Update app.py

app.py

@@ -18,6 +18,7 @@ from docx.enum.text import WD_PARAGRAPH_ALIGNMENT
 import os
 import json
 from sklearn.linear_model import Ridge, Lasso, ElasticNet
+from sklearn.model_selection import KFold  # CORRECCIÓN IMPORTANTE
 from sklearn.model_selection import cross_val_score
 from sklearn.decomposition import PCA
 from sklearn.preprocessing import StandardScaler
@@ -650,15 +651,39 @@ class RSM_BoxBehnken:
                       f'{self.x1_name}_sq', f'{self.x2_name}_sq', f'{self.x3_name}_sq']]
         y = self.data[self.y_name]
         
-        # Calcular R² de validación cruzada
-        r2_scores = cross_val_score(self.model_simplified, X, y, cv=cv_folds)
+        # Crear los folds
+        kf = KFold(n_splits=cv_folds, shuffle=True, random_state=42)
         
-        # Calcular MSE de validación cruzada
-        mse_scores = -cross_val_score(
-            self.model_simplified, X, y, 
-            cv=cv_folds, 
-            scoring='neg_mean_squared_error'
-        )
+        # Almacenar los resultados
+        r2_scores = []
+        mse_scores = []
+        
+        # Realizar la validación cruzada
+        for train_index, test_index in kf.split(X):
+            X_train, X_test = X.iloc[train_index], X.iloc[test_index]
+            y_train, y_test = y.iloc[train_index], y.iloc[test_index]
+            
+            # Crear un DataFrame para el ajuste del modelo
+            train_data = X_train.copy()
+            train_data[self.y_name] = y_train
+            
+            # Ajustar el modelo en los datos de entrenamiento
+            formula = f'{self.y_name} ~ {self.x1_name} + {self.x2_name} + {self.x3_name} + ' \
+                      f'{self.x1_name}_sq + {self.x2_name}_sq + {self.x3_name}_sq'
+            model = smf.ols(formula, data=train_data).fit()
+            
+            # Predecir en los datos de prueba
+            y_pred = model.predict(X_test)
+            
+            # Calcular R²
+            ss_total = np.sum((y_test - np.mean(y_test))**2)
+            ss_residual = np.sum((y_test - y_pred)**2)
+            r2 = 1 - (ss_residual / ss_total)
+            r2_scores.append(r2)
+            
+            # Calcular MSE
+            mse = np.mean((y_test - y_pred)**2)
+            mse_scores.append(mse)
         
         # Crear tabla de resultados
         cv_results = pd.DataFrame({
@@ -668,7 +693,7 @@ class RSM_BoxBehnken:
         })
         
         # Agregar promedio
-        cv_results.loc['Promedio'] = ['Promedio', cv_results['R²'].mean(), cv_results['MSE'].mean()]
+        cv_results.loc['Promedio'] = ['Promedio', np.mean(r2_scores), np.mean(mse_scores)]
         
         return cv_results