Add package-mode=false and plot final ROC curve

notebooks/exploratory.ipynb

@@ -281,7 +281,7 @@
     "# ROC-AUC score\n",
     "from sklearn.metrics import roc_auc_score, roc_curve\n",
     "\n",
-    "y_scores = np.max(y_proba, axis=1)\n",
+    "y_scores = y_proba[:, 1]\n",
     "\n",
     "roc_auc = roc_auc_score(y_test, y_scores)\n",
     "print(f\"ROC-AUC Score: {roc_auc:.4f}\")\n",
@@ -394,14 +394,15 @@
    "source": [
     "# Predict on test data\n",
     "y_pred_best = xgb_clf_best.predict(X_test)\n",
-    "y_proba_best = xgb_clf_best.predict_proba(X_test)[:, 1]\n",
+    "y_proba_best = xgb_clf_best.predict_proba(X_test)\n",
     "\n",
     "# Accuracy\n",
     "accuracy_best = accuracy_score(y_test, y_pred_best)\n",
     "print(f\"Optimized Accuracy: {accuracy_best:.4f}\")\n",
     "\n",
     "# ROC-AUC score\n",
-    "roc_auc_best = roc_auc_score(y_test, y_proba_best)\n",
+    "y_scores_best = y_proba_best[:, 1]\n",
+    "roc_auc_best = roc_auc_score(y_test, y_scores_best)\n",
     "print(f\"Optimized ROC-AUC Score: {roc_auc_best:.4f}\")"
    ]
   },
@@ -410,7 +411,17 @@
    "execution_count": null,
    "metadata": {},
    "outputs": [],
-   "source": []
+   "source": [
+    "# Plot ROC curve\n",
+    "fpr_best, tpr_best, thresholds_best = roc_curve(y_test, y_scores_best)\n",
+    "plt.plot(fpr_best, tpr_best, label=f\"AUC = {roc_auc_best:.4f}\")\n",
+    "plt.plot([0, 1], [0, 1], linestyle=\"--\")\n",
+    "plt.xlabel(\"False Positive Rate\")\n",
+    "plt.ylabel(\"True Positive Rate\")\n",
+    "plt.title(\"ROC Curve\")\n",
+    "plt.legend()\n",
+    "plt.show()"
+   ]
   }
  ],
  "metadata": {

pyproject.toml

@@ -1,4 +1,5 @@
 [tool.poetry]
+package-mode = false
 name = "credit-risk-modeling"
 version = "0.1.0"
 description = "Credit Risk Modeling using XGBoost"