final_model.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Modules\n",
    "from final_model import *\n",
    "\n",
    "\n",
    "# Files \n",
    "x_train_csv = \"dataset/x_train.csv\"\n",
    "y_train_csv = \"dataset/y_train.csv\"\n",
    "x_test_csv = \"dataset/x_test.csv\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Final features selected**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "features_dict = {\n",
    "    # Binary variables\n",
    "    'BPMEDS': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "    'TOLDHI2': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "    'DIABETE3': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "    'HLTHPLN1': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "    'QLACTLM2': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "    'EXERANY2': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "    'ADDEPEV2': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "    'ASTHMA3': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "    'SMOKE100': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "    'CVDSTRK3': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "    'HAVARTH3': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "    'CHCCOPD1': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "    'SEX': {'type': 'binary', 'missing_values': [np.nan]},\n",
    "    'BPHIGH4': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "    'USEEQUIP': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "    'BLIND': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "    'DECIDE': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "    'DIFFWALK': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "    'DIFFDRES': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "    'DIFFALON': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "    'ALCDAY5': {'type': 'binary', 'missing_values': [np.nan]},\n",
    "    'CHECKUP1': {'type': 'binary', 'missing_values': [np.nan]},\n",
    "    'STRENGTH': {'type': 'binary', 'missing_values': [np.nan]},\n",
    "    'AVG_FRUITS_VEGS': {'type': 'binary', 'missing_values': [np.nan]},\n",
    "    '_RFBING5': {'type': 'binary', 'missing_values': [9, np.nan]},\n",
    "     '_AGE65YR': {'type': 'binary', 'missing_values': [ np.nan]},\n",
    "     '_RFDRHV5': {'type': 'binary', 'missing_values': [9, np.nan]},\n",
    "     '_TOTINDA': {'type': 'binary', 'missing_values': [9, np.nan]},\n",
    "     '_PAINDX1': {'type': 'binary', 'missing_values': [9, np.nan]},\n",
    "     '_PASTRNG': {'type': 'binary', 'missing_values': [9, np.nan]},\n",
    "     '_PASTAE1': {'type': 'binary', 'missing_values': [9, np.nan]},\n",
    "     '_FLSHOT6': {'type': 'binary', 'missing_values': [ 9,np.nan]},\n",
    "     '_PNEUMO2': {'type': 'binary', 'missing_values': [ 9,np.nan]},\n",
    "     '_AIDTST3': {'type': 'binary', 'missing_values': [9, np.nan]},\n",
    "     '_RFBMI5': {'type': 'binary', 'missing_values': [9, np.nan]},\n",
    "     '_HISPANC': {'type': 'binary', 'missing_values': [9, np.nan]},\n",
    "     '_CASTHM1': {'type': 'binary', 'missing_values': [9, np.nan]},\n",
    "     '_LTASTH1': {'type': 'binary', 'missing_values': [9, np.nan]},\n",
    "     '_RFCHOL': {'type': 'binary', 'missing_values': [9, np.nan]},\n",
    "     '_RFHYPE5': {'type': 'binary', 'missing_values': [9, np.nan]},\n",
    "     '_HCVU651': {'type': 'binary', 'missing_values': [9, np.nan]},\n",
    "     '_RFHLTH': {'type': 'binary', 'missing_values': [9, np.nan]},\n",
    "     'MEDCOST': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "     'BLOODCHO': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "     'ASTHNOW': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "     'CHCSCNCR': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "     'CHCCOPD1': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "     'CHCOCNCR': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "     'CHCKIDNY': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "     'VETERAN3': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "     'INTERNET': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "     'LMTJOIN3': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "     'ARTHDIS2': {'type': 'binary', 'missing_values': [7, 9, np.nan]},\n",
    "     \n",
    "\n",
    "\n",
    "\n",
    "    # Categorical variable\n",
    "    '_RACEGR3': {\n",
    "        'type': 'categorical',\n",
    "        'missing_values': [9, np.nan],\n",
    "        'categories': [1, 2, 3, 4, 5]\n",
    "    },\n",
    "\n",
    "    # Numeric variables\n",
    "    #'_DRNKWEK': {'type': 'numeric', 'missing_values': [99900, np.nan], 'range': (1, 6)},\n",
    "    '_PAREC1': {'type': 'numeric', 'missing_values': [9, np.nan], 'range': (1, 6)},\n",
    "    '_PA150R2': {'type': 'numeric', 'missing_values': [ 9,np.nan], 'range': (1, 6)},\n",
    "    '_PACAT1': {'type': 'numeric', 'missing_values': [9], 'range': (1, 6)},\n",
    "    'FC60_': {'type': 'numeric', 'missing_values': [ 99900], 'range': (1, 6)},\n",
    "    'MAXVO2_': {'type': 'numeric', 'missing_values': [ 99900], 'range': (1, 6)},\n",
    "    'EDUCA': {'type': 'numeric', 'missing_values': [9, np.nan], 'range': (1, 6)},  # Education level\n",
    "    '_INCOMG': {'type': 'numeric', 'missing_values': [9, np.nan], 'range': (1, 6)},\n",
    "    'BMI': {'type': 'numeric', 'missing_values': [7777, 9999, np.nan], 'range': (1, 4)},\n",
    "    'MENTHLTH': {'type': 'numeric', 'missing_values': [77, 99, np.nan], 'map_value': {88: 0}, 'range': (1, 30)},\n",
    "    '_AGEG5YR': {'type': 'numeric', 'missing_values': [14, np.nan], 'range': (1, 13)},\n",
    "    'PHYSHLTH': {'type': 'numeric', 'missing_values': [77, 99, np.nan], 'map_value': {88: 0}, 'range': (1, 30)}\n",
    "}\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Read x_train and y_train\n",
    "X_train, features = csv_to_array(\"dataset/x_train.csv\")\n",
    "Y_train, labels = csv_to_array(\"dataset/y_train.csv\")\n",
    "x_test , ids = csv_to_array(\"dataset/x_test.csv\", id=True)\n",
    "# Flatten y_train to convert into 1d array\n",
    "Y_train = Y_train.flatten()\n",
    "# Y modification\n",
    "Y_train = np.where(Y_train == -1, 0, Y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**K-fold cross validation results for final model**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\beaiz\\AppData\\Local\\Temp\\ipykernel_28756\\1134805394.py:1172: RuntimeWarning: Mean of empty slice\n",
      "  average_column = np.nanmean(selected_columns, axis=1)\n",
      "100%|██████████| 10/10 [10:21<00:00, 62.12s/it]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average F1 score: 0.7741522424360323\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "x_train, features = X_preprocessing(X_train, features)\n",
    "#X_test, _ = X_preprocessing(x_test, features)\n",
    "X_train_balanced, Y_train_balanced = balance_dataset(x_train, Y_train)\n",
    "X_train_transformed = transform_features_column_median(X_train_balanced, features_dict, features )\n",
    "max_iters=1500\n",
    "gamma= 0.001\n",
    "beta1=0.9\n",
    "beta2=0.97\n",
    "param_l2 = 0.001\n",
    "prob_threshold=0.52\n",
    "batch=4096\n",
    "decay_rate=0.96\n",
    "decay_steps=100\n",
    "K = 10\n",
    "# Run evaluation with current parameter combination\n",
    "F1_score = K_fold_cross_validation_f1_score(X_train_transformed, Y_train_balanced, K, \n",
    "    seed=5,\n",
    "    max_iters=max_iters,\n",
    "    gamma=gamma,\n",
    "    reg_norm= ['l2',param_l2],\n",
    "    prob_threshold=prob_threshold,\n",
    "    batch_size=batch,\n",
    "    beta1=beta1,\n",
    "    beta2=beta2,\n",
    "    decay_rate=decay_rate, \n",
    "    decay_steps=decay_steps\n",
    ")\n",
    "\n",
    "print(f'Average F1 score: {F1_score}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Plot generation for visualization**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\beaiz\\AppData\\Local\\Temp\\ipykernel_28756\\1134805394.py:1172: RuntimeWarning: Mean of empty slice\n",
      "  average_column = np.nanmean(selected_columns, axis=1)\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "accuracy:0.7810181190681622 , F1 score: 0.7817712811693895\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x600 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "x_train, features = X_preprocessing(X_train, features)\n",
    "#X_test, _ = X_preprocessing(x_test, features)\n",
    "X_train_balanced, Y_train_balanced = balance_dataset(x_train, Y_train)\n",
    "X_train_transformed = transform_features_column_median(X_train_balanced, features_dict, features )\n",
    "#X_test_transformed = transform_features_column_median(X_test, features_dict_big, features )\n",
    "N = Y_train_balanced.shape[0]\n",
    "max_iters=1500\n",
    "gamma= 0.001\n",
    "beta1=0.9\n",
    "beta2=0.97\n",
    "reg_norm= ['l2', 0.001]\n",
    "prob_threshold=0.52\n",
    "batch_size=4096\n",
    "decay_rate=0.96\n",
    "decay_steps=100\n",
    "plot=True\n",
    "initial_w =  np.random.rand(np.shape(X_train_transformed)[1] ) * 0\n",
    "N = np.shape(Y_train_balanced)[0] # number of samples\n",
    "random_index = np.random.permutation(N)\n",
    "# Proportion trainig: 0.9; test: 0.1\n",
    "N_train = int(0.9 * N)\n",
    "training_index = random_index[:N_train]\n",
    "test_index = random_index[N_train:]\n",
    "x_tr = X_train_transformed[training_index]\n",
    "x_te = X_train_transformed[test_index]\n",
    "y_tr = Y_train_balanced[training_index]\n",
    "y_te = Y_train_balanced[test_index]\n",
    "\n",
    "w, losses_tr, losses_te = reg_logistic_regression(\n",
    "    y_tr, x_tr, y_te, x_te, initial_w, max_iters=max_iters,gamma= gamma,reg_norm= reg_norm,batch_size= batch_size, beta1=beta1, beta2=beta2,\n",
    "    decay_rate=decay_rate, decay_steps=decay_steps)\n",
    "\n",
    "# TEST \n",
    "#y_prob = sigmoid(X_test_transformed @ w)\n",
    "#y_pred =np.where(y_prob < prob_threshold, -1, 1) # if below threshold 0 otherwise 1\n",
    "\n",
    "# VALIDATION\n",
    "y_prob_tr = sigmoid(x_te @ w)\n",
    "y_pred_tr =np.where(y_prob_tr < prob_threshold, 0, 1) # if below threshold 0 otherwise 1\n",
    "accuracy, F1_score = F1_score_f(y_te, y_pred_tr,plot=plot)\n",
    "print(f'accuracy:{accuracy} , F1 score: {F1_score}')\n",
    "\n",
    "\n",
    "\n",
    "# Analyze feature distributions for FN and FP\n",
    "#analyze_feature_distribution(x_te, y_te, FN_indices, FP_indices)\n",
    "\n",
    "\n",
    "if plot:\n",
    "    plot_roc_curve(y_te, y_prob_tr)\n",
    "    plot_loss_iter(losses_train=losses_tr, losses_test=losses_te, max_iters=max_iters, gamma=gamma)\n",
    "    plot_prediction_vs_feature(x_te, y_te, w, feature_index=0)\n",
    "\n",
    "\n",
    "# Define the output filename\n",
    "name = \"submission.csv\"\n",
    "\n",
    "\n",
    "# Create the CSV submission file\n",
    "#create_csv_submission(ids, y_pred, name)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "ML_EPFL",
   "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": "3.8.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}