006_roc_auc

roc_auc.py

from sklearn.metrics import roc_auc_score, roc_curve

y_test_int = y_test.replace({'Good': 1, 'Bad': 0})
auc_lr = roc_auc_score(y_test_int, probs_lr)
fpr_lr, tpr_lr, thresholds_lr = roc_curve(y_test_int, probs_lr)

auc_dt = roc_auc_score(y_test_int, probs_dt)
fpr_dt, tpr_dt, thresholds_dt = roc_curve(y_test_int, probs_dt)

auc_rf = roc_auc_score(y_test_int, probs_rf)
fpr_rf, tpr_rf, thresholds_rf = roc_curve(y_test_int, probs_rf)

auc_xg = roc_auc_score(y_test_int, probs_xg)
fpr_xg, tpr_xg, thresholds_xg = roc_curve(y_test_int, probs_xg)

plt.figure(figsize=(12, 7))
plt.plot(fpr_lr, tpr_lr, label=f'AUC (Logistic Regression) = {auc_lr:.2f}')
plt.plot(fpr_dt, tpr_dt, label=f'AUC (Decision Tree) = {auc_dt:.2f}')
plt.plot(fpr_rf, tpr_rf, label=f'AUC (Random Forests) = {auc_rf:.2f}')
plt.plot(fpr_xg, tpr_xg, label=f'AUC (XGBoost) = {auc_xg:.2f}')
plt.plot([0, 1], [0, 1], color='blue', linestyle='--', label='Baseline')
plt.title('ROC Curve', size=20)
plt.xlabel('False Positive Rate', size=14)
plt.ylabel('True Positive Rate', size=14)
plt.legend();