Pipe_Line_Random_Forest.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "cd91071d",
   "metadata": {},
   "source": [
    "# بسم الله الرحمن الرحيم"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "id": "b9f50056",
   "metadata": {},
   "outputs": [],
   "source": [
    "from utils import *\n",
    "import pandas as pd \n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from matplotlib import pyplot as plt \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "id": "9f472df1",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv(\"dataset/body_level_classification_train.csv\")\n",
    "df_original= cleaning_data(df).copy()\n",
    "all_class_train,all_class_test=data_spliting(df_original,1/3)\n",
    "## shuffling \n",
    "all_class_train = all_class_train.sample(frac=1.0, random_state=42)\n",
    "all_class_test = all_class_test.sample(frac=1.0, random_state=42)\n",
    "X_train=all_class_train.drop([\"Body_Level\"],axis=1).to_numpy()\n",
    "Y_train=all_class_train[\"Body_Level\"].to_numpy()\n",
    "X_test =all_class_test.drop([\"Body_Level\"],axis=1).to_numpy()\n",
    "Y_test =all_class_test[\"Body_Level\"].to_numpy()\n",
    "\n",
    "\n",
    "X_train, scaler= data_scaling(X_train)\n",
    "X_val, X_test, Y_val, Y_test = train_test_split(X_test\n",
    "                                                     , Y_test, test_size=0.5, random_state=42)\n",
    "X_val = scaler.transform(X_val)\n",
    "X_test=scaler.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "id": "36909894",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{3: 0.5429515418502202,\n",
       " 2: 0.9095940959409594,\n",
       " 1: 1.8395522388059702,\n",
       " 0: 1.9409448818897639}"
      ]
     },
     "execution_count": 149,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "class_weights = all_class_train[\"Body_Level\"].value_counts().to_dict()\n",
    "total_num = sum(class_weights.values())\n",
    "class_weights= { i : (total_num/(4* j )) for i ,j in zip(class_weights.keys(),class_weights.values())}\n",
    "class_weights"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eb54bbaf",
   "metadata": {},
   "source": [
    "# Applying bare Random Forest "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "id": "ef60e071",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(random_state=42)"
      ]
     },
     "execution_count": 150,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bare_RF = RandomForestClassifier(random_state=42)\n",
    "bare_RF.fit(np.concatenate([X_train,X_val], axis=0)\n",
    "                   ,np.concatenate([Y_train,Y_val],axis=0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "id": "15863b60",
   "metadata": {},
   "outputs": [],
   "source": [
    "bare_RF_accuracy , repo = test_model(bare_RF,X_train,Y_train,axis=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "id": "8d3ab18f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.0\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       1.00      1.00      1.00       127\n",
      "           1       1.00      1.00      1.00       134\n",
      "           2       1.00      1.00      1.00       271\n",
      "           3       1.00      1.00      1.00       454\n",
      "\n",
      "    accuracy                           1.00       986\n",
      "   macro avg       1.00      1.00      1.00       986\n",
      "weighted avg       1.00      1.00      1.00       986\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(bare_RF_accuracy)\n",
    "print(repo)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "id": "151ff223",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9308943089430894\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       1.00      0.93      0.96        28\n",
      "           1       0.88      0.83      0.86        36\n",
      "           2       0.89      0.94      0.91        80\n",
      "           3       0.96      0.96      0.96       102\n",
      "\n",
      "    accuracy                           0.93       246\n",
      "   macro avg       0.93      0.92      0.92       246\n",
      "weighted avg       0.93      0.93      0.93       246\n",
      "\n"
     ]
    }
   ],
   "source": [
    "bare_RF_accuracy_test , repo = test_model(bare_RF,X_test,Y_test,axis=False)\n",
    "print(bare_RF_accuracy_test)\n",
    "print(repo)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "56226ead",
   "metadata": {},
   "source": [
    "## Applying Random Forest With Class Weights"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "id": "593bfadb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(class_weight={0: 1.9409448818897639,\n",
       "                                     1: 1.8395522388059702,\n",
       "                                     2: 0.9095940959409594,\n",
       "                                     3: 0.5429515418502202},\n",
       "                       random_state=42)"
      ]
     },
     "execution_count": 154,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "weights_RF_model = RandomForestClassifier(random_state=42,class_weight=class_weights)\n",
    "weights_RF_model.fit(np.concatenate([X_train,X_val], axis=0)\n",
    "                   ,np.concatenate([Y_train,Y_val],axis=0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "id": "fcd4f66c",
   "metadata": {},
   "outputs": [],
   "source": [
    "weights_RF , repo = test_model(weights_RF_model,X_train,Y_train,axis=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "id": "9c5c01cb",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.0\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       1.00      1.00      1.00       127\n",
      "           1       1.00      1.00      1.00       134\n",
      "           2       1.00      1.00      1.00       271\n",
      "           3       1.00      1.00      1.00       454\n",
      "\n",
      "    accuracy                           1.00       986\n",
      "   macro avg       1.00      1.00      1.00       986\n",
      "weighted avg       1.00      1.00      1.00       986\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(weights_RF)\n",
    "print(repo)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "id": "c1cf9f09",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.926829268292683\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.96      0.89      0.93        28\n",
      "           1       0.86      0.83      0.85        36\n",
      "           2       0.90      0.94      0.92        80\n",
      "           3       0.96      0.96      0.96       102\n",
      "\n",
      "    accuracy                           0.93       246\n",
      "   macro avg       0.92      0.91      0.91       246\n",
      "weighted avg       0.93      0.93      0.93       246\n",
      "\n"
     ]
    }
   ],
   "source": [
    "weights_RF_test , repo = test_model(weights_RF_model,X_test,Y_test,axis=False)\n",
    "print(weights_RF_test)\n",
    "print(repo)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "014dcc87",
   "metadata": {},
   "source": [
    "## HyperParameter Tuning on Random Forest "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "id": "48c78ac2",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 28 candidates, totalling 140 fits\n",
      "[CV] END ......................max_depth=10, n_estimators=50; total time=   0.1s\n",
      "[CV] END ......................max_depth=10, n_estimators=50; total time=   0.1s\n",
      "[CV] END ......................max_depth=10, n_estimators=50; total time=   0.1s\n",
      "[CV] END ......................max_depth=10, n_estimators=50; total time=   0.1s\n",
      "[CV] END ......................max_depth=10, n_estimators=50; total time=   0.1s\n",
      "[CV] END .....................max_depth=10, n_estimators=100; total time=   0.3s\n",
      "[CV] END .....................max_depth=10, n_estimators=100; total time=   0.2s\n",
      "[CV] END .....................max_depth=10, n_estimators=100; total time=   0.3s\n",
      "[CV] END .....................max_depth=10, n_estimators=100; total time=   0.2s\n",
      "[CV] END .....................max_depth=10, n_estimators=100; total time=   0.2s\n",
      "[CV] END .....................max_depth=10, n_estimators=200; total time=   0.4s\n",
      "[CV] END .....................max_depth=10, n_estimators=200; total time=   0.4s\n",
      "[CV] END .....................max_depth=10, n_estimators=200; total time=   0.5s\n",
      "[CV] END .....................max_depth=10, n_estimators=200; total time=   0.4s\n",
      "[CV] END .....................max_depth=10, n_estimators=200; total time=   0.4s\n",
      "[CV] END .....................max_depth=10, n_estimators=300; total time=   0.5s\n",
      "[CV] END .....................max_depth=10, n_estimators=300; total time=   0.6s\n",
      "[CV] END .....................max_depth=10, n_estimators=300; total time=   0.6s\n",
      "[CV] END .....................max_depth=10, n_estimators=300; total time=   0.6s\n",
      "[CV] END .....................max_depth=10, n_estimators=300; total time=   0.7s\n",
      "[CV] END .....................max_depth=10, n_estimators=400; total time=   0.9s\n",
      "[CV] END .....................max_depth=10, n_estimators=400; total time=   0.8s\n",
      "[CV] END .....................max_depth=10, n_estimators=400; total time=   1.1s\n",
      "[CV] END .....................max_depth=10, n_estimators=400; total time=   1.4s\n",
      "[CV] END .....................max_depth=10, n_estimators=400; total time=   1.2s\n",
      "[CV] END .....................max_depth=10, n_estimators=500; total time=   1.4s\n",
      "[CV] END .....................max_depth=10, n_estimators=500; total time=   1.3s\n",
      "[CV] END .....................max_depth=10, n_estimators=500; total time=   1.0s\n",
      "[CV] END .....................max_depth=10, n_estimators=500; total time=   1.0s\n",
      "[CV] END .....................max_depth=10, n_estimators=500; total time=   1.2s\n",
      "[CV] END .....................max_depth=10, n_estimators=600; total time=   1.6s\n",
      "[CV] END .....................max_depth=10, n_estimators=600; total time=   1.5s\n",
      "[CV] END .....................max_depth=10, n_estimators=600; total time=   1.7s\n",
      "[CV] END .....................max_depth=10, n_estimators=600; total time=   1.5s\n",
      "[CV] END .....................max_depth=10, n_estimators=600; total time=   1.7s\n",
      "[CV] END ......................max_depth=20, n_estimators=50; total time=   0.2s\n",
      "[CV] END ......................max_depth=20, n_estimators=50; total time=   0.1s\n",
      "[CV] END ......................max_depth=20, n_estimators=50; total time=   0.2s\n",
      "[CV] END ......................max_depth=20, n_estimators=50; total time=   0.2s\n",
      "[CV] END ......................max_depth=20, n_estimators=50; total time=   0.2s\n",
      "[CV] END .....................max_depth=20, n_estimators=100; total time=   0.4s\n",
      "[CV] END .....................max_depth=20, n_estimators=100; total time=   0.3s\n",
      "[CV] END .....................max_depth=20, n_estimators=100; total time=   0.3s\n",
      "[CV] END .....................max_depth=20, n_estimators=100; total time=   0.2s\n",
      "[CV] END .....................max_depth=20, n_estimators=100; total time=   0.3s\n",
      "[CV] END .....................max_depth=20, n_estimators=200; total time=   0.6s\n",
      "[CV] END .....................max_depth=20, n_estimators=200; total time=   0.7s\n",
      "[CV] END .....................max_depth=20, n_estimators=200; total time=   0.6s\n",
      "[CV] END .....................max_depth=20, n_estimators=200; total time=   0.6s\n",
      "[CV] END .....................max_depth=20, n_estimators=200; total time=   0.6s\n",
      "[CV] END .....................max_depth=20, n_estimators=300; total time=   0.9s\n",
      "[CV] END .....................max_depth=20, n_estimators=300; total time=   0.9s\n",
      "[CV] END .....................max_depth=20, n_estimators=300; total time=   0.7s\n",
      "[CV] END .....................max_depth=20, n_estimators=300; total time=   0.9s\n",
      "[CV] END .....................max_depth=20, n_estimators=300; total time=   0.9s\n",
      "[CV] END .....................max_depth=20, n_estimators=400; total time=   1.3s\n",
      "[CV] END .....................max_depth=20, n_estimators=400; total time=   1.3s\n",
      "[CV] END .....................max_depth=20, n_estimators=400; total time=   1.2s\n",
      "[CV] END .....................max_depth=20, n_estimators=400; total time=   1.0s\n",
      "[CV] END .....................max_depth=20, n_estimators=400; total time=   1.4s\n",
      "[CV] END .....................max_depth=20, n_estimators=500; total time=   1.4s\n",
      "[CV] END .....................max_depth=20, n_estimators=500; total time=   1.3s\n",
      "[CV] END .....................max_depth=20, n_estimators=500; total time=   1.3s\n",
      "[CV] END .....................max_depth=20, n_estimators=500; total time=   1.0s\n",
      "[CV] END .....................max_depth=20, n_estimators=500; total time=   1.1s\n",
      "[CV] END .....................max_depth=20, n_estimators=600; total time=   1.2s\n",
      "[CV] END .....................max_depth=20, n_estimators=600; total time=   1.2s\n",
      "[CV] END .....................max_depth=20, n_estimators=600; total time=   1.3s\n",
      "[CV] END .....................max_depth=20, n_estimators=600; total time=   1.2s\n",
      "[CV] END .....................max_depth=20, n_estimators=600; total time=   1.2s\n",
      "[CV] END ......................max_depth=30, n_estimators=50; total time=   0.1s\n",
      "[CV] END ......................max_depth=30, n_estimators=50; total time=   0.1s\n",
      "[CV] END ......................max_depth=30, n_estimators=50; total time=   0.1s\n",
      "[CV] END ......................max_depth=30, n_estimators=50; total time=   0.1s\n",
      "[CV] END ......................max_depth=30, n_estimators=50; total time=   0.1s\n",
      "[CV] END .....................max_depth=30, n_estimators=100; total time=   0.2s\n",
      "[CV] END .....................max_depth=30, n_estimators=100; total time=   0.2s\n",
      "[CV] END .....................max_depth=30, n_estimators=100; total time=   0.2s\n",
      "[CV] END .....................max_depth=30, n_estimators=100; total time=   0.2s\n",
      "[CV] END .....................max_depth=30, n_estimators=100; total time=   0.3s\n",
      "[CV] END .....................max_depth=30, n_estimators=200; total time=   0.4s\n",
      "[CV] END .....................max_depth=30, n_estimators=200; total time=   0.4s\n",
      "[CV] END .....................max_depth=30, n_estimators=200; total time=   0.4s\n",
      "[CV] END .....................max_depth=30, n_estimators=200; total time=   0.5s\n",
      "[CV] END .....................max_depth=30, n_estimators=200; total time=   0.5s\n",
      "[CV] END .....................max_depth=30, n_estimators=300; total time=   0.7s\n",
      "[CV] END .....................max_depth=30, n_estimators=300; total time=   0.7s\n",
      "[CV] END .....................max_depth=30, n_estimators=300; total time=   0.6s\n",
      "[CV] END .....................max_depth=30, n_estimators=300; total time=   0.7s\n",
      "[CV] END .....................max_depth=30, n_estimators=300; total time=   0.6s\n",
      "[CV] END .....................max_depth=30, n_estimators=400; total time=   0.9s\n",
      "[CV] END .....................max_depth=30, n_estimators=400; total time=   0.9s\n",
      "[CV] END .....................max_depth=30, n_estimators=400; total time=   0.9s\n",
      "[CV] END .....................max_depth=30, n_estimators=400; total time=   0.9s\n",
      "[CV] END .....................max_depth=30, n_estimators=400; total time=   0.9s\n",
      "[CV] END .....................max_depth=30, n_estimators=500; total time=   1.1s\n",
      "[CV] END .....................max_depth=30, n_estimators=500; total time=   1.1s\n",
      "[CV] END .....................max_depth=30, n_estimators=500; total time=   1.1s\n",
      "[CV] END .....................max_depth=30, n_estimators=500; total time=   1.4s\n",
      "[CV] END .....................max_depth=30, n_estimators=500; total time=   1.2s\n",
      "[CV] END .....................max_depth=30, n_estimators=600; total time=   1.4s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV] END .....................max_depth=30, n_estimators=600; total time=   1.4s\n",
      "[CV] END .....................max_depth=30, n_estimators=600; total time=   1.3s\n",
      "[CV] END .....................max_depth=30, n_estimators=600; total time=   1.3s\n",
      "[CV] END .....................max_depth=30, n_estimators=600; total time=   1.3s\n",
      "[CV] END ....................max_depth=None, n_estimators=50; total time=   0.1s\n",
      "[CV] END ....................max_depth=None, n_estimators=50; total time=   0.1s\n",
      "[CV] END ....................max_depth=None, n_estimators=50; total time=   0.1s\n",
      "[CV] END ....................max_depth=None, n_estimators=50; total time=   0.1s\n",
      "[CV] END ....................max_depth=None, n_estimators=50; total time=   0.1s\n",
      "[CV] END ...................max_depth=None, n_estimators=100; total time=   0.2s\n",
      "[CV] END ...................max_depth=None, n_estimators=100; total time=   0.2s\n",
      "[CV] END ...................max_depth=None, n_estimators=100; total time=   0.3s\n",
      "[CV] END ...................max_depth=None, n_estimators=100; total time=   0.2s\n",
      "[CV] END ...................max_depth=None, n_estimators=100; total time=   0.2s\n",
      "[CV] END ...................max_depth=None, n_estimators=200; total time=   0.4s\n",
      "[CV] END ...................max_depth=None, n_estimators=200; total time=   0.4s\n",
      "[CV] END ...................max_depth=None, n_estimators=200; total time=   0.4s\n",
      "[CV] END ...................max_depth=None, n_estimators=200; total time=   0.4s\n",
      "[CV] END ...................max_depth=None, n_estimators=200; total time=   0.4s\n",
      "[CV] END ...................max_depth=None, n_estimators=300; total time=   0.6s\n",
      "[CV] END ...................max_depth=None, n_estimators=300; total time=   0.7s\n",
      "[CV] END ...................max_depth=None, n_estimators=300; total time=   0.6s\n",
      "[CV] END ...................max_depth=None, n_estimators=300; total time=   0.7s\n",
      "[CV] END ...................max_depth=None, n_estimators=300; total time=   0.7s\n",
      "[CV] END ...................max_depth=None, n_estimators=400; total time=   0.9s\n",
      "[CV] END ...................max_depth=None, n_estimators=400; total time=   0.9s\n",
      "[CV] END ...................max_depth=None, n_estimators=400; total time=   0.9s\n",
      "[CV] END ...................max_depth=None, n_estimators=400; total time=   1.0s\n",
      "[CV] END ...................max_depth=None, n_estimators=400; total time=   1.1s\n",
      "[CV] END ...................max_depth=None, n_estimators=500; total time=   1.1s\n",
      "[CV] END ...................max_depth=None, n_estimators=500; total time=   1.0s\n",
      "[CV] END ...................max_depth=None, n_estimators=500; total time=   1.1s\n",
      "[CV] END ...................max_depth=None, n_estimators=500; total time=   1.0s\n",
      "[CV] END ...................max_depth=None, n_estimators=500; total time=   1.2s\n",
      "[CV] END ...................max_depth=None, n_estimators=600; total time=   1.3s\n",
      "[CV] END ...................max_depth=None, n_estimators=600; total time=   1.3s\n",
      "[CV] END ...................max_depth=None, n_estimators=600; total time=   1.3s\n",
      "[CV] END ...................max_depth=None, n_estimators=600; total time=   1.2s\n",
      "[CV] END ...................max_depth=None, n_estimators=600; total time=   1.4s\n",
      "Best parameters:  {'max_depth': 20, 'n_estimators': 200}\n",
      "Best score:  0.9471873868536257\n"
     ]
    }
   ],
   "source": [
    "\n",
    "# Define the random forest model\n",
    "best_RF_model = RandomForestClassifier(random_state=42)\n",
    "\n",
    "# Define the parameter grid to search over\n",
    "param_grid = {\n",
    "    'n_estimators': [50, 100, 200,300,400,500,600],\n",
    "    'max_depth': [10, 20, 30, None],\n",
    "}\n",
    "\n",
    "# Define the grid search object\n",
    "grid_search = GridSearchCV(estimator=best_RF_model, param_grid=param_grid, cv=5,verbose=2)\n",
    "\n",
    "# Fit the grid search object to the data\n",
    "grid_search.fit(np.concatenate([X_train,X_val], axis=0)\n",
    "                   ,np.concatenate([Y_train,Y_val],axis=0))\n",
    "\n",
    "# Print the best parameters and best score\n",
    "print(\"Best parameters: \", grid_search.best_params_)\n",
    "print(\"Best score: \", grid_search.best_score_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "id": "6934f3b2",
   "metadata": {},
   "outputs": [],
   "source": [
    "cv_results = grid_search.cv_results_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "id": "eb720c19",
   "metadata": {},
   "outputs": [],
   "source": [
    "best_RF_model_accuracy , repo = test_model(grid_search.best_estimator_,X_test,Y_test,axis=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "id": "194e9857",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9308943089430894\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       1.00      0.93      0.96        28\n",
      "           1       0.86      0.86      0.86        36\n",
      "           2       0.90      0.93      0.91        80\n",
      "           3       0.96      0.96      0.96       102\n",
      "\n",
      "    accuracy                           0.93       246\n",
      "   macro avg       0.93      0.92      0.92       246\n",
      "weighted avg       0.93      0.93      0.93       246\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(best_RF_model_accuracy )\n",
    "print(repo)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "id": "7366433a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'mean_fit_time': array([0.09650674, 0.21562166, 0.39444904, 0.58546119, 1.02023306,\n",
       "        1.11603432, 1.48647699, 0.1548882 , 0.27653098, 0.58450894,\n",
       "        0.80679479, 1.1791182 , 1.14662828, 1.1433413 , 0.10606866,\n",
       "        0.19887366, 0.41947422, 0.63028426, 0.83147035, 1.12926598,\n",
       "        1.26710887, 0.11043153, 0.20437198, 0.40504966, 0.62685761,\n",
       "        0.88684611, 1.03342948, 1.21532655]),\n",
       " 'std_fit_time': array([0.010446  , 0.0201871 , 0.02368314, 0.06851658, 0.20289397,\n",
       "        0.14703374, 0.08793916, 0.03072102, 0.05207899, 0.02516233,\n",
       "        0.09266163, 0.11408904, 0.13580347, 0.0292674 , 0.00569471,\n",
       "        0.0228225 , 0.02679266, 0.02226686, 0.0160014 , 0.11581054,\n",
       "        0.03801445, 0.00809513, 0.02117874, 0.01028289, 0.02633414,\n",
       "        0.08018844, 0.08431763, 0.04646372]),\n",
       " 'mean_score_time': array([0.00747404, 0.01695771, 0.02476807, 0.0379066 , 0.06129885,\n",
       "        0.07132502, 0.0877687 , 0.00923576, 0.01614351, 0.03375916,\n",
       "        0.04415803, 0.06224413, 0.06133771, 0.07560401, 0.00763869,\n",
       "        0.01396556, 0.0278543 , 0.04108486, 0.05542417, 0.06407247,\n",
       "        0.09026589, 0.0075366 , 0.01387596, 0.02758021, 0.0433764 ,\n",
       "        0.05579939, 0.06170387, 0.07602668]),\n",
       " 'std_score_time': array([0.00118923, 0.00242085, 0.00263504, 0.00194175, 0.00768739,\n",
       "        0.01525262, 0.01241939, 0.00127419, 0.00193732, 0.00302371,\n",
       "        0.00929352, 0.00886112, 0.00811932, 0.00892928, 0.00080056,\n",
       "        0.00213474, 0.0044534 , 0.00469614, 0.00403157, 0.00643748,\n",
       "        0.01159767, 0.00078948, 0.00176668, 0.00265767, 0.00401222,\n",
       "        0.00606215, 0.00420109, 0.00593201]),\n",
       " 'param_max_depth': masked_array(data=[10, 10, 10, 10, 10, 10, 10, 20, 20, 20, 20, 20, 20, 20,\n",
       "                    30, 30, 30, 30, 30, 30, 30, None, None, None, None,\n",
       "                    None, None, None],\n",
       "              mask=[False, False, False, False, False, False, False, False,\n",
       "                    False, False, False, False, False, False, False, False,\n",
       "                    False, False, False, False, False, False, False, False,\n",
       "                    False, False, False, False],\n",
       "        fill_value='?',\n",
       "             dtype=object),\n",
       " 'param_n_estimators': masked_array(data=[50, 100, 200, 300, 400, 500, 600, 50, 100, 200, 300,\n",
       "                    400, 500, 600, 50, 100, 200, 300, 400, 500, 600, 50,\n",
       "                    100, 200, 300, 400, 500, 600],\n",
       "              mask=[False, False, False, False, False, False, False, False,\n",
       "                    False, False, False, False, False, False, False, False,\n",
       "                    False, False, False, False, False, False, False, False,\n",
       "                    False, False, False, False],\n",
       "        fill_value='?',\n",
       "             dtype=object),\n",
       " 'params': [{'max_depth': 10, 'n_estimators': 50},\n",
       "  {'max_depth': 10, 'n_estimators': 100},\n",
       "  {'max_depth': 10, 'n_estimators': 200},\n",
       "  {'max_depth': 10, 'n_estimators': 300},\n",
       "  {'max_depth': 10, 'n_estimators': 400},\n",
       "  {'max_depth': 10, 'n_estimators': 500},\n",
       "  {'max_depth': 10, 'n_estimators': 600},\n",
       "  {'max_depth': 20, 'n_estimators': 50},\n",
       "  {'max_depth': 20, 'n_estimators': 100},\n",
       "  {'max_depth': 20, 'n_estimators': 200},\n",
       "  {'max_depth': 20, 'n_estimators': 300},\n",
       "  {'max_depth': 20, 'n_estimators': 400},\n",
       "  {'max_depth': 20, 'n_estimators': 500},\n",
       "  {'max_depth': 20, 'n_estimators': 600},\n",
       "  {'max_depth': 30, 'n_estimators': 50},\n",
       "  {'max_depth': 30, 'n_estimators': 100},\n",
       "  {'max_depth': 30, 'n_estimators': 200},\n",
       "  {'max_depth': 30, 'n_estimators': 300},\n",
       "  {'max_depth': 30, 'n_estimators': 400},\n",
       "  {'max_depth': 30, 'n_estimators': 500},\n",
       "  {'max_depth': 30, 'n_estimators': 600},\n",
       "  {'max_depth': None, 'n_estimators': 50},\n",
       "  {'max_depth': None, 'n_estimators': 100},\n",
       "  {'max_depth': None, 'n_estimators': 200},\n",
       "  {'max_depth': None, 'n_estimators': 300},\n",
       "  {'max_depth': None, 'n_estimators': 400},\n",
       "  {'max_depth': None, 'n_estimators': 500},\n",
       "  {'max_depth': None, 'n_estimators': 600}],\n",
       " 'split0_test_score': array([0.95951417, 0.951417  , 0.95951417, 0.95951417, 0.96356275,\n",
       "        0.96356275, 0.95951417, 0.94736842, 0.951417  , 0.95951417,\n",
       "        0.951417  , 0.95951417, 0.95951417, 0.95546559, 0.94736842,\n",
       "        0.951417  , 0.95951417, 0.951417  , 0.95951417, 0.95951417,\n",
       "        0.95546559, 0.94736842, 0.951417  , 0.95951417, 0.951417  ,\n",
       "        0.95951417, 0.95951417, 0.95546559]),\n",
       " 'split1_test_score': array([0.94308943, 0.93902439, 0.94715447, 0.94308943, 0.95121951,\n",
       "        0.95934959, 0.95934959, 0.94308943, 0.94715447, 0.95934959,\n",
       "        0.95528455, 0.95528455, 0.95528455, 0.95528455, 0.94308943,\n",
       "        0.94715447, 0.95934959, 0.95528455, 0.95528455, 0.95528455,\n",
       "        0.95528455, 0.94308943, 0.94715447, 0.95934959, 0.95528455,\n",
       "        0.95528455, 0.95528455, 0.95528455]),\n",
       " 'split2_test_score': array([0.91056911, 0.91056911, 0.91056911, 0.91056911, 0.91056911,\n",
       "        0.91056911, 0.90650407, 0.93089431, 0.94308943, 0.93902439,\n",
       "        0.92276423, 0.93089431, 0.93089431, 0.92682927, 0.93089431,\n",
       "        0.93902439, 0.93902439, 0.92276423, 0.93089431, 0.93089431,\n",
       "        0.92682927, 0.93089431, 0.93902439, 0.93902439, 0.92276423,\n",
       "        0.93089431, 0.93089431, 0.92682927]),\n",
       " 'split3_test_score': array([0.91056911, 0.92276423, 0.92682927, 0.92682927, 0.93089431,\n",
       "        0.92682927, 0.93089431, 0.92276423, 0.92682927, 0.92682927,\n",
       "        0.91869919, 0.92682927, 0.92276423, 0.92682927, 0.92276423,\n",
       "        0.92682927, 0.92682927, 0.91869919, 0.92682927, 0.92276423,\n",
       "        0.92682927, 0.92276423, 0.92682927, 0.92682927, 0.91869919,\n",
       "        0.92682927, 0.92276423, 0.92682927]),\n",
       " 'split4_test_score': array([0.92276423, 0.93495935, 0.93089431, 0.93089431, 0.93902439,\n",
       "        0.93902439, 0.93495935, 0.93902439, 0.93902439, 0.95121951,\n",
       "        0.94308943, 0.94715447, 0.94715447, 0.95121951, 0.93902439,\n",
       "        0.93902439, 0.95121951, 0.94308943, 0.94715447, 0.94715447,\n",
       "        0.95121951, 0.93902439, 0.93902439, 0.95121951, 0.94308943,\n",
       "        0.94715447, 0.94715447, 0.95121951]),\n",
       " 'mean_test_score': array([0.92930121, 0.93174682, 0.93499226, 0.93417926, 0.93905401,\n",
       "        0.93986702, 0.9382443 , 0.93662816, 0.94150291, 0.94718739,\n",
       "        0.93825088, 0.94393535, 0.94312235, 0.94312564, 0.93662816,\n",
       "        0.9406899 , 0.94718739, 0.93825088, 0.94393535, 0.94312235,\n",
       "        0.94312564, 0.93662816, 0.9406899 , 0.94718739, 0.93825088,\n",
       "        0.94393535, 0.94312235, 0.94312564]),\n",
       " 'std_test_score': array([0.01922214, 0.01399559, 0.01690658, 0.01639208, 0.0180369 ,\n",
       "        0.01985003, 0.01984992, 0.00880828, 0.00841671, 0.01262707,\n",
       "        0.01489302, 0.01299657, 0.01411982, 0.01339229, 0.00880828,\n",
       "        0.00842053, 0.01262707, 0.01489302, 0.01299657, 0.01411982,\n",
       "        0.01339229, 0.00880828, 0.00842053, 0.01262707, 0.01489302,\n",
       "        0.01299657, 0.01411982, 0.01339229]),\n",
       " 'rank_test_score': array([28, 27, 25, 26, 17, 16, 21, 22, 13,  1, 18,  4, 10,  7, 22, 14,  1,\n",
       "        18,  4, 10,  7, 22, 14,  1, 18,  4, 10,  7], dtype=int32)}"
      ]
     },
     "execution_count": 162,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cv_results = grid_search.cv_results_\n",
    "cv_results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 165,
   "id": "f7eb7b1a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.929301207991837 {'max_depth': 10, 'n_estimators': 50}\n",
      "0.9317468154438628 {'max_depth': 10, 'n_estimators': 100}\n",
      "0.9349922649024061 {'max_depth': 10, 'n_estimators': 200}\n",
      "0.9341792567723248 {'max_depth': 10, 'n_estimators': 300}\n",
      "0.9390540140219216 {'max_depth': 10, 'n_estimators': 400}\n",
      "0.9398670221520028 {'max_depth': 10, 'n_estimators': 500}\n",
      "0.9382442974227312 {'max_depth': 10, 'n_estimators': 600}\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "cv_results = grid_search.cv_results_\n",
    "estimators_max_depth_10 = []\n",
    "accuracies_max_depth_10 = []\n",
    "\n",
    "cv_results = grid_search.cv_results_\n",
    "\n",
    "\n",
    "for mean_score, params in zip(cv_results['mean_test_score'], cv_results['params']):\n",
    "    if(params[\"max_depth\"]==10):\n",
    "        estimators_max_depth_10.append(params[\"n_estimators\"])\n",
    "        accuracies_max_depth_10.append(mean_score)\n",
    "        print(mean_score,params)\n",
    "        \n",
    "estimators_max_depth_10\n",
    "accuracies_max_depth_10\n",
    "\n",
    "plt.plot(estimators_max_depth_10,accuracies_max_depth_10,label='accuracy')\n",
    "plt.title(\"const depth 10  variable n_estimators\")\n",
    "plt.axhline(1, color='red',label='training-accuracy') \n",
    "plt.legend()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6a037e8e",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 166,
   "id": "a30adb73",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9366281557552419 {'max_depth': 20, 'n_estimators': 50}\n",
      "0.9415029130048385 {'max_depth': 20, 'n_estimators': 100}\n",
      "0.9471873868536257 {'max_depth': 20, 'n_estimators': 200}\n",
      "0.9382508804845134 {'max_depth': 20, 'n_estimators': 300}\n",
      "0.9439353543333004 {'max_depth': 20, 'n_estimators': 400}\n",
      "0.9431223462032191 {'max_depth': 20, 'n_estimators': 500}\n",
      "0.94312563773411 {'max_depth': 20, 'n_estimators': 600}\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "estimators_max_depth_20 = []\n",
    "accuracies_max_depth_20 = []\n",
    "\n",
    "for mean_score, params in zip(cv_results['mean_test_score'], cv_results['params']):\n",
    "    if(params[\"max_depth\"]==20):\n",
    "        estimators_max_depth_20.append(params[\"n_estimators\"])\n",
    "        accuracies_max_depth_20.append(mean_score)\n",
    "        print(mean_score,params)\n",
    "\n",
    "        \n",
    "\n",
    "plt.plot(estimators_max_depth_20,accuracies_max_depth_20)\n",
    "plt.title(\"const depth 20  variable n_estimators\")\n",
    "plt.axhline(1, color='red',label='training-accuracy') \n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "id": "d17b1d6a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9366281557552419 {'max_depth': 30, 'n_estimators': 50}\n",
      "0.9406899048747572 {'max_depth': 30, 'n_estimators': 100}\n",
      "0.9471873868536257 {'max_depth': 30, 'n_estimators': 200}\n",
      "0.9382508804845134 {'max_depth': 30, 'n_estimators': 300}\n",
      "0.9439353543333004 {'max_depth': 30, 'n_estimators': 400}\n",
      "0.9431223462032191 {'max_depth': 30, 'n_estimators': 500}\n",
      "0.94312563773411 {'max_depth': 30, 'n_estimators': 600}\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "estimators_max_depth_30 = []\n",
    "accuracies_max_depth_30 = []\n",
    "\n",
    "for mean_score, params in zip(cv_results['mean_test_score'], cv_results['params']):\n",
    "    if(params[\"max_depth\"]==30):\n",
    "        estimators_max_depth_30.append(params[\"n_estimators\"])\n",
    "        accuracies_max_depth_30.append(mean_score)\n",
    "        print(mean_score,params)\n",
    "\n",
    "\n",
    "\n",
    "plt.plot(estimators_max_depth_30,accuracies_max_depth_30)\n",
    "\n",
    "plt.title(\"const depth 30  variable n_estimators\")\n",
    "plt.axhline(1, color='red',label='training-accuracy') \n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n",
    "plt.show()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "id": "521cd8de",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9366281557552419 {'max_depth': None, 'n_estimators': 50}\n",
      "0.9406899048747572 {'max_depth': None, 'n_estimators': 100}\n",
      "0.9471873868536257 {'max_depth': None, 'n_estimators': 200}\n",
      "0.9382508804845134 {'max_depth': None, 'n_estimators': 300}\n",
      "0.9439353543333004 {'max_depth': None, 'n_estimators': 400}\n",
      "0.9431223462032191 {'max_depth': None, 'n_estimators': 500}\n",
      "0.94312563773411 {'max_depth': None, 'n_estimators': 600}\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "estimators_max_depth_None = []\n",
    "accuracies_max_depth_None = []\n",
    "\n",
    "for mean_score, params in zip(cv_results['mean_test_score'], cv_results['params']):\n",
    "    if(params[\"max_depth\"]==None):\n",
    "        estimators_max_depth_None.append(params[\"n_estimators\"])\n",
    "        accuracies_max_depth_None.append(mean_score)\n",
    "        print(mean_score,params)\n",
    "\n",
    "\n",
    "\n",
    "plt.plot(estimators_max_depth_None,accuracies_max_depth_None)\n",
    "plt.show()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "id": "d2763ce1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "20\n",
      "200\n"
     ]
    }
   ],
   "source": [
    "print(cv_results[\"params\"][np.argmax(cv_results[\"mean_test_score\"])][\"max_depth\"])\n",
    "print(cv_results[\"params\"][np.argmax(cv_results[\"mean_test_score\"])][\"n_estimators\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "id": "ae151c15",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(class_weight={0: 1.9409448818897639,\n",
       "                                     1: 1.8395522388059702,\n",
       "                                     2: 0.9095940959409594,\n",
       "                                     3: 0.5429515418502202},\n",
       "                       max_depth=20, n_estimators=400)"
      ]
     },
     "execution_count": 121,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "best_model_RF= RandomForestClassifier(n_estimators=400,max_depth=20,class_weight=class_weights)\n",
    "# concat class weights with it \n",
    "best_model_RF.fit(np.concatenate([X_train,X_val], axis=0)\n",
    "                   ,np.concatenate([Y_train,Y_val],axis=0))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "id": "4d139bd5",
   "metadata": {},
   "outputs": [],
   "source": [
    "best_RF_model_accuracy , repo = test_model(best_model_RF,X_test,Y_test,axis=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "id": "b1453c1e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9349593495934959\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       1.00      0.93      0.96        28\n",
      "           1       0.86      0.86      0.86        36\n",
      "           2       0.91      0.93      0.92        80\n",
      "           3       0.96      0.97      0.97       102\n",
      "\n",
      "    accuracy                           0.93       246\n",
      "   macro avg       0.93      0.92      0.93       246\n",
      "weighted avg       0.94      0.93      0.94       246\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(best_RF_model_accuracy)\n",
    "print(repo)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bd6097cd",
   "metadata": {},
   "source": [
    "# oversampling "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "id": "0c777ed0",
   "metadata": {},
   "outputs": [],
   "source": [
    "oversampled= over_sampling(all_class_train,sampling_ratio=0.5)\n",
    "class_weights = oversampled[\"Body_Level\"].value_counts().to_dict()\n",
    "total_num = sum(class_weights.values())\n",
    "class_weights= { i : (total_num/(4* j )) for i ,j in zip(class_weights.keys(),class_weights.values())}\n",
    "class_weights\n",
    "\n",
    "X_train=oversampled.drop([\"Body_Level\"],axis=1).to_numpy()\n",
    "Y_train=oversampled[\"Body_Level\"].to_numpy()\n",
    "X_test =all_class_test.drop([\"Body_Level\"],axis=1).to_numpy()\n",
    "Y_test =all_class_test[\"Body_Level\"].to_numpy()\n",
    "X_train, scaler= data_scaling(X_train)\n",
    "X_val, X_test, Y_val, Y_test = train_test_split(X_test\n",
    "                                                     , Y_test, test_size=0.5, random_state=42)\n",
    "X_val = scaler.transform(X_val)\n",
    "X_test=scaler.transform(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b747d794",
   "metadata": {},
   "source": [
    "# Applying bare Random Forest after oversampling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "id": "d0b15a9b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(random_state=42)"
      ]
     },
     "execution_count": 125,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bare_RF_oversampled = RandomForestClassifier(random_state=42)\n",
    "bare_RF_oversampled.fit(np.concatenate([X_train,X_val], axis=0)\n",
    "                   ,np.concatenate([Y_train,Y_val],axis=0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "id": "903d76e7",
   "metadata": {},
   "outputs": [],
   "source": [
    "bare_RF_oversampled_accuracy , repo = test_model(bare_RF,X_train,Y_train,axis=False)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "id": "1ff88032",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.929384965831435\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       1.00      1.00      1.00       257\n",
      "           1       1.00      0.81      0.89       262\n",
      "           2       0.86      0.88      0.87       344\n",
      "           3       0.92      1.00      0.96       454\n",
      "\n",
      "    accuracy                           0.93      1317\n",
      "   macro avg       0.94      0.92      0.93      1317\n",
      "weighted avg       0.93      0.93      0.93      1317\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(bare_RF_oversampled_accuracy)\n",
    "print(repo)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "id": "129782ea",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(class_weight={0: 1.2811284046692606,\n",
       "                                     1: 1.256679389312977,\n",
       "                                     2: 0.9571220930232558,\n",
       "                                     3: 0.7252202643171806},\n",
       "                       random_state=42)"
      ]
     },
     "execution_count": 128,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "weights_RF_oversampled = RandomForestClassifier(random_state=42, class_weight=class_weights)\n",
    "weights_RF_oversampled.fit(np.concatenate([X_train,X_val], axis=0)\n",
    "                   ,np.concatenate([Y_train,Y_val],axis=0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "id": "bdbe5462",
   "metadata": {},
   "outputs": [],
   "source": [
    "weights_RF_oversampled_accuracy , repo = test_model(weights_RF_oversampled,\n",
    "                                                    X_train,Y_train,axis=False)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "id": "cc4052b4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.0\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       1.00      1.00      1.00       257\n",
      "           1       1.00      1.00      1.00       262\n",
      "           2       1.00      1.00      1.00       344\n",
      "           3       1.00      1.00      1.00       454\n",
      "\n",
      "    accuracy                           1.00      1317\n",
      "   macro avg       1.00      1.00      1.00      1317\n",
      "weighted avg       1.00      1.00      1.00      1317\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(weights_RF_oversampled_accuracy)\n",
    "print(repo)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "id": "d65fbd0a",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 28 candidates, totalling 140 fits\n",
      "[CV] END ......................max_depth=10, n_estimators=50; total time=   0.2s\n",
      "[CV] END ......................max_depth=10, n_estimators=50; total time=   0.1s\n",
      "[CV] END ......................max_depth=10, n_estimators=50; total time=   0.1s\n",
      "[CV] END ......................max_depth=10, n_estimators=50; total time=   0.1s\n",
      "[CV] END ......................max_depth=10, n_estimators=50; total time=   0.2s\n",
      "[CV] END .....................max_depth=10, n_estimators=100; total time=   0.2s\n",
      "[CV] END .....................max_depth=10, n_estimators=100; total time=   0.2s\n",
      "[CV] END .....................max_depth=10, n_estimators=100; total time=   0.2s\n",
      "[CV] END .....................max_depth=10, n_estimators=100; total time=   0.3s\n",
      "[CV] END .....................max_depth=10, n_estimators=100; total time=   0.2s\n",
      "[CV] END .....................max_depth=10, n_estimators=200; total time=   0.5s\n",
      "[CV] END .....................max_depth=10, n_estimators=200; total time=   0.4s\n",
      "[CV] END .....................max_depth=10, n_estimators=200; total time=   0.5s\n",
      "[CV] END .....................max_depth=10, n_estimators=200; total time=   0.5s\n",
      "[CV] END .....................max_depth=10, n_estimators=200; total time=   0.5s\n",
      "[CV] END .....................max_depth=10, n_estimators=300; total time=   0.7s\n",
      "[CV] END .....................max_depth=10, n_estimators=300; total time=   0.8s\n",
      "[CV] END .....................max_depth=10, n_estimators=300; total time=   0.7s\n",
      "[CV] END .....................max_depth=10, n_estimators=300; total time=   0.7s\n",
      "[CV] END .....................max_depth=10, n_estimators=300; total time=   0.7s\n",
      "[CV] END .....................max_depth=10, n_estimators=400; total time=   1.0s\n",
      "[CV] END .....................max_depth=10, n_estimators=400; total time=   1.0s\n",
      "[CV] END .....................max_depth=10, n_estimators=400; total time=   0.9s\n",
      "[CV] END .....................max_depth=10, n_estimators=400; total time=   0.9s\n",
      "[CV] END .....................max_depth=10, n_estimators=400; total time=   0.9s\n",
      "[CV] END .....................max_depth=10, n_estimators=500; total time=   1.5s\n",
      "[CV] END .....................max_depth=10, n_estimators=500; total time=   1.2s\n",
      "[CV] END .....................max_depth=10, n_estimators=500; total time=   1.6s\n",
      "[CV] END .....................max_depth=10, n_estimators=500; total time=   1.5s\n",
      "[CV] END .....................max_depth=10, n_estimators=500; total time=   1.1s\n",
      "[CV] END .....................max_depth=10, n_estimators=600; total time=   1.9s\n",
      "[CV] END .....................max_depth=10, n_estimators=600; total time=   1.3s\n",
      "[CV] END .....................max_depth=10, n_estimators=600; total time=   1.5s\n",
      "[CV] END .....................max_depth=10, n_estimators=600; total time=   1.4s\n",
      "[CV] END .....................max_depth=10, n_estimators=600; total time=   1.5s\n",
      "[CV] END ......................max_depth=20, n_estimators=50; total time=   0.1s\n",
      "[CV] END ......................max_depth=20, n_estimators=50; total time=   0.2s\n",
      "[CV] END ......................max_depth=20, n_estimators=50; total time=   0.1s\n",
      "[CV] END ......................max_depth=20, n_estimators=50; total time=   0.1s\n",
      "[CV] END ......................max_depth=20, n_estimators=50; total time=   0.1s\n",
      "[CV] END .....................max_depth=20, n_estimators=100; total time=   0.3s\n",
      "[CV] END .....................max_depth=20, n_estimators=100; total time=   0.3s\n",
      "[CV] END .....................max_depth=20, n_estimators=100; total time=   0.2s\n",
      "[CV] END .....................max_depth=20, n_estimators=100; total time=   0.3s\n",
      "[CV] END .....................max_depth=20, n_estimators=100; total time=   0.2s\n",
      "[CV] END .....................max_depth=20, n_estimators=200; total time=   0.5s\n",
      "[CV] END .....................max_depth=20, n_estimators=200; total time=   0.5s\n",
      "[CV] END .....................max_depth=20, n_estimators=200; total time=   0.5s\n",
      "[CV] END .....................max_depth=20, n_estimators=200; total time=   0.5s\n",
      "[CV] END .....................max_depth=20, n_estimators=200; total time=   0.5s\n",
      "[CV] END .....................max_depth=20, n_estimators=300; total time=   0.7s\n",
      "[CV] END .....................max_depth=20, n_estimators=300; total time=   0.7s\n",
      "[CV] END .....................max_depth=20, n_estimators=300; total time=   0.8s\n",
      "[CV] END .....................max_depth=20, n_estimators=300; total time=   0.8s\n",
      "[CV] END .....................max_depth=20, n_estimators=300; total time=   0.8s\n",
      "[CV] END .....................max_depth=20, n_estimators=400; total time=   1.1s\n",
      "[CV] END .....................max_depth=20, n_estimators=400; total time=   1.1s\n",
      "[CV] END .....................max_depth=20, n_estimators=400; total time=   1.0s\n",
      "[CV] END .....................max_depth=20, n_estimators=400; total time=   1.2s\n",
      "[CV] END .....................max_depth=20, n_estimators=400; total time=   1.2s\n",
      "[CV] END .....................max_depth=20, n_estimators=500; total time=   1.5s\n",
      "[CV] END .....................max_depth=20, n_estimators=500; total time=   1.8s\n",
      "[CV] END .....................max_depth=20, n_estimators=500; total time=   1.9s\n",
      "[CV] END .....................max_depth=20, n_estimators=500; total time=   1.6s\n",
      "[CV] END .....................max_depth=20, n_estimators=500; total time=   1.2s\n",
      "[CV] END .....................max_depth=20, n_estimators=600; total time=   1.3s\n",
      "[CV] END .....................max_depth=20, n_estimators=600; total time=   1.4s\n",
      "[CV] END .....................max_depth=20, n_estimators=600; total time=   1.3s\n",
      "[CV] END .....................max_depth=20, n_estimators=600; total time=   1.3s\n",
      "[CV] END .....................max_depth=20, n_estimators=600; total time=   1.9s\n",
      "[CV] END ......................max_depth=30, n_estimators=50; total time=   0.2s\n",
      "[CV] END ......................max_depth=30, n_estimators=50; total time=   0.2s\n",
      "[CV] END ......................max_depth=30, n_estimators=50; total time=   0.2s\n",
      "[CV] END ......................max_depth=30, n_estimators=50; total time=   0.2s\n",
      "[CV] END ......................max_depth=30, n_estimators=50; total time=   0.2s\n",
      "[CV] END .....................max_depth=30, n_estimators=100; total time=   0.3s\n",
      "[CV] END .....................max_depth=30, n_estimators=100; total time=   0.3s\n",
      "[CV] END .....................max_depth=30, n_estimators=100; total time=   0.4s\n",
      "[CV] END .....................max_depth=30, n_estimators=100; total time=   0.3s\n",
      "[CV] END .....................max_depth=30, n_estimators=100; total time=   0.3s\n",
      "[CV] END .....................max_depth=30, n_estimators=200; total time=   0.7s\n",
      "[CV] END .....................max_depth=30, n_estimators=200; total time=   0.6s\n",
      "[CV] END .....................max_depth=30, n_estimators=200; total time=   0.7s\n",
      "[CV] END .....................max_depth=30, n_estimators=200; total time=   0.6s\n",
      "[CV] END .....................max_depth=30, n_estimators=200; total time=   0.6s\n",
      "[CV] END .....................max_depth=30, n_estimators=300; total time=   0.9s\n",
      "[CV] END .....................max_depth=30, n_estimators=300; total time=   0.8s\n",
      "[CV] END .....................max_depth=30, n_estimators=300; total time=   1.1s\n",
      "[CV] END .....................max_depth=30, n_estimators=300; total time=   1.1s\n",
      "[CV] END .....................max_depth=30, n_estimators=300; total time=   0.9s\n",
      "[CV] END .....................max_depth=30, n_estimators=400; total time=   1.0s\n",
      "[CV] END .....................max_depth=30, n_estimators=400; total time=   1.0s\n",
      "[CV] END .....................max_depth=30, n_estimators=400; total time=   1.4s\n",
      "[CV] END .....................max_depth=30, n_estimators=400; total time=   1.3s\n",
      "[CV] END .....................max_depth=30, n_estimators=400; total time=   1.4s\n",
      "[CV] END .....................max_depth=30, n_estimators=500; total time=   1.3s\n",
      "[CV] END .....................max_depth=30, n_estimators=500; total time=   1.1s\n",
      "[CV] END .....................max_depth=30, n_estimators=500; total time=   1.2s\n",
      "[CV] END .....................max_depth=30, n_estimators=500; total time=   1.2s\n",
      "[CV] END .....................max_depth=30, n_estimators=500; total time=   1.2s\n",
      "[CV] END .....................max_depth=30, n_estimators=600; total time=   1.5s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV] END .....................max_depth=30, n_estimators=600; total time=   1.5s\n",
      "[CV] END .....................max_depth=30, n_estimators=600; total time=   1.4s\n",
      "[CV] END .....................max_depth=30, n_estimators=600; total time=   1.4s\n",
      "[CV] END .....................max_depth=30, n_estimators=600; total time=   1.4s\n",
      "[CV] END ....................max_depth=None, n_estimators=50; total time=   0.1s\n",
      "[CV] END ....................max_depth=None, n_estimators=50; total time=   0.1s\n",
      "[CV] END ....................max_depth=None, n_estimators=50; total time=   0.1s\n",
      "[CV] END ....................max_depth=None, n_estimators=50; total time=   0.2s\n",
      "[CV] END ....................max_depth=None, n_estimators=50; total time=   0.1s\n",
      "[CV] END ...................max_depth=None, n_estimators=100; total time=   0.3s\n",
      "[CV] END ...................max_depth=None, n_estimators=100; total time=   0.3s\n",
      "[CV] END ...................max_depth=None, n_estimators=100; total time=   0.3s\n",
      "[CV] END ...................max_depth=None, n_estimators=100; total time=   0.4s\n",
      "[CV] END ...................max_depth=None, n_estimators=100; total time=   0.2s\n",
      "[CV] END ...................max_depth=None, n_estimators=200; total time=   0.8s\n",
      "[CV] END ...................max_depth=None, n_estimators=200; total time=   0.7s\n",
      "[CV] END ...................max_depth=None, n_estimators=200; total time=   0.5s\n",
      "[CV] END ...................max_depth=None, n_estimators=200; total time=   0.5s\n",
      "[CV] END ...................max_depth=None, n_estimators=200; total time=   0.6s\n",
      "[CV] END ...................max_depth=None, n_estimators=300; total time=   0.7s\n",
      "[CV] END ...................max_depth=None, n_estimators=300; total time=   0.8s\n",
      "[CV] END ...................max_depth=None, n_estimators=300; total time=   0.9s\n",
      "[CV] END ...................max_depth=None, n_estimators=300; total time=   0.7s\n",
      "[CV] END ...................max_depth=None, n_estimators=300; total time=   0.8s\n",
      "[CV] END ...................max_depth=None, n_estimators=400; total time=   1.0s\n",
      "[CV] END ...................max_depth=None, n_estimators=400; total time=   1.0s\n",
      "[CV] END ...................max_depth=None, n_estimators=400; total time=   1.0s\n",
      "[CV] END ...................max_depth=None, n_estimators=400; total time=   1.0s\n",
      "[CV] END ...................max_depth=None, n_estimators=400; total time=   1.0s\n",
      "[CV] END ...................max_depth=None, n_estimators=500; total time=   1.3s\n",
      "[CV] END ...................max_depth=None, n_estimators=500; total time=   1.2s\n",
      "[CV] END ...................max_depth=None, n_estimators=500; total time=   1.4s\n",
      "[CV] END ...................max_depth=None, n_estimators=500; total time=   1.4s\n",
      "[CV] END ...................max_depth=None, n_estimators=500; total time=   1.2s\n",
      "[CV] END ...................max_depth=None, n_estimators=600; total time=   2.1s\n",
      "[CV] END ...................max_depth=None, n_estimators=600; total time=   2.1s\n",
      "[CV] END ...................max_depth=None, n_estimators=600; total time=   1.9s\n",
      "[CV] END ...................max_depth=None, n_estimators=600; total time=   1.9s\n",
      "[CV] END ...................max_depth=None, n_estimators=600; total time=   2.2s\n",
      "Best parameters:  {'max_depth': 30, 'n_estimators': 600}\n",
      "Best score:  0.9698779388875236\n"
     ]
    }
   ],
   "source": [
    "\n",
    "# Define the random forest model\n",
    "best_RF_oversampled_model = RandomForestClassifier(random_state=42,class_weight=class_weights)\n",
    "\n",
    "# Define the parameter grid to search over\n",
    "param_grid = {\n",
    "    'n_estimators': [50, 100, 200,300,400,500,600],\n",
    "    'max_depth': [10, 20, 30, None],\n",
    "}\n",
    "\n",
    "# Define the grid search object\n",
    "grid_search = GridSearchCV(estimator=best_RF_oversampled_model, param_grid=param_grid, cv=5,verbose=2)\n",
    "\n",
    "# Fit the grid search object to the data\n",
    "grid_search.fit(np.concatenate([X_train,X_val], axis=0)\n",
    "                   ,np.concatenate([Y_train,Y_val],axis=0))\n",
    "\n",
    "# Print the best parameters and best score\n",
    "print(\"Best parameters: \", grid_search.best_params_)\n",
    "print(\"Best score: \", grid_search.best_score_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "id": "2f1aa571",
   "metadata": {},
   "outputs": [],
   "source": [
    "best_RF_oversampled_model_accuracy , repo = test_model(grid_search.best_estimator_\n",
    "                                                       ,X_test,Y_test,axis=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "id": "2b6c2e17",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9308943089430894\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       1.00      0.93      0.96        28\n",
      "           1       0.86      0.86      0.86        36\n",
      "           2       0.90      0.93      0.91        80\n",
      "           3       0.96      0.96      0.96       102\n",
      "\n",
      "    accuracy                           0.93       246\n",
      "   macro avg       0.93      0.92      0.92       246\n",
      "weighted avg       0.93      0.93      0.93       246\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(best_RF_oversampled_model_accuracy )\n",
    "print(repo)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "554806da",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}