Update src/streamlit_app.py

src/streamlit_app.py

@@ -870,13 +870,33 @@ with tabs[4]:
                         # SHAP direction analysis
                         expl = shap.TreeExplainer(model)
                         shap_vals = expl.shap_values(sample_X)
+                        
+                        # --- Normalize SHAP output structure (handles list, ndarray, or multi-dim cases) ---
+                        if isinstance(shap_vals, list):  # e.g., for multiclass models
+                            shap_vals = shap_vals[0]
+                        
+                        shap_vals = np.array(shap_vals)
+                        
+                        # If SHAP output has >2 dims, reduce to (n_samples, n_features)
+                        if shap_vals.ndim > 2:
+                            shap_vals = shap_vals.reshape(shap_vals.shape[0], -1)
+                        
+                        # Align SHAP features to DataFrame
+                        if shap_vals.shape[1] != sample_X.shape[1]:
+                            min_feats = min(shap_vals.shape[1], sample_X.shape[1])
+                            shap_vals = shap_vals[:, :min_feats]
+                            sample_X = sample_X.iloc[:, :min_feats]
+                        
+                        # Compute robust means
                         mean_abs = np.abs(shap_vals).mean(axis=0)
                         mean_sign = np.sign(shap_vals).mean(axis=0)
+                        
                         importance = pd.DataFrame({
                             "Feature": sample_X.columns,
                             "Mean |SHAP|": mean_abs,
                             "Mean SHAP Sign": mean_sign
                         }).sort_values("Mean |SHAP|", ascending=False)
+
                         
                         # Display Top 5 Drivers
                         st.markdown("### Top 5 Operational Drivers Influencing Target")