12910/0/12910_idx5_x2551_y1151_class0.png    1316  \n",
      "data/12910/0/12910_idx5_x2001_y101_class0.png     1225  \n",
      "(194265, 7)\n",
      "\n",
      "X_val : \n",
      "                                                   red   green    blue  \\\n",
      "data/10277/0/10277_idx5_x301_y2051_class0.png   1.8916  3.3608  3.1964   \n",
      "data/10277/0/10277_idx5_x1701_y1251_class0.png  1.9904  2.1932  2.2724   \n",
      "data/10277/0/10277_idx5_x751_y551_class0.png    3.0532  2.3560  2.4284   \n",
      "data/10277/0/10277_idx5_x501_y1901_class0.png   1.9820  2.4620  2.7300   \n",
      "data/10277/0/10277_idx5_x901_y751_class0.png    2.5864  2.2684  2.3436   \n",
      "\n",
      "                                                    gray  gray_var  edges  \\\n",
      "data/10277/0/10277_idx5_x301_y2051_class0.png   0.867761  0.030824    181   \n",
      "data/10277/0/10277_idx5_x1701_y1251_class0.png  0.480924  0.015185    507   \n",
      "data/10277/0/10277_idx5_x751_y551_class0.png    0.807837  0.022346    474   \n",
      "data/10277/0/10277_idx5_x501_y1901_class0.png   0.703649  0.039559    499   \n",
      "data/10277/0/10277_idx5_x901_y751_class0.png    0.695074  0.033165    507   \n",
      "\n",
      "                                                ridges  \n",
      "data/10277/0/10277_idx5_x301_y2051_class0.png     1597  \n",
      "data/10277/0/10277_idx5_x1701_y1251_class0.png    1233  \n",
      "data/10277/0/10277_idx5_x751_y551_class0.png      1389  \n",
      "data/10277/0/10277_idx5_x501_y1901_class0.png     1327  \n",
      "data/10277/0/10277_idx5_x901_y751_class0.png      1300  \n",
      "(27753, 7)\n",
      "\n",
      "X_test : \n",
      "                                                red   green    blue      gray  \\\n",
      "data/9257/0/9257_idx5_x301_y1001_class0.png  2.5784  2.7000  2.2112  0.609239   \n",
      "data/9257/0/9257_idx5_x2351_y251_class0.png  2.4596  2.0084  2.7724  0.501843   \n",
      "data/9257/0/9257_idx5_x1101_y401_class0.png  2.3712  2.5872  2.7416  0.593688   \n",
      "data/9257/0/9257_idx5_x751_y1001_class0.png  2.8744  2.5176  2.3424  0.599057   \n",
      "data/9257/0/9257_idx5_x1201_y551_class0.png  2.7900  2.2444  2.4596  0.641368   \n",
      "\n",
      "                                             gray_var  edges  ridges  \n",
      "data/9257/0/9257_idx5_x301_y1001_class0.png  0.003757     51    1260  \n",
      "data/9257/0/9257_idx5_x2351_y251_class0.png  0.024991    479    1183  \n",
      "data/9257/0/9257_idx5_x1101_y401_class0.png  0.027011    504    1264  \n",
      "data/9257/0/9257_idx5_x751_y1001_class0.png  0.002871     33    1245  \n",
      "data/9257/0/9257_idx5_x1201_y551_class0.png  0.031924    280    1333  \n",
      "(55505, 7)\n",
      "\n",
      "y_train : \n",
      "data/12910/0/12910_idx5_x3001_y701_class0.png     0\n",
      "data/12910/0/12910_idx5_x1501_y401_class0.png     0\n",
      "data/12910/0/12910_idx5_x2251_y1101_class0.png    0\n",
      "data/12910/0/12910_idx5_x2551_y1151_class0.png    0\n",
      "data/12910/0/12910_idx5_x2001_y101_class0.png     0\n",
      "Name: group, dtype: int64\n",
      "(194265,)\n",
      "\n",
      "y_val : \n",
      "data/10277/0/10277_idx5_x301_y2051_class0.png     0\n",
      "data/10277/0/10277_idx5_x1701_y1251_class0.png    0\n",
      "data/10277/0/10277_idx5_x751_y551_class0.png      0\n",
      "data/10277/0/10277_idx5_x501_y1901_class0.png     0\n",
      "data/10277/0/10277_idx5_x901_y751_class0.png      0\n",
      "Name: group, dtype: int64\n",
      "(27753,)\n",
      "\n",
      "y_test : \n",
      "data/9257/0/9257_idx5_x301_y1001_class0.png    0\n",
      "data/9257/0/9257_idx5_x2351_y251_class0.png    0\n",
      "data/9257/0/9257_idx5_x1101_y401_class0.png    0\n",
      "data/9257/0/9257_idx5_x751_y1001_class0.png    0\n",
      "data/9257/0/9257_idx5_x1201_y551_class0.png    0\n",
      "Name: group, dtype: int64\n",
      "(55505,)\n"
     ]
    }
   ],
   "source": [
    "X= df[['red','green','blue','gray','gray_var','edges','ridges']]\n",
    "y=df['group']\n",
    "  \n",
    "# using the train test split function\n",
    "X_train, X_test, y_train, y_test = train_test_split(X,y, test_size=0.2, shuffle=False)\n",
    "X_train, X_val, y_train, y_val = train_test_split(X_train, y_train, test_size=0.125, shuffle=False)\n",
    "\n",
    "print('X_train : ')\n",
    "print(X_train.head())\n",
    "print(X_train.shape)\n",
    "print('')\n",
    "print('X_val : ')\n",
    "print(X_val.head())\n",
    "print(X_val.shape)\n",
    "print('')\n",
    "print('X_test : ')\n",
    "print(X_test.head())\n",
    "print(X_test.shape)\n",
    "print('')\n",
    "print('y_train : ')\n",
    "print(y_train.head())\n",
    "print(y_train.shape)\n",
    "print('')\n",
    "print('y_val : ')\n",
    "print(y_val.head())\n",
    "print(y_val.shape)\n",
    "print('')\n",
    "print('y_test : ')\n",
    "print(y_test.head())\n",
    "print(y_test.shape)\n",
    "\n",
    "\n",
    "\n",
    "#Split: 70% train, 10% validation and 20% test."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "44e3cdfd-b77d-4060-948b-a2460e0ce569",
   "metadata": {},
   "source": [
    "## Single decision tree classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "3b3b63bc-71db-46c0-b7b5-a06c89b2c95d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy and balancend accuracy for max_depth of 1 is: 0.6743775447699348 , 0.6910898622593671\n",
      "Accuracy and balancend accuracy for max_depth of 2 is: 0.7307678449176666 , 0.6279309405299985\n",
      "Accuracy and balancend accuracy for max_depth of 3 is: 0.7611069073613663 , 0.6324538093264291\n",
      "Accuracy and balancend accuracy for max_depth of 4 is: 0.7845277987965265 , 0.6755620317682303\n",
      "Accuracy and balancend accuracy for max_depth of 5 is: 0.7894281699275754 , 0.6556362113575418\n",
      "Accuracy and balancend accuracy for max_depth of 6 is: 0.7789067848520881 , 0.6725507266021542\n",
      "Accuracy and balancend accuracy for max_depth of 7 is: 0.7786905920080712 , 0.6923540581623155\n",
      "Accuracy and balancend accuracy for max_depth of 8 is: 0.7880589485821352 , 0.6908705346625522\n",
      "Accuracy and balancend accuracy for max_depth of 9 is: 0.7859690844233056 , 0.6954887484388006\n",
      "Accuracy and balancend accuracy for max_depth of 10 is: 0.7940402839332684 , 0.6957826363219817\n",
      "Accuracy and balancend accuracy for max_depth of 11 is: 0.7935718661045653 , 0.691336120331681\n",
      "Accuracy and balancend accuracy for max_depth of 12 is: 0.786833855799373 , 0.6881730058132269\n",
      "Accuracy and balancend accuracy for max_depth of 13 is: 0.7831946095917558 , 0.6854909694561461\n",
      "Accuracy and balancend accuracy for max_depth of 14 is: 0.7733578351889886 , 0.6796793216285776\n",
      "Accuracy and balancend accuracy for max_depth of 15 is: 0.7691060425899903 , 0.6746267205628991\n",
      "Accuracy and balancend accuracy for max_depth of 16 is: 0.766439664180449 , 0.6754109906386082\n",
      "Accuracy and balancend accuracy for max_depth of 17 is: 0.759341332468562 , 0.668489260928737\n",
      "Accuracy and balancend accuracy for max_depth of 18 is: 0.754585089900191 , 0.6661508335387618\n",
      "Accuracy and balancend accuracy for max_depth of 19 is: 0.7483515295643715 , 0.6637414735809901\n",
      "Accuracy and balancend accuracy for max_depth of 20 is: 0.7463697618275502 , 0.6609351131745764\n",
      "Accuracy and balancend accuracy for max_depth of 21 is: 0.7416135192591792 , 0.6585347584946658\n",
      "Accuracy and balancend accuracy for max_depth of 22 is: 0.7402082657730695 , 0.659128850759243\n",
      "Accuracy and balancend accuracy for max_depth of 23 is: 0.737145533816164 , 0.6560077750971546\n",
      "Accuracy and balancend accuracy for max_depth of 24 is: 0.7326054840918099 , 0.6538066035062369\n",
      "Accuracy and balancend accuracy for max_depth of 25 is: 0.7296868806975823 , 0.6492907767517356\n",
      "Accuracy and balancend accuracy for max_depth of 26 is: 0.7279573379454474 , 0.6496170580277912\n",
      "Accuracy and balancend accuracy for max_depth of 27 is: 0.7281014665081252 , 0.6510709622724098\n",
      "Accuracy and balancend accuracy for max_depth of 28 is: 0.7257233452239398 , 0.6467744204356813\n",
      "Accuracy and balancend accuracy for max_depth of 29 is: 0.7248225417072028 , 0.6483699764206828\n",
      "Accuracy and balancend accuracy for max_depth of 30 is: 0.7215436169062804 , 0.6451735722494722\n",
      "Accuracy and balancend accuracy for max_depth of 31 is: 0.7232371275177458 , 0.6473634338942611\n",
      "Accuracy and balancend accuracy for max_depth of 32 is: 0.7213994883436025 , 0.6443389409042084\n",
      "Accuracy and balancend accuracy for max_depth of 33 is: 0.7203545562641877 , 0.6438613197452367\n",
      "Accuracy and balancend accuracy for max_depth of 34 is: 0.7219039383129752 , 0.6452784773346973\n",
      "Accuracy and balancend accuracy for max_depth of 35 is: 0.7219039383129752 , 0.6442257134057944\n",
      "Accuracy and balancend accuracy for max_depth of 36 is: 0.7218679061723057 , 0.6463083652970905\n",
      "Accuracy and balancend accuracy for max_depth of 37 is: 0.7202824919828487 , 0.6437536405222822\n",
      "Accuracy and balancend accuracy for max_depth of 38 is: 0.7213274240622636 , 0.6455317347698986\n",
      "Accuracy and balancend accuracy for max_depth of 39 is: 0.7190573992000865 , 0.6434712759804401\n",
      "Accuracy and balancend accuracy for max_depth of 40 is: 0.7209671026555687 , 0.64561261155448\n",
      "Accuracy and balancend accuracy for max_depth of 41 is: 0.721435520484272 , 0.6456622899593627\n",
      "Accuracy and balancend accuracy for max_depth of 42 is: 0.720606781248874 , 0.6458173429189323\n",
      "Accuracy and balancend accuracy for max_depth of 43 is: 0.7185169170900443 , 0.6424469362935061\n",
      "Accuracy and balancend accuracy for max_depth of 44 is: 0.7209310705148992 , 0.6449085353986803\n",
      "Accuracy and balancend accuracy for max_depth of 45 is: 0.7212553597809246 , 0.6440616551683637\n",
      "Accuracy and balancend accuracy for max_depth of 46 is: 0.7220120347349837 , 0.6460902327134516\n",
      "Accuracy and balancend accuracy for max_depth of 47 is: 0.7205707491082045 , 0.6460421761121646\n",
      "Accuracy and balancend accuracy for max_depth of 48 is: 0.7209310705148992 , 0.6442892624993257\n",
      "Accuracy and balancend accuracy for max_depth of 49 is: 0.7185889813713833 , 0.6430500338359445\n",
      "Execution time: 215.51437854766846s\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Single decision tree with different depths, to see what is proper.\n",
    "acc_for_depths = []\n",
    "bal_accs = []\n",
    "\n",
    "st = time.time()\n",
    "for d in range(1,50):\n",
    "    model = DecisionTreeClassifier(max_depth = d)\n",
    "    model = model.fit(X_train, y_train)\n",
    "    y_pred = model.predict(X_val)\n",
    "    accuracy = model.score(X_val, y_val)\n",
    "    balanced_accuracy = balanced_accuracy_score(y_val, y_pred)\n",
    "    acc_for_depths.append(accuracy)\n",
    "    bal_accs.append(balanced_accuracy)\n",
    "    print(f'Accuracy and balancend accuracy for max_depth of {d} is: {accuracy} , {balanced_accuracy}')\n",
    "    \n",
    "et = time.time()\n",
    "elapsed_time = et - st\n",
    "print(f'Execution time: {elapsed_time}s')\n",
    "    \n",
    "plt.plot(range(1,50), acc_for_depths, label='Accuracy')\n",
    "plt.plot(range(1,50), bal_accs, label='Balanced accuracy')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "#Observing in the plot that max_depth = 11 gives the highest accuracy when validating."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e747761a-4d42-4c15-907d-1482d21916de",
   "metadata": {},
   "source": [
    "### Final test of decision tree with max_depth = 10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "01b54559-af94-42b4-85a4-ba43398558fc",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy for max_depth of 10 is: 0.7990991802540311\n",
      "Balanced accuracy for max_depth of 10 is: 0.7284824557294114\n",
      "Execution time: 2.8073630332946777s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x7f1a917b35b0>"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "st = time.time()\n",
    "depth = 10\n",
    "model = DecisionTreeClassifier(max_depth = depth)\n",
    "model.fit(X_train, y_train)\n",
    "\n",
    "predicted_value = model.predict(X_test)\n",
    "\n",
    "#To plot the actual decision tree\n",
    "#tree.plot_tree(model)\n",
    "\n",
    "accuracy = model.score(X_test, y_test)\n",
    "balanced_accuracy = balanced_accuracy_score(y_test, predicted_value)\n",
    "\n",
    "print(f'Accuracy for max_depth of {depth} is: {accuracy}')\n",
    "print(f'Balanced accuracy for max_depth of {depth} is: {balanced_accuracy}')\n",
    "    \n",
    "et = time.time()\n",
    "elapsed_time = et - st\n",
    "print(f'Execution time: {elapsed_time}s')\n",
    "\n",
    "ConfusionMatrixDisplay(confusion_matrix(y_test, predicted_value)).plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2cd50809-5ce7-4533-8ed8-b1be1506cd2a",
   "metadata": {},
   "source": [
    "## Random forest classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "06074cce-4355-4e9b-94c4-f9ba68395aaa",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy and balancend accuracy for max_depth of 1 is: 0.7875544986127626 , 0.5\n",
      "Accuracy and balancend accuracy for max_depth of 2 is: 0.7875544986127626 , 0.5\n",
      "Accuracy and balancend accuracy for max_depth of 3 is: 0.7954455374193781 , 0.5271798064624946\n",
      "Accuracy and balancend accuracy for max_depth of 4 is: 0.7934277375418873 , 0.6494436957592056\n",
      "Accuracy and balancend accuracy for max_depth of 5 is: 0.7951933124346917 , 0.6670372772410083\n",
      "Accuracy and balancend accuracy for max_depth of 6 is: 0.7955536338413866 , 0.6743257479587468\n",
      "Accuracy and balancend accuracy for max_depth of 7 is: 0.7944006053399633 , 0.6834401561301786\n",
      "Accuracy and balancend accuracy for max_depth of 8 is: 0.7956256981227254 , 0.6895436859259543\n",
      "Accuracy and balancend accuracy for max_depth of 9 is: 0.7977515944222247 , 0.691326858979568\n",
      "Accuracy and balancend accuracy for max_depth of 10 is: 0.7973912730155298 , 0.6920889359534396\n",
      "Accuracy and balancend accuracy for max_depth of 11 is: 0.7993370086116817 , 0.6961728934819884\n",
      "Accuracy and balancend accuracy for max_depth of 12 is: 0.7982200122509279 , 0.6973834845081907\n",
      "Accuracy and balancend accuracy for max_depth of 13 is: 0.8000936835657406 , 0.6963436523290124\n",
      "Accuracy and balancend accuracy for max_depth of 14 is: 0.798832558642309 , 0.6977104486489181\n",
      "Accuracy and balancend accuracy for max_depth of 15 is: 0.8003098764097575 , 0.697967163086521\n",
      "Accuracy and balancend accuracy for max_depth of 16 is: 0.7981839801102584 , 0.6953170079738108\n",
      "Accuracy and balancend accuracy for max_depth of 17 is: 0.7994811371743595 , 0.6973790885668651\n",
      "Accuracy and balancend accuracy for max_depth of 18 is: 0.7984362050949447 , 0.6957248488991197\n",
      "Accuracy and balancend accuracy for max_depth of 19 is: 0.7994090728930205 , 0.6973952639237814\n",
      "Accuracy and balancend accuracy for max_depth of 20 is: 0.7984722372356141 , 0.6961812158951775\n",
      "Accuracy and balancend accuracy for max_depth of 21 is: 0.7987244622203005 , 0.696650984110422\n",
      "Accuracy and balancend accuracy for max_depth of 22 is: 0.7975354015782078 , 0.695276804316251\n",
      "Accuracy and balancend accuracy for max_depth of 23 is: 0.7982560443915973 , 0.6961058873860555\n",
      "Accuracy and balancend accuracy for max_depth of 24 is: 0.7986884300796311 , 0.6968758173036542\n",
      "Accuracy and balancend accuracy for max_depth of 25 is: 0.7986163657982921 , 0.6964585016310223\n",
      "Accuracy and balancend accuracy for max_depth of 26 is: 0.797859690844233 , 0.6961019609141919\n",
      "Accuracy and balancend accuracy for max_depth of 27 is: 0.7965985659208014 , 0.6932577015184862\n",
      "Accuracy and balancend accuracy for max_depth of 28 is: 0.7983641408136057 , 0.695617169676165\n",
      "Accuracy and balancend accuracy for max_depth of 29 is: 0.7977876265628941 , 0.6941983902831088\n",
      "Accuracy and balancend accuracy for max_depth of 30 is: 0.797175080171513 , 0.6951099719410907\n",
      "Accuracy and balancend accuracy for max_depth of 31 is: 0.7966706302021404 , 0.6955328358891819\n",
      "Accuracy and balancend accuracy for max_depth of 32 is: 0.7987965265016395 , 0.6973160089427957\n",
      "Accuracy and balancend accuracy for max_depth of 33 is: 0.7980758836882499 , 0.6959295802635721\n",
      "Accuracy and balancend accuracy for max_depth of 34 is: 0.7969588873274961 , 0.6944772978225495\n",
      "Accuracy and balancend accuracy for max_depth of 35 is: 0.7982920765322667 , 0.6958191269028877\n",
      "Accuracy and balancend accuracy for max_depth of 36 is: 0.7961661802327676 , 0.6924877716008875\n",
      "Accuracy and balancend accuracy for max_depth of 37 is: 0.7978236587035635 , 0.6953978847583923\n",
      "Accuracy and balancend accuracy for max_depth of 38 is: 0.7991208157676648 , 0.6950448010439635\n",
      "Accuracy and balancend accuracy for max_depth of 39 is: 0.7991208157676648 , 0.6967787651621564\n",
      "Accuracy and balancend accuracy for max_depth of 40 is: 0.7986163657982921 , 0.6948483920927002\n",
      "Accuracy and balancend accuracy for max_depth of 41 is: 0.7968868230461572 , 0.6918925270021763\n",
      "Accuracy and balancend accuracy for max_depth of 42 is: 0.7974993694375383 , 0.6937057461013549\n",
      "Accuracy and balancend accuracy for max_depth of 43 is: 0.797175080171513 , 0.6936237169826396\n",
      "Accuracy and balancend accuracy for max_depth of 44 is: 0.7980758836882499 , 0.6938240524057664\n",
      "Accuracy and balancend accuracy for max_depth of 45 is: 0.799373040752351 , 0.6948333690699178\n",
      "Accuracy and balancend accuracy for max_depth of 46 is: 0.7978236587035635 , 0.6942831935395539\n",
      "Accuracy and balancend accuracy for max_depth of 47 is: 0.7995532014556985 , 0.6956289490917559\n",
      "Accuracy and balancend accuracy for max_depth of 48 is: 0.79944510503369 , 0.6959318849318399\n",
      "Accuracy and balancend accuracy for max_depth of 49 is: 0.7989766872049868 , 0.6957583519470862\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Random forest classification with different values for max_depth\n",
    "acc_for_depths = []\n",
    "bal_accs = []\n",
    "for i in range(1,50):\n",
    "    model = RandomForestClassifier(max_depth = i)\n",
    "\n",
    "    # fit the regressor with x and y data\n",
    "    model.fit(X_train, y_train) \n",
    "\n",
    "    y_pred = model.predict(X_val)  # test the output by changing values\n",
    "    #print(y_pred)\n",
    "\n",
    "    #print(\"accuracy on training set: %f\" % regressor.score(X_train, y_train))\n",
    "    #print(\"accuracy on validation set: %f\" % regressor.score(X_val, y_val))\n",
    "\n",
    "    # Create confusion matrix\n",
    "    #confusion = confusion_matrix(y_val, y_pred)\n",
    "    #ConfusionMatrixDisplay(confusion).plot()\n",
    "    # Display accuracy score\n",
    "    accuracy = model.score(X_val, y_val)\n",
    "    balanced_accuracy = balanced_accuracy_score(y_val, y_pred)\n",
    "    print(f'Accuracy and balancend accuracy for max_depth of {i} is: {accuracy} , {balanced_accuracy}')\n",
    "    acc_for_depths.append(accuracy)\n",
    "    bal_accs.append(balanced_accuracy)\n",
    "    \n",
    "plt.plot(range(1,50), acc_for_depths, label='Accuracy')\n",
    "plt.plot(range(1,50), bal_accs, label='Balanced accuracy')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d68869c7-8315-49da-ac7d-5ce1bd25435f",
   "metadata": {},
   "source": [
    "### Final test of random forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "5d2fa05e-b5b5-42ed-ab34-6bee6ee58a8d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy and balanced accuracy for random forest with max_depth = 15: 0.8034951806143591 , 0.7305675493670385\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x7f1a91312700>"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "depth = 15 #As found optimal above\n",
    "model = RandomForestClassifier(max_depth = depth)\n",
    "model.fit(X_train, y_train)\n",
    "y_pred = model.predict(X_test)\n",
    "accuracy = model.score(X_test, y_test)\n",
    "balanced_accuracy = balanced_accuracy_score(y_test, y_pred)\n",
    "print(f'Accuracy and balanced accuracy for random forest with max_depth = {depth}: {accuracy} , {balanced_accuracy}')\n",
    "\n",
    "# Create confusion matrix\n",
    "confusion = confusion_matrix(y_test, y_pred)\n",
    "ConfusionMatrixDisplay(confusion).plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fe9372ed-cd8e-4f2b-9b23-f38a963cba2a",
   "metadata": {},
   "source": [
    "### Feature importance for the random forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "5c195173-8766-45b2-b82b-bc45bad99ebd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.0, 1.0)"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(model.feature_importances_, 'o')\n",
    "plt.xticks(range(X_test.shape[1]), X_test.columns, rotation=90)\n",
    "plt.ylim(0, 1)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}