Data leakage

import pandas as pd
import numpy as np
from sklearn.model_selection import train_test_split, GridSearchCV
from sklearn.svm import SVC
from sklearn.preprocessing import MinMaxScaler, OrdinalEncoder
from sklearn.pipeline import Pipeline
from sklearn.compose import ColumnTransformer
from sklearn.metrics import roc_auc_score

df = pd.read_csv("../data/heart.csv")

categoricas = ['Sex', 'ChestPainType', 'RestingECG', 'ExerciseAngina', 'ST_Slope']
numericas = ['Age', 'RestingBP', 'Cholesterol', 'FastingBS', 'MaxHR', 'Oldpeak']

X = df.drop("HeartDisease", axis=1)
y = df["HeartDisease"]

# Flujo CON fuga
encoder_leaky = OrdinalEncoder()
X_leaky = X.copy()
X_leaky[categoricas] = encoder_leaky.fit_transform(X_leaky[categoricas])
X_leaky["leaky_feature"] = y + np.random.normal(0, 0.01, size=len(y))

scaler = MinMaxScaler()
X_scaled = scaler.fit_transform(X_leaky)
X_train_l, X_test_l, y_train_l, y_test_l = train_test_split(X_scaled, y, test_size=0.2, random_state=42)

grid_l = GridSearchCV(SVC(probability=True), param_grid={"C": [0.1, 1, 10], "gamma": [0.01, 0.1]}, cv=5, scoring="roc_auc")
grid_l.fit(X_train_l, y_train_l)
auc_l = roc_auc_score(y_test_l, grid_l.predict_proba(X_test_l)[:, 1])

# Flujo SIN fuga
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

preprocessor = ColumnTransformer([
    ('cat', OrdinalEncoder(), categoricas),
    ('num', MinMaxScaler(), numericas)
])
pipe = Pipeline([("prep", preprocessor), ("svc", SVC(probability=True))])
param_grid = {"svc__C": [0.1, 1, 10], "svc__gamma": [0.01, 0.1]}
grid = GridSearchCV(pipe, param_grid, cv=5, scoring="roc_auc")
grid.fit(X_train, y_train)
auc = roc_auc_score(y_test, grid.predict_proba(X_test)[:, 1])

print(f"AUC con fuga: {auc_l:.3f}")
print(f"AUC sin fuga: {auc:.3f}")