{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Retrain LR Model — Clean Pipeline (v3)\n",
    "\n",
    "## Changes from v1\n",
    "1. **WHQ_WHD050**: Removed — redundant body composition measure, counterintuitive SHAP\n",
    "2. **BIX_BIDFAT**: Removed — redundant with BMI\n",
    "3. **BMX_BMXBMI**: Added — BMI replaces weight+fat (universal, unit-agnostic)\n",
    "4. **BPX_BPXPLS**: Added — Pulse/heart rate (clinically relevant to panic, non-redundant)\n",
    "5. **Remove IMQ_IMQ020**: Auto-filled with imputer median, not real user input\n",
    "6. **Remove DEMO_RIDAGEMN**: Training mean ~30y (std=5.5) causes disproportionate SHAP for anyone >35y\n",
    "7. **Remove BPX_PEASCTM1**: Not collected in form\n",
    "8. **Remove BMX_BMXTRI**: Not collected in form\n",
    "9. **Remove DEMO_INDFMINC**: Collinear with DEMO_INDFMPIR (both derived from income), causes contradictory SHAP\n",
    "\n",
    "## Pipeline\n",
    "Same rigorous pipeline as `panic_disorder_LR_optimization.ipynb`:\n",
    "- 9 candidate features → median imputation → 1:3 balance → 75/25 split → StandardScaler → RFECV → Grid Search → Threshold optimization\n",
    "\n",
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Setup and Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Libraries loaded successfully!\n",
      "Analysis started: 2026-03-17 17:54:07\n"
     ]
    }
   ],
   "source": [
    "# Core libraries\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "\n",
    "# Visualization\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "plt.style.use('seaborn-v0_8-whitegrid')\n",
    "\n",
    "# Sklearn\n",
    "from sklearn.model_selection import train_test_split, StratifiedKFold, cross_validate, GridSearchCV, learning_curve\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.feature_selection import RFECV\n",
    "from sklearn.impute import SimpleImputer\n",
    "\n",
    "# Metrics\n",
    "from sklearn.metrics import (\n",
    "    balanced_accuracy_score, precision_score, recall_score, f1_score,\n",
    "    roc_auc_score, average_precision_score, matthews_corrcoef,\n",
    "    confusion_matrix, roc_curve, precision_recall_curve, make_scorer\n",
    ")\n",
    "\n",
    "# Model persistence\n",
    "import joblib\n",
    "import os\n",
    "import json\n",
    "from datetime import datetime\n",
    "\n",
    "# Set random seed\n",
    "RANDOM_STATE = 42\n",
    "np.random.seed(RANDOM_STATE)\n",
    "\n",
    "# Create output directories\n",
    "os.makedirs('results_v2', exist_ok=True)\n",
    "os.makedirs('figures_v2', exist_ok=True)\n",
    "os.makedirs('models_v2', exist_ok=True)\n",
    "\n",
    "print(\"Libraries loaded successfully!\")\n",
    "print(f\"Analysis started: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "# PHASE 1: Load Data & Clean\n",
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1.1 Load Data and Define 9 Candidate Features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using 9 candidate features\n",
      "Removed: ['IMQ_IMQ020', 'DEMO_RIDAGEMN', 'BPX_PEASCTM1', 'BMX_BMXTRI', 'DEMO_INDFMINC', 'WHQ_WHD050', 'BIX_BIDFAT']\n",
      "Features: ['BPX_BPXDAR', 'MPQ_MPQ070', 'DEMO_INDFMPIR', 'BMX_BMXBMI', 'MCQ_MCQ250F', 'CVX_CVDEXSTS', 'BPQ_BPQ010', 'ALQ_ALQ150', 'BPX_BPXPLS']\n",
      "\n",
      "Dataset shape: (6581, 1819)\n"
     ]
    }
   ],
   "source": [
    "# Load dataset\n",
    "df = pd.read_csv('NHANES_panic_subset_with_target_modified.csv')\n",
    "\n",
    "# 9 candidate features (v3: replaced WHQ_WHD050+BIX_BIDFAT with BMX_BMXBMI+BPX_BPXPLS)\n",
    "FEATURES = [\n",
    "    'BPX_BPXDAR',      # Diastolic BP\n",
    "    'MPQ_MPQ070',      # Low back pain\n",
    "    'DEMO_INDFMPIR',   # Poverty-income ratio (adjusted by household size)\n",
    "    'BMX_BMXBMI',      # Body Mass Index (replaces weight + body fat)\n",
    "    'MCQ_MCQ250F',     # Family history HTA/stroke\n",
    "    'CVX_CVDEXSTS',    # Cardiovascular fitness\n",
    "    'BPQ_BPQ010',      # Told high blood pressure\n",
    "    'ALQ_ALQ150',      # Regular alcohol consumption\n",
    "    'BPX_BPXPLS',      # Pulse / heart rate (bpm)\n",
    "]\n",
    "\n",
    "REMOVED_FEATURES = [\n",
    "    'IMQ_IMQ020', 'DEMO_RIDAGEMN', 'BPX_PEASCTM1', 'BMX_BMXTRI',\n",
    "    'DEMO_INDFMINC', 'WHQ_WHD050', 'BIX_BIDFAT'\n",
    "]\n",
    "TARGET_COL = 'target'\n",
    "\n",
    "print(f\"Using {len(FEATURES)} candidate features\")\n",
    "print(f\"Removed: {REMOVED_FEATURES}\")\n",
    "print(f\"Features: {FEATURES}\")\n",
    "print(f\"\\nDataset shape: {df.shape}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1.2 Data Quality Check"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "=== BMX_BMXBMI (Body Mass Index) ===\n",
      "  Non-null: 6348/6581 (96.5%)\n",
      "  Mean: 27.8, Std: 7.1\n",
      "  Min: 14.4, Max: 65.0\n",
      "\n",
      "=== BPX_BPXPLS (Pulse / Heart Rate) ===\n",
      "  Non-null: 6229/6581 (94.7%)\n",
      "  Mean: 74.5, Std: 12.8\n",
      "  Min: 44.0, Max: 122.0\n",
      "\n",
      "=== Correlation with target ===\n",
      "  BPX_BPXDAR          : r=+0.0008, p=0.9496\n",
      "  MPQ_MPQ070          : r=-0.0910, p=0.0000\n",
      "  DEMO_INDFMPIR       : r=-0.0792, p=0.0000\n",
      "  BMX_BMXBMI          : r=+0.0315, p=0.0121\n",
      "  MCQ_MCQ250F         : r=-0.0766, p=0.0000\n",
      "  CVX_CVDEXSTS        : r=+0.1177, p=0.0000\n",
      "  BPQ_BPQ010          : r=-0.0500, p=0.0000\n",
      "  ALQ_ALQ150          : r=-0.0189, p=0.1735\n",
      "  BPX_BPXPLS          : r=+0.1005, p=0.0000\n"
     ]
    }
   ],
   "source": [
    "# Quick sanity check on new features\n",
    "print(\"=== BMX_BMXBMI (Body Mass Index) ===\")\n",
    "print(f\"  Non-null: {df['BMX_BMXBMI'].notna().sum()}/{len(df)} ({df['BMX_BMXBMI'].notna().mean()*100:.1f}%)\")\n",
    "print(f\"  Mean: {df['BMX_BMXBMI'].mean():.1f}, Std: {df['BMX_BMXBMI'].std():.1f}\")\n",
    "print(f\"  Min: {df['BMX_BMXBMI'].min():.1f}, Max: {df['BMX_BMXBMI'].max():.1f}\")\n",
    "\n",
    "print(\"\\n=== BPX_BPXPLS (Pulse / Heart Rate) ===\")\n",
    "print(f\"  Non-null: {df['BPX_BPXPLS'].notna().sum()}/{len(df)} ({df['BPX_BPXPLS'].notna().mean()*100:.1f}%)\")\n",
    "print(f\"  Mean: {df['BPX_BPXPLS'].mean():.1f}, Std: {df['BPX_BPXPLS'].std():.1f}\")\n",
    "print(f\"  Min: {df['BPX_BPXPLS'].min():.1f}, Max: {df['BPX_BPXPLS'].max():.1f}\")\n",
    "\n",
    "# Check for any suspicious values\n",
    "print(\"\\n=== Correlation with target ===\")\n",
    "from scipy.stats import pointbiserialr\n",
    "for f in FEATURES:\n",
    "    valid = df[[f, 'target']].dropna()\n",
    "    if len(valid) > 10:\n",
    "        corr, pval = pointbiserialr(valid['target'], valid[f])\n",
    "        print(f\"  {f:20s}: r={corr:+.4f}, p={pval:.4f}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1.3 Feature Availability & Missing Values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Available features: 9/9\n",
      "\n",
      "Missing values per feature:\n",
      "  BPX_BPXDAR: 429 (6.5%)\n",
      "  MPQ_MPQ070: 0 (0.0%)\n",
      "  DEMO_INDFMPIR: 423 (6.4%)\n",
      "  BMX_BMXBMI: 233 (3.5%)\n",
      "  MCQ_MCQ250F: 0 (0.0%)\n",
      "  CVX_CVDEXSTS: 84 (1.3%)\n",
      "  BPQ_BPQ010: 0 (0.0%)\n",
      "  ALQ_ALQ150: 1366 (20.8%)\n",
      "  BPX_BPXPLS: 352 (5.3%)\n"
     ]
    }
   ],
   "source": [
    "# Check feature availability and missing values\n",
    "available_features = [f for f in FEATURES if f in df.columns]\n",
    "missing_features = [f for f in FEATURES if f not in df.columns]\n",
    "\n",
    "print(f\"Available features: {len(available_features)}/{len(FEATURES)}\")\n",
    "if missing_features:\n",
    "    print(f\"WARNING - Missing: {missing_features}\")\n",
    "\n",
    "# Show missing values per feature\n",
    "print(\"\\nMissing values per feature:\")\n",
    "for feat in available_features:\n",
    "    missing = df[feat].isnull().sum()\n",
    "    pct = missing / len(df) * 100\n",
    "    print(f\"  {feat}: {missing} ({pct:.1f}%)\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1.4 Imputation & Balancing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Dataset after imputation:\n",
      "  Total samples: 6581\n",
      "  PD cases: 178\n",
      "  Controls: 6403\n",
      "  Imbalance ratio: 1:36.0\n",
      "\n",
      "  BMX_BMXBMI imputed median: 26.1\n",
      "  BPX_BPXPLS imputed median: 74.0 bpm\n"
     ]
    }
   ],
   "source": [
    "# Create subset and apply median imputation\n",
    "df_subset = df[available_features + [TARGET_COL]].copy()\n",
    "\n",
    "# Impute missing values with median\n",
    "imputer = SimpleImputer(strategy='median')\n",
    "X_imputed = pd.DataFrame(\n",
    "    imputer.fit_transform(df_subset[available_features]),\n",
    "    columns=available_features\n",
    ")\n",
    "y = df_subset[TARGET_COL].values\n",
    "\n",
    "print(f\"Dataset after imputation:\")\n",
    "print(f\"  Total samples: {len(X_imputed)}\")\n",
    "print(f\"  PD cases: {sum(y == 1)}\")\n",
    "print(f\"  Controls: {sum(y == 0)}\")\n",
    "print(f\"  Imbalance ratio: 1:{sum(y == 0) / sum(y == 1):.1f}\")\n",
    "\n",
    "# Verify imputed medians\n",
    "print(f\"\\n  BMX_BMXBMI imputed median: {X_imputed['BMX_BMXBMI'].median():.1f}\")\n",
    "print(f\"  BPX_BPXPLS imputed median: {X_imputed['BPX_BPXPLS'].median():.1f} bpm\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Balanced dataset (1:3 ratio):\n",
      "  Total samples: 712\n",
      "  PD cases: 178\n",
      "  Controls: 534\n"
     ]
    }
   ],
   "source": [
    "# Create balanced dataset (1:3 ratio)\n",
    "\n",
    "# Separate classes\n",
    "X_pd = X_imputed[y == 1]\n",
    "y_pd = y[y == 1]\n",
    "X_ctrl = X_imputed[y == 0]\n",
    "y_ctrl = y[y == 0]\n",
    "\n",
    "n_pd = len(y_pd)\n",
    "n_ctrl_sample = n_pd * 3  # 1:3 ratio\n",
    "\n",
    "# Random undersample controls\n",
    "np.random.seed(RANDOM_STATE)\n",
    "ctrl_indices = np.random.choice(len(X_ctrl), size=n_ctrl_sample, replace=False)\n",
    "X_ctrl_sampled = X_ctrl.iloc[ctrl_indices]\n",
    "y_ctrl_sampled = y_ctrl[ctrl_indices]\n",
    "\n",
    "# Combine\n",
    "X = pd.concat([X_pd, X_ctrl_sampled]).reset_index(drop=True)\n",
    "y = np.concatenate([y_pd, y_ctrl_sampled])\n",
    "\n",
    "# Shuffle\n",
    "shuffle_idx = np.random.permutation(len(X))\n",
    "X = X.iloc[shuffle_idx].reset_index(drop=True)\n",
    "y = y[shuffle_idx]\n",
    "\n",
    "print(f\"Balanced dataset (1:3 ratio):\")\n",
    "print(f\"  Total samples: {len(X)}\")\n",
    "print(f\"  PD cases: {sum(y == 1)}\")\n",
    "print(f\"  Controls: {sum(y == 0)}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1.5 Train/Test Split & Scaling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train set: 534 samples (PD: 134, Ctrl: 400)\n",
      "Test set: 178 samples (PD: 44, Ctrl: 134)\n",
      "\n",
      "Features standardized.\n"
     ]
    }
   ],
   "source": [
    "# Train-test split (stratified)\n",
    "X_train, X_test, y_train, y_test = train_test_split(\n",
    "    X, y, test_size=0.25, random_state=RANDOM_STATE, stratify=y\n",
    ")\n",
    "\n",
    "print(f\"Train set: {len(X_train)} samples (PD: {sum(y_train == 1)}, Ctrl: {sum(y_train == 0)})\")\n",
    "print(f\"Test set: {len(X_test)} samples (PD: {sum(y_test == 1)}, Ctrl: {sum(y_test == 0)})\")\n",
    "\n",
    "# Standardize features\n",
    "scaler = StandardScaler()\n",
    "X_train_scaled = scaler.fit_transform(X_train)\n",
    "X_test_scaled = scaler.transform(X_test)\n",
    "\n",
    "print(\"\\nFeatures standardized.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "# PHASE 2: Baseline & Feature Selection (RFECV)\n",
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2.1 Baseline Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "============================================================\n",
      "BASELINE LOGISTIC REGRESSION (9 features, all candidates)\n",
      "============================================================\n",
      "Balanced Accuracy: 0.6825\n",
      "Recall: 0.4545\n",
      "Precision: 0.6250\n",
      "F1 Score: 0.5263\n",
      "ROC-AUC: 0.8262\n"
     ]
    }
   ],
   "source": [
    "# Train baseline logistic regression with all 9 features\n",
    "lr_baseline = LogisticRegression(random_state=RANDOM_STATE, max_iter=1000)\n",
    "lr_baseline.fit(X_train_scaled, y_train)\n",
    "\n",
    "# Baseline predictions\n",
    "y_pred_baseline = lr_baseline.predict(X_test_scaled)\n",
    "y_proba_baseline = lr_baseline.predict_proba(X_test_scaled)[:, 1]\n",
    "\n",
    "# Baseline metrics\n",
    "print(\"=\" * 60)\n",
    "print(\"BASELINE LOGISTIC REGRESSION (9 features, all candidates)\")\n",
    "print(\"=\" * 60)\n",
    "print(f\"Balanced Accuracy: {balanced_accuracy_score(y_test, y_pred_baseline):.4f}\")\n",
    "print(f\"Recall: {recall_score(y_test, y_pred_baseline):.4f}\")\n",
    "print(f\"Precision: {precision_score(y_test, y_pred_baseline):.4f}\")\n",
    "print(f\"F1 Score: {f1_score(y_test, y_pred_baseline):.4f}\")\n",
    "print(f\"ROC-AUC: {roc_auc_score(y_test, y_proba_baseline):.4f}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Feature Importance (by coefficient magnitude):\n",
      "      Feature  Coefficient  Abs_Coefficient\n",
      " CVX_CVDEXSTS     0.654985         0.654985\n",
      "DEMO_INDFMPIR    -0.611877         0.611877\n",
      "  MCQ_MCQ250F    -0.398057         0.398057\n",
      "   ALQ_ALQ150    -0.390770         0.390770\n",
      "   MPQ_MPQ070    -0.369092         0.369092\n",
      "   BPQ_BPQ010    -0.319855         0.319855\n",
      "   BPX_BPXPLS     0.307324         0.307324\n",
      "   BMX_BMXBMI    -0.093507         0.093507\n",
      "   BPX_BPXDAR     0.072508         0.072508\n"
     ]
    }
   ],
   "source": [
    "# Feature importance (coefficients)\n",
    "coef_df = pd.DataFrame({\n",
    "    'Feature': available_features,\n",
    "    'Coefficient': lr_baseline.coef_[0],\n",
    "    'Abs_Coefficient': np.abs(lr_baseline.coef_[0])\n",
    "}).sort_values('Abs_Coefficient', ascending=False)\n",
    "\n",
    "print(\"\\nFeature Importance (by coefficient magnitude):\")\n",
    "print(coef_df.to_string(index=False))\n",
    "\n",
    "coef_df.to_csv('results_v2/feature_importance_baseline.csv', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2.2 RFECV Feature Selection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Running Recursive Feature Elimination...\n",
      "\n",
      "Optimal number of features: 9\n",
      "Optimal features: ['BPX_BPXDAR', 'MPQ_MPQ070', 'DEMO_INDFMPIR', 'BMX_BMXBMI', 'MCQ_MCQ250F', 'CVX_CVDEXSTS', 'BPQ_BPQ010', 'ALQ_ALQ150', 'BPX_BPXPLS']\n",
      "Dropped by RFECV: []\n",
      "Best CV score: 0.6429\n"
     ]
    }
   ],
   "source": [
    "# Recursive Feature Elimination with Cross-Validation\n",
    "print(\"Running Recursive Feature Elimination...\")\n",
    "\n",
    "rfecv = RFECV(\n",
    "    estimator=LogisticRegression(random_state=RANDOM_STATE, max_iter=1000),\n",
    "    step=1,\n",
    "    cv=StratifiedKFold(5),\n",
    "    scoring='balanced_accuracy',\n",
    "    n_jobs=-1\n",
    ")\n",
    "\n",
    "rfecv.fit(X_train_scaled, y_train)\n",
    "\n",
    "# Optimal features\n",
    "optimal_features = [f for f, selected in zip(available_features, rfecv.support_) if selected]\n",
    "dropped_features = [f for f, selected in zip(available_features, rfecv.support_) if not selected]\n",
    "\n",
    "print(f\"\\nOptimal number of features: {rfecv.n_features_}\")\n",
    "print(f\"Optimal features: {optimal_features}\")\n",
    "print(f\"Dropped by RFECV: {dropped_features}\")\n",
    "print(f\"Best CV score: {rfecv.cv_results_['mean_test_score'].max():.4f}\")\n",
    "\n",
    "# Save\n",
    "with open('results_v2/optimal_features.txt', 'w') as f:\n",
    "    for feat in optimal_features:\n",
    "        f.write(f\"{feat}\\n\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot RFECV scores\n",
    "n_features_range = range(1, len(available_features) + 1)\n",
    "mean_scores = rfecv.cv_results_['mean_test_score']\n",
    "std_scores = rfecv.cv_results_['std_test_score']\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(10, 6))\n",
    "ax.fill_between(n_features_range,\n",
    "                mean_scores - std_scores,\n",
    "                mean_scores + std_scores,\n",
    "                alpha=0.2, color='blue')\n",
    "ax.plot(n_features_range, mean_scores, 'o-', color='blue', linewidth=2)\n",
    "ax.axvline(x=rfecv.n_features_, color='red', linestyle='--', label=f'Optimal: {rfecv.n_features_} features')\n",
    "ax.set_xlabel('Number of Features')\n",
    "ax.set_ylabel('Balanced Accuracy (CV)')\n",
    "ax.set_title('RFECV — Feature Selection (v3, BMI+Pulse)')\n",
    "ax.legend()\n",
    "ax.grid(True, alpha=0.3)\n",
    "plt.tight_layout()\n",
    "plt.savefig('figures_v2/rfecv_scores.png', dpi=300, bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2.3 Compare Feature Sets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Comparing feature sets:\n",
      "============================================================\n",
      "all_9 (9 features): Bal.Acc=0.6414 (+/-0.0933), Recall=0.3654\n",
      "rfe_optimal (9 features): Bal.Acc=0.6414 (+/-0.0933), Recall=0.3654\n",
      "\n",
      "Best feature set: all_9 with 9 features\n",
      "Selected features: ['BPX_BPXDAR', 'MPQ_MPQ070', 'DEMO_INDFMPIR', 'BMX_BMXBMI', 'MCQ_MCQ250F', 'CVX_CVDEXSTS', 'BPQ_BPQ010', 'ALQ_ALQ150', 'BPX_BPXPLS']\n"
     ]
    }
   ],
   "source": [
    "# Define feature sets to test\n",
    "feature_sets = {\n",
    "    'all_9': available_features,\n",
    "    'rfe_optimal': optimal_features,\n",
    "}\n",
    "\n",
    "# Test each feature set\n",
    "print(\"\\nComparing feature sets:\")\n",
    "print(\"=\" * 60)\n",
    "\n",
    "feature_set_results = []\n",
    "for name, features in feature_sets.items():\n",
    "    X_temp_train = X_train[features]\n",
    "    scaler_temp = StandardScaler()\n",
    "    X_temp_scaled = scaler_temp.fit_transform(X_temp_train)\n",
    "\n",
    "    lr_temp = LogisticRegression(random_state=RANDOM_STATE, max_iter=1000)\n",
    "    cv_scores = cross_validate(\n",
    "        lr_temp, X_temp_scaled, y_train,\n",
    "        cv=StratifiedKFold(10, shuffle=True, random_state=RANDOM_STATE),\n",
    "        scoring=['balanced_accuracy', 'recall', 'precision']\n",
    "    )\n",
    "\n",
    "    result = {\n",
    "        'Feature_Set': name,\n",
    "        'N_Features': len(features),\n",
    "        'Balanced_Acc': cv_scores['test_balanced_accuracy'].mean(),\n",
    "        'Balanced_Acc_Std': cv_scores['test_balanced_accuracy'].std(),\n",
    "        'Recall': cv_scores['test_recall'].mean(),\n",
    "        'Precision': cv_scores['test_precision'].mean(),\n",
    "    }\n",
    "    feature_set_results.append(result)\n",
    "\n",
    "    print(f\"{name} ({len(features)} features): Bal.Acc={result['Balanced_Acc']:.4f} (+/-{result['Balanced_Acc_Std']:.4f}), Recall={result['Recall']:.4f}\")\n",
    "\n",
    "# Select best feature set\n",
    "feature_set_df = pd.DataFrame(feature_set_results)\n",
    "feature_set_df.to_csv('results_v2/feature_set_comparison.csv', index=False)\n",
    "\n",
    "best_set_name = feature_set_df.loc[feature_set_df['Balanced_Acc'].idxmax(), 'Feature_Set']\n",
    "best_features = feature_sets[best_set_name]\n",
    "print(f\"\\nBest feature set: {best_set_name} with {len(best_features)} features\")\n",
    "print(f\"Selected features: {best_features}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "# PHASE 3: Hyperparameter Optimization\n",
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3.1 Prepare Best Feature Set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using 9 features for optimization\n",
      "Features: ['BPX_BPXDAR', 'MPQ_MPQ070', 'DEMO_INDFMPIR', 'BMX_BMXBMI', 'MCQ_MCQ250F', 'CVX_CVDEXSTS', 'BPQ_BPQ010', 'ALQ_ALQ150', 'BPX_BPXPLS']\n"
     ]
    }
   ],
   "source": [
    "# Use best features from RFECV\n",
    "X_train_best = X_train[best_features]\n",
    "X_test_best = X_test[best_features]\n",
    "\n",
    "# Scale\n",
    "scaler_best = StandardScaler()\n",
    "X_train_best_scaled = scaler_best.fit_transform(X_train_best)\n",
    "X_test_best_scaled = scaler_best.transform(X_test_best)\n",
    "\n",
    "print(f\"Using {len(best_features)} features for optimization\")\n",
    "print(f\"Features: {best_features}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3.2 Comprehensive Grid Search"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Grid search with 240 combinations\n",
      "Fitting 10 folds for each of 240 candidates, totalling 2400 fits\n",
      "\n",
      "Best CV Balanced Accuracy: 0.7682\n",
      "Best Parameters: {'C': 0.05, 'class_weight': {0: 1, 1: 3}, 'max_iter': 1000, 'penalty': 'l2', 'solver': 'saga'}\n"
     ]
    }
   ],
   "source": [
    "# Define comprehensive parameter grid (same as v1)\n",
    "param_grid = {\n",
    "    'C': [0.001, 0.01, 0.05, 0.1, 0.5, 1, 5, 10, 50, 100],\n",
    "    'penalty': ['l1', 'l2'],\n",
    "    'solver': ['liblinear', 'saga'],\n",
    "    'class_weight': [None, 'balanced', {0: 1, 1: 3}, {0: 1, 1: 5}, {0: 1, 1: 7}, {0: 1, 1: 10}],\n",
    "    'max_iter': [1000],\n",
    "}\n",
    "\n",
    "n_combos = len(param_grid['C']) * len(param_grid['penalty']) * len(param_grid['solver']) * len(param_grid['class_weight'])\n",
    "print(f\"Grid search with {n_combos} combinations\")\n",
    "\n",
    "# Run grid search\n",
    "grid_search = GridSearchCV(\n",
    "    estimator=LogisticRegression(random_state=RANDOM_STATE),\n",
    "    param_grid=param_grid,\n",
    "    cv=StratifiedKFold(10, shuffle=True, random_state=RANDOM_STATE),\n",
    "    scoring='balanced_accuracy',\n",
    "    n_jobs=-1,\n",
    "    verbose=1,\n",
    "    return_train_score=True\n",
    ")\n",
    "\n",
    "grid_search.fit(X_train_best_scaled, y_train)\n",
    "\n",
    "print(f\"\\nBest CV Balanced Accuracy: {grid_search.best_score_:.4f}\")\n",
    "print(f\"Best Parameters: {grid_search.best_params_}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Top 10 configurations:\n",
      "                                                                                                 params  mean_test_score  std_test_score  mean_train_score\n",
      "59       {'C': 0.05, 'class_weight': {0: 1, 1: 3}, 'max_iter': 1000, 'penalty': 'l2', 'solver': 'saga'}         0.768214        0.071003          0.780396\n",
      "79          {'C': 0.1, 'class_weight': 'balanced', 'max_iter': 1000, 'penalty': 'l2', 'solver': 'saga'}         0.768214        0.071003          0.781365\n",
      "83        {'C': 0.1, 'class_weight': {0: 1, 1: 3}, 'max_iter': 1000, 'penalty': 'l2', 'solver': 'saga'}         0.768214        0.071003          0.781500\n",
      "130    {'C': 1, 'class_weight': {0: 1, 1: 3}, 'max_iter': 1000, 'penalty': 'l2', 'solver': 'liblinear'}         0.768214        0.072423          0.782195\n",
      "128    {'C': 1, 'class_weight': {0: 1, 1: 3}, 'max_iter': 1000, 'penalty': 'l1', 'solver': 'liblinear'}         0.768214        0.072423          0.780389\n",
      "102    {'C': 0.5, 'class_weight': 'balanced', 'max_iter': 1000, 'penalty': 'l2', 'solver': 'liblinear'}         0.768214        0.072423          0.781084\n",
      "126      {'C': 1, 'class_weight': 'balanced', 'max_iter': 1000, 'penalty': 'l2', 'solver': 'liblinear'}         0.768214        0.072423          0.782334\n",
      "124      {'C': 1, 'class_weight': 'balanced', 'max_iter': 1000, 'penalty': 'l1', 'solver': 'liblinear'}         0.768214        0.072423          0.779972\n",
      "106  {'C': 0.5, 'class_weight': {0: 1, 1: 3}, 'max_iter': 1000, 'penalty': 'l2', 'solver': 'liblinear'}         0.768214        0.072423          0.781084\n",
      "107       {'C': 0.5, 'class_weight': {0: 1, 1: 3}, 'max_iter': 1000, 'penalty': 'l2', 'solver': 'saga'}         0.768214        0.072423          0.782334\n"
     ]
    }
   ],
   "source": [
    "# Save grid search results\n",
    "grid_results_df = pd.DataFrame(grid_search.cv_results_)\n",
    "grid_results_df = grid_results_df.sort_values('rank_test_score')\n",
    "grid_results_df.to_csv('results_v2/grid_search_results.csv', index=False)\n",
    "\n",
    "# Show top 10 configurations\n",
    "print(\"\\nTop 10 configurations:\")\n",
    "print(grid_results_df[['params', 'mean_test_score', 'std_test_score', 'mean_train_score']].head(10).to_string())\n",
    "\n",
    "# Save best parameters\n",
    "params_to_save = grid_search.best_params_.copy()\n",
    "if isinstance(params_to_save.get('class_weight'), dict):\n",
    "    params_to_save['class_weight'] = str(params_to_save['class_weight'])\n",
    "with open('results_v2/best_parameters.json', 'w') as f:\n",
    "    json.dump(params_to_save, f, indent=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Overfitting check:\n",
      "  Train score: 0.7804\n",
      "  CV score: 0.7682\n",
      "  Gap: 0.0122\n",
      "  OK: No severe overfitting.\n"
     ]
    }
   ],
   "source": [
    "# Check for overfitting\n",
    "best_idx = grid_search.best_index_\n",
    "train_score = grid_search.cv_results_['mean_train_score'][best_idx]\n",
    "test_score = grid_search.cv_results_['mean_test_score'][best_idx]\n",
    "gap = train_score - test_score\n",
    "\n",
    "print(f\"Overfitting check:\")\n",
    "print(f\"  Train score: {train_score:.4f}\")\n",
    "print(f\"  CV score: {test_score:.4f}\")\n",
    "print(f\"  Gap: {gap:.4f}\")\n",
    "\n",
    "if gap > 0.05:\n",
    "    print(\"  WARNING: Potential overfitting detected!\")\n",
    "else:\n",
    "    print(\"  OK: No severe overfitting.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "# PHASE 4: Threshold Optimization\n",
    "---"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Threshold Analysis:\n",
      " Threshold  Balanced_Accuracy   Recall  Precision  Specificity       F1  TP  FP  TN  FN\n",
      "      0.10           0.529512 0.954545   0.259259     0.104478 0.407767  42 120  14   2\n",
      "      0.15           0.566825 0.954545   0.276316     0.179104 0.428571  42 110  24   2\n",
      "      0.20           0.611601 0.954545   0.300000     0.268657 0.456522  42  98  36   2\n",
      "      0.25           0.645014 0.931818   0.322835     0.358209 0.479532  41  86  48   3\n",
      "      0.30           0.693521 0.931818   0.359649     0.455224 0.518987  41  73  61   3\n",
      "      0.35           0.745760 0.931818   0.410000     0.559701 0.569444  41  59  75   3\n",
      "      0.40           0.794267 0.931818   0.471264     0.656716 0.625954  41  46  88   3\n",
      "      0.45           0.801391 0.886364   0.506494     0.716418 0.644628  39  38  96   5\n",
      "      0.50           0.823779 0.886364   0.549296     0.761194 0.678261  39  32 102   5\n",
      "      0.55           0.751696 0.727273   0.516129     0.776119 0.603774  32  30 104  12\n",
      "      0.60           0.736092 0.636364   0.560000     0.835821 0.595745  28  22 112  16\n",
      "      0.65           0.697931 0.522727   0.575000     0.873134 0.547619  23  17 117  21\n",
      "      0.70           0.682666 0.477273   0.583333     0.888060 0.525000  21  15 119  23\n",
      "      0.75           0.648406 0.386364   0.586207     0.910448 0.465753  17  12 122  27\n",
      "      0.80           0.568691 0.204545   0.500000     0.932836 0.290323   9   9 125  35\n",
      "      0.85           0.575984 0.181818   0.666667     0.970149 0.285714   8   4 130  36\n"
     ]
    }
   ],
   "source": [
    "# Get best model\n",
    "best_model = grid_search.best_estimator_\n",
    "\n",
    "# Get predicted probabilities\n",
    "y_proba = best_model.predict_proba(X_test_best_scaled)[:, 1]\n",
    "\n",
    "# Test thresholds\n",
    "thresholds = np.arange(0.1, 0.9, 0.05)\n",
    "threshold_results = []\n",
    "\n",
    "for thresh in thresholds:\n",
    "    y_pred_thresh = (y_proba >= thresh).astype(int)\n",
    "    tn, fp, fn, tp = confusion_matrix(y_test, y_pred_thresh).ravel()\n",
    "\n",
    "    result = {\n",
    "        'Threshold': thresh,\n",
    "        'Balanced_Accuracy': balanced_accuracy_score(y_test, y_pred_thresh),\n",
    "        'Recall': recall_score(y_test, y_pred_thresh),\n",
    "        'Precision': precision_score(y_test, y_pred_thresh, zero_division=0),\n",
    "        'Specificity': tn / (tn + fp) if (tn + fp) > 0 else 0,\n",
    "        'F1': f1_score(y_test, y_pred_thresh, zero_division=0),\n",
    "        'TP': tp, 'FP': fp, 'TN': tn, 'FN': fn\n",
    "    }\n",
    "    threshold_results.append(result)\n",
    "\n",
    "threshold_df = pd.DataFrame(threshold_results)\n",
    "print(\"Threshold Analysis:\")\n",
    "print(threshold_df.to_string(index=False))\n",
    "\n",
    "threshold_df.to_csv('results_v2/threshold_optimization.csv', index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Optimal threshold for screening: 0.50\n",
      "At this threshold:\n",
      "  Balanced Accuracy: 0.8238\n",
      "  Recall: 0.8864\n",
      "  Precision: 0.5493\n",
      "  F1: 0.6783\n"
     ]
    }
   ],
   "source": [
    "# Find optimal threshold for screening (prioritize recall >= 0.75)\n",
    "high_recall_df = threshold_df[threshold_df['Recall'] >= 0.75]\n",
    "\n",
    "if len(high_recall_df) > 0:\n",
    "    # Among high recall, choose best balanced accuracy\n",
    "    optimal_idx = high_recall_df['Balanced_Accuracy'].idxmax()\n",
    "    optimal_threshold = threshold_df.loc[optimal_idx, 'Threshold']\n",
    "else:\n",
    "    # Fall back to threshold that maximizes balanced accuracy\n",
    "    optimal_idx = threshold_df['Balanced_Accuracy'].idxmax()\n",
    "    optimal_threshold = threshold_df.loc[optimal_idx, 'Threshold']\n",
    "\n",
    "print(f\"\\nOptimal threshold for screening: {optimal_threshold:.2f}\")\n",
    "print(f\"At this threshold:\")\n",
    "print(f\"  Balanced Accuracy: {threshold_df.loc[optimal_idx, 'Balanced_Accuracy']:.4f}\")\n",
    "print(f\"  Recall: {threshold_df.loc[optimal_idx, 'Recall']:.4f}\")\n",
    "print(f\"  Precision: {threshold_df.loc[optimal_idx, 'Precision']:.4f}\")\n",
    "print(f\"  F1: {threshold_df.loc[optimal_idx, 'F1']:.4f}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot threshold curves\n",
    "fig, ax = plt.subplots(figsize=(10, 6))\n",
    "\n",
    "ax.plot(threshold_df['Threshold'], threshold_df['Recall'], 'b-', label='Recall', linewidth=2)\n",
    "ax.plot(threshold_df['Threshold'], threshold_df['Precision'], 'r-', label='Precision', linewidth=2)\n",
    "ax.plot(threshold_df['Threshold'], threshold_df['Balanced_Accuracy'], 'g-', label='Balanced Accuracy', linewidth=2)\n",
    "ax.plot(threshold_df['Threshold'], threshold_df['F1'], 'orange', label='F1 Score', linewidth=2)\n",
    "\n",
    "ax.axvline(x=optimal_threshold, color='k', linestyle='--', label=f'Optimal ({optimal_threshold:.2f})')\n",
    "\n",
    "ax.set_xlabel('Threshold')\n",
    "ax.set_ylabel('Score')\n",
    "ax.set_title('Threshold Optimization (v3 — BMI + Pulse)')\n",
    "ax.legend()\n",
    "ax.grid(True, alpha=0.3)\n",
    "ax.set_xlim([0.1, 0.9])\n",
    "ax.set_ylim([0, 1])\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.savefig('figures_v2/threshold_curve.png', dpi=300, bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "# PHASE 5: Final Evaluation\n",
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5.1 Final Cross-Validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "============================================================\n",
      "FINAL CROSS-VALIDATION RESULTS (10-fold)\n",
      "============================================================\n",
      "balanced_accuracy: 0.7682 (+/- 0.0710)\n",
      "recall: 0.8214 (+/- 0.1139)\n",
      "precision: 0.4930 (+/- 0.0683)\n",
      "f1: 0.6150 (+/- 0.0815)\n",
      "roc_auc: 0.8002 (+/- 0.0831)\n"
     ]
    }
   ],
   "source": [
    "# Final model with best parameters\n",
    "final_model = LogisticRegression(**grid_search.best_params_, random_state=RANDOM_STATE)\n",
    "\n",
    "# Rigorous 10-fold CV\n",
    "cv_results = cross_validate(\n",
    "    final_model,\n",
    "    X_train_best_scaled,\n",
    "    y_train,\n",
    "    cv=StratifiedKFold(10, shuffle=True, random_state=RANDOM_STATE),\n",
    "    scoring={\n",
    "        'balanced_accuracy': make_scorer(balanced_accuracy_score),\n",
    "        'recall': make_scorer(recall_score),\n",
    "        'precision': make_scorer(precision_score, zero_division=0),\n",
    "        'f1': make_scorer(f1_score),\n",
    "        'roc_auc': 'roc_auc',\n",
    "    },\n",
    "    return_train_score=True\n",
    ")\n",
    "\n",
    "print(\"=\" * 60)\n",
    "print(\"FINAL CROSS-VALIDATION RESULTS (10-fold)\")\n",
    "print(\"=\" * 60)\n",
    "for metric in ['balanced_accuracy', 'recall', 'precision', 'f1', 'roc_auc']:\n",
    "    test_scores = cv_results[f'test_{metric}']\n",
    "    print(f\"{metric}: {test_scores.mean():.4f} (+/- {test_scores.std():.4f})\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5.2 Test Set Evaluation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "============================================================\n",
      "FINAL TEST SET RESULTS (threshold=0.50)\n",
      "============================================================\n",
      "Balanced Accuracy: 0.8238\n",
      "Recall: 0.8864\n",
      "Precision: 0.5493\n",
      "Specificity: 0.7612\n",
      "F1 Score: 0.6783\n",
      "MCC: 0.5705\n",
      "ROC-AUC: 0.8258\n",
      "PR-AUC: 0.5996\n",
      "\n",
      "Confusion Matrix:\n",
      "  TP: 39, FP: 32\n",
      "  FN: 5, TN: 102\n"
     ]
    }
   ],
   "source": [
    "# Train final model on all training data\n",
    "final_model.fit(X_train_best_scaled, y_train)\n",
    "\n",
    "# Predict with optimal threshold\n",
    "y_test_proba = final_model.predict_proba(X_test_best_scaled)[:, 1]\n",
    "y_test_pred = (y_test_proba >= optimal_threshold).astype(int)\n",
    "\n",
    "# Calculate metrics\n",
    "tn, fp, fn, tp = confusion_matrix(y_test, y_test_pred).ravel()\n",
    "\n",
    "test_metrics = {\n",
    "    'Balanced Accuracy': balanced_accuracy_score(y_test, y_test_pred),\n",
    "    'Recall': recall_score(y_test, y_test_pred),\n",
    "    'Precision': precision_score(y_test, y_test_pred, zero_division=0),\n",
    "    'Specificity': tn / (tn + fp),\n",
    "    'F1 Score': f1_score(y_test, y_test_pred),\n",
    "    'MCC': matthews_corrcoef(y_test, y_test_pred),\n",
    "    'ROC-AUC': roc_auc_score(y_test, y_test_proba),\n",
    "    'PR-AUC': average_precision_score(y_test, y_test_proba),\n",
    "}\n",
    "\n",
    "print(\"\\n\" + \"=\" * 60)\n",
    "print(f\"FINAL TEST SET RESULTS (threshold={optimal_threshold:.2f})\")\n",
    "print(\"=\" * 60)\n",
    "for metric, value in test_metrics.items():\n",
    "    print(f\"{metric}: {value:.4f}\")\n",
    "\n",
    "print(f\"\\nConfusion Matrix:\")\n",
    "print(f\"  TP: {tp}, FP: {fp}\")\n",
    "print(f\"  FN: {fn}, TN: {tn}\")\n",
    "\n",
    "# Save results\n",
    "with open('results_v2/final_test_results.json', 'w') as f:\n",
    "    json.dump(test_metrics, f, indent=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5.3 Model Coefficients"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Final Model Coefficients:\n",
      "      Feature  Coefficient  Abs_Coefficient  Odds_Ratio\n",
      " CVX_CVDEXSTS     0.628246         0.628246    1.874320\n",
      "DEMO_INDFMPIR    -0.524009         0.524009    0.592142\n",
      "   ALQ_ALQ150    -0.403025         0.403025    0.668296\n",
      "  MCQ_MCQ250F    -0.355381         0.355381    0.700907\n",
      "   BPX_BPXPLS     0.329413         0.329413    1.390152\n",
      "   MPQ_MPQ070    -0.315533         0.315533    0.729400\n",
      "   BPQ_BPQ010    -0.304315         0.304315    0.737628\n",
      "   BMX_BMXBMI    -0.085668         0.085668    0.917899\n",
      "   BPX_BPXDAR     0.069489         0.069489    1.071961\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Extract final coefficients\n",
    "coef_final = pd.DataFrame({\n",
    "    'Feature': best_features,\n",
    "    'Coefficient': final_model.coef_[0],\n",
    "    'Abs_Coefficient': np.abs(final_model.coef_[0])\n",
    "}).sort_values('Abs_Coefficient', ascending=False)\n",
    "\n",
    "coef_final['Odds_Ratio'] = np.exp(coef_final['Coefficient'])\n",
    "\n",
    "print(\"\\nFinal Model Coefficients:\")\n",
    "print(coef_final.to_string(index=False))\n",
    "\n",
    "coef_final.to_csv('results_v2/final_model_coefficients.csv', index=False)\n",
    "\n",
    "# Plot coefficients\n",
    "fig, ax = plt.subplots(figsize=(10, 6))\n",
    "colors = ['green' if c > 0 else 'red' for c in coef_final['Coefficient']]\n",
    "ax.barh(coef_final['Feature'], coef_final['Coefficient'], color=colors, alpha=0.7)\n",
    "ax.set_xlabel('Coefficient')\n",
    "ax.set_ylabel('Feature')\n",
    "ax.set_title('LR Coefficients — v3 (BMI + Pulse)')\n",
    "ax.axvline(x=0, color='k', linewidth=0.5)\n",
    "plt.tight_layout()\n",
    "plt.savefig('figures_v2/coefficient_plot.png', dpi=300, bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5.4 ROC and PR Curves"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1400x500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# ROC and PR curves\n",
    "fig, axes = plt.subplots(1, 2, figsize=(14, 5))\n",
    "\n",
    "# ROC curve\n",
    "fpr, tpr, _ = roc_curve(y_test, y_test_proba)\n",
    "roc_auc = roc_auc_score(y_test, y_test_proba)\n",
    "\n",
    "axes[0].plot(fpr, tpr, 'b-', linewidth=2, label=f'v2 LR (AUC = {roc_auc:.3f})')\n",
    "axes[0].plot([0, 1], [0, 1], 'k--', linewidth=1, label='Random')\n",
    "axes[0].set_xlabel('False Positive Rate')\n",
    "axes[0].set_ylabel('True Positive Rate')\n",
    "axes[0].set_title('ROC Curve')\n",
    "axes[0].legend()\n",
    "axes[0].grid(True, alpha=0.3)\n",
    "\n",
    "# PR curve\n",
    "prec_curve, rec_curve, _ = precision_recall_curve(y_test, y_test_proba)\n",
    "pr_auc = average_precision_score(y_test, y_test_proba)\n",
    "\n",
    "axes[1].plot(rec_curve, prec_curve, 'b-', linewidth=2, label=f'v2 LR (AP = {pr_auc:.3f})')\n",
    "axes[1].set_xlabel('Recall')\n",
    "axes[1].set_ylabel('Precision')\n",
    "axes[1].set_title('Precision-Recall Curve')\n",
    "axes[1].legend()\n",
    "axes[1].grid(True, alpha=0.3)\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.savefig('figures_v2/roc_pr_curves_final.png', dpi=300, bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5.5 Confusion Matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x500 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Confusion matrix visualization\n",
    "fig, axes = plt.subplots(1, 2, figsize=(12, 5))\n",
    "\n",
    "cm = confusion_matrix(y_test, y_test_pred)\n",
    "\n",
    "# Raw counts\n",
    "sns.heatmap(cm, annot=True, fmt='d', cmap='Blues', ax=axes[0],\n",
    "            xticklabels=['Control', 'PD'], yticklabels=['Control', 'PD'])\n",
    "axes[0].set_xlabel('Predicted')\n",
    "axes[0].set_ylabel('Actual')\n",
    "axes[0].set_title('Confusion Matrix (Counts)')\n",
    "\n",
    "# Normalized\n",
    "cm_norm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
    "sns.heatmap(cm_norm, annot=True, fmt='.2%', cmap='Blues', ax=axes[1],\n",
    "            xticklabels=['Control', 'PD'], yticklabels=['Control', 'PD'])\n",
    "axes[1].set_xlabel('Predicted')\n",
    "axes[1].set_ylabel('Actual')\n",
    "axes[1].set_title('Confusion Matrix (Normalized)')\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.savefig('figures_v2/confusion_matrix_final.png', dpi=300, bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5.6 Learning Curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Learning curve\n",
    "train_sizes, train_scores, val_scores = learning_curve(\n",
    "    final_model,\n",
    "    X_train_best_scaled,\n",
    "    y_train,\n",
    "    cv=5,\n",
    "    scoring='balanced_accuracy',\n",
    "    train_sizes=np.linspace(0.1, 1.0, 10),\n",
    "    random_state=RANDOM_STATE,\n",
    "    n_jobs=-1\n",
    ")\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(10, 6))\n",
    "\n",
    "ax.fill_between(train_sizes,\n",
    "                train_scores.mean(axis=1) - train_scores.std(axis=1),\n",
    "                train_scores.mean(axis=1) + train_scores.std(axis=1),\n",
    "                alpha=0.1, color='blue')\n",
    "ax.fill_between(train_sizes,\n",
    "                val_scores.mean(axis=1) - val_scores.std(axis=1),\n",
    "                val_scores.mean(axis=1) + val_scores.std(axis=1),\n",
    "                alpha=0.1, color='orange')\n",
    "\n",
    "ax.plot(train_sizes, train_scores.mean(axis=1), 'o-', color='blue', label='Training Score')\n",
    "ax.plot(train_sizes, val_scores.mean(axis=1), 'o-', color='orange', label='Validation Score')\n",
    "\n",
    "ax.set_xlabel('Training Set Size')\n",
    "ax.set_ylabel('Balanced Accuracy')\n",
    "ax.set_title('Learning Curve — v2 (clean pipeline)')\n",
    "ax.legend(loc='lower right')\n",
    "ax.grid(True, alpha=0.3)\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.savefig('figures_v2/learning_curve_final.png', dpi=300, bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "# PHASE 6: Save Artefacts\n",
    "---"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "All artefacts saved to models_v2/:\n",
      "  - optimized_lr_model.pkl\n",
      "  - scaler_optimized.pkl\n",
      "  - imputer.pkl\n",
      "  - feature_names_optimized.txt\n",
      "  - model_config.json\n"
     ]
    }
   ],
   "source": [
    "# Save model and preprocessing\n",
    "joblib.dump(final_model, 'models_v2/optimized_lr_model.pkl')\n",
    "joblib.dump(scaler_best, 'models_v2/scaler_optimized.pkl')\n",
    "joblib.dump(imputer, 'models_v2/imputer.pkl')\n",
    "\n",
    "# Save feature names\n",
    "with open('models_v2/feature_names_optimized.txt', 'w') as f:\n",
    "    for feat in best_features:\n",
    "        f.write(f\"{feat}\\n\")\n",
    "\n",
    "# Save comprehensive model config\n",
    "model_config = {\n",
    "    'version': 'v3_bmi_pulse',\n",
    "    'date': datetime.now().isoformat(),\n",
    "    'candidate_features': available_features,\n",
    "    'selected_features': best_features,\n",
    "    'n_features': len(best_features),\n",
    "    'removed_features': REMOVED_FEATURES,\n",
    "    'changes_v3': 'Replaced WHQ_WHD050+BIX_BIDFAT with BMX_BMXBMI+BPX_BPXPLS',\n",
    "    'optimal_threshold': float(optimal_threshold),\n",
    "    'best_params': {k: str(v) if isinstance(v, dict) else v for k, v in grid_search.best_params_.items()},\n",
    "    'test_metrics': {k: float(v) for k, v in test_metrics.items()},\n",
    "    'imputer_features': available_features,\n",
    "    'imputer_strategy': 'median',\n",
    "}\n",
    "\n",
    "with open('models_v2/model_config.json', 'w') as f:\n",
    "    json.dump(model_config, f, indent=2)\n",
    "\n",
    "print(\"All artefacts saved to models_v2/:\")\n",
    "print(\"  - optimized_lr_model.pkl\")\n",
    "print(\"  - scaler_optimized.pkl\")\n",
    "print(\"  - imputer.pkl\")\n",
    "print(\"  - feature_names_optimized.txt\")\n",
    "print(\"  - model_config.json\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "# Final Summary\n",
    "---"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "================================================================================\n",
      "RETRAIN SUMMARY (v3 — BMI + Pulse)\n",
      "================================================================================\n",
      "\n",
      "CHANGES FROM v2:\n",
      "  1. Replaced WHQ_WHD050 (weight lbs) + BIX_BIDFAT (body fat kg) → BMX_BMXBMI (BMI)\n",
      "  2. Added BPX_BPXPLS (pulse/heart rate) — clinically relevant, non-redundant\n",
      "\n",
      "FINAL MODEL PERFORMANCE:\n",
      "  Balanced Accuracy: 0.8238\n",
      "  Recall (Sensitivity): 0.8864\n",
      "  Precision: 0.5493\n",
      "  Specificity: 0.7612\n",
      "  F1 Score: 0.6783\n",
      "  ROC-AUC: 0.8258\n",
      "\n",
      "BEST PARAMETERS:\n",
      "  {'C': 0.05, 'class_weight': {0: 1, 1: 3}, 'max_iter': 1000, 'penalty': 'l2', 'solver': 'saga'}\n",
      "\n",
      "OPTIMAL THRESHOLD: 0.50\n",
      "\n",
      "FEATURES USED: 9\n",
      "  ['BPX_BPXDAR', 'MPQ_MPQ070', 'DEMO_INDFMPIR', 'BMX_BMXBMI', 'MCQ_MCQ250F', 'CVX_CVDEXSTS', 'BPQ_BPQ010', 'ALQ_ALQ150', 'BPX_BPXPLS']\n",
      "\n",
      "COMPARISON:\n",
      "  v1: Bal.Acc 80.2%, Recall 90.9%, 11 features (with IMQ_IMQ020 + DEMO_RIDAGEMN)\n",
      "  v2: Bal.Acc 78.3%, Recall 86.4%, 7 features (clean, weight+fat)\n",
      "  v3: Bal.Acc 82.4%, Recall 88.6%, 9 features (BMI+Pulse)\n",
      "\n",
      "FILES GENERATED:\n",
      "  models_v2/optimized_lr_model.pkl\n",
      "  models_v2/scaler_optimized.pkl\n",
      "  models_v2/imputer.pkl\n",
      "  models_v2/model_config.json\n",
      "  models_v2/feature_names_optimized.txt\n",
      "  results_v2/*.csv, results_v2/*.json\n",
      "  figures_v2/*.png\n",
      "\n",
      "RESULT: EXCELLENT - Report as-is\n",
      "\n",
      "Analysis complete!\n"
     ]
    }
   ],
   "source": [
    "print(\"\\n\" + \"=\" * 80)\n",
    "print(\"RETRAIN SUMMARY (v3 — BMI + Pulse)\")\n",
    "print(\"=\" * 80)\n",
    "\n",
    "print(f\"\"\"\n",
    "CHANGES FROM v2:\n",
    "  1. Replaced WHQ_WHD050 (weight lbs) + BIX_BIDFAT (body fat kg) → BMX_BMXBMI (BMI)\n",
    "  2. Added BPX_BPXPLS (pulse/heart rate) — clinically relevant, non-redundant\n",
    "\n",
    "FINAL MODEL PERFORMANCE:\n",
    "  Balanced Accuracy: {test_metrics['Balanced Accuracy']:.4f}\n",
    "  Recall (Sensitivity): {test_metrics['Recall']:.4f}\n",
    "  Precision: {test_metrics['Precision']:.4f}\n",
    "  Specificity: {test_metrics['Specificity']:.4f}\n",
    "  F1 Score: {test_metrics['F1 Score']:.4f}\n",
    "  ROC-AUC: {test_metrics['ROC-AUC']:.4f}\n",
    "\n",
    "BEST PARAMETERS:\n",
    "  {grid_search.best_params_}\n",
    "\n",
    "OPTIMAL THRESHOLD: {optimal_threshold:.2f}\n",
    "\n",
    "FEATURES USED: {len(best_features)}\n",
    "  {best_features}\n",
    "\n",
    "COMPARISON:\n",
    "  v1: Bal.Acc 80.2%, Recall 90.9%, 11 features (with IMQ_IMQ020 + DEMO_RIDAGEMN)\n",
    "  v2: Bal.Acc 78.3%, Recall 86.4%, 7 features (clean, weight+fat)\n",
    "  v3: Bal.Acc {test_metrics['Balanced Accuracy']:.1%}, Recall {test_metrics['Recall']:.1%}, {len(best_features)} features (BMI+Pulse)\n",
    "\n",
    "FILES GENERATED:\n",
    "  models_v2/optimized_lr_model.pkl\n",
    "  models_v2/scaler_optimized.pkl\n",
    "  models_v2/imputer.pkl\n",
    "  models_v2/model_config.json\n",
    "  models_v2/feature_names_optimized.txt\n",
    "  results_v2/*.csv, results_v2/*.json\n",
    "  figures_v2/*.png\n",
    "\"\"\")\n",
    "\n",
    "# Decision tree\n",
    "ba = test_metrics['Balanced Accuracy']\n",
    "if ba >= 0.82:\n",
    "    print(\"RESULT: EXCELLENT - Report as-is\")\n",
    "elif ba >= 0.75:\n",
    "    print(\"RESULT: VERY GOOD - Clinically useful screening tool\")\n",
    "elif ba >= 0.70:\n",
    "    print(\"RESULT: ACCEPTABLE - Modest but interpretable screening tool\")\n",
    "else:\n",
    "    print(\"RESULT: Below target - May need further investigation\")\n",
    "\n",
    "print(\"\\nAnalysis complete!\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  Copied models_v2/optimized_lr_model.pkl → ../Panic.2/calculator_panic/models_v2/optimized_lr_model.pkl\n",
      "  Copied models_v2/scaler_optimized.pkl → ../Panic.2/calculator_panic/models_v2/scaler_optimized.pkl\n",
      "  Copied models_v2/imputer.pkl → ../Panic.2/calculator_panic/models_v2/imputer.pkl\n",
      "  Copied models_v2/model_config.json → ../Panic.2/calculator_panic/models_v2/model_config.json\n",
      "  Copied models_v2/feature_names_optimized.txt → ../Panic.2/calculator_panic/models_v2/feature_names_optimized.txt\n",
      "\n",
      "All artefacts copied to ../Panic.2/calculator_panic/models_v2/\n"
     ]
    }
   ],
   "source": [
    "# Copy artefacts to calculator_panic/models_v2/ for deployment\n",
    "import shutil\n",
    "\n",
    "calculator_models_dir = '../Panic.2/calculator_panic/models_v2/'\n",
    "os.makedirs(calculator_models_dir, exist_ok=True)\n",
    "\n",
    "files_to_copy = [\n",
    "    'models_v2/optimized_lr_model.pkl',\n",
    "    'models_v2/scaler_optimized.pkl',\n",
    "    'models_v2/imputer.pkl',\n",
    "    'models_v2/model_config.json',\n",
    "    'models_v2/feature_names_optimized.txt',\n",
    "]\n",
    "\n",
    "for src in files_to_copy:\n",
    "    dst = os.path.join(calculator_models_dir, os.path.basename(src))\n",
    "    shutil.copy2(src, dst)\n",
    "    print(f\"  Copied {src} → {dst}\")\n",
    "\n",
    "print(f\"\\nAll artefacts copied to {calculator_models_dir}\")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.10 (tensorflow)",
   "language": "python",
   "name": "tensorflow"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}