Update src/streamlit_app.py

src/streamlit_app.py

@@ -651,7 +651,15 @@ with tabs[4]:
             for fam in families_to_try:
                 log(f"Tuning family: {fam}")
                 st.caption(f"Tuning family: {fam}")
-                tuned_results.append(tune_family(fam, X, y, n_trials=max_trials))
+                result = tune_family(fam, X, y, n_trials=max_trials)
+                model_obj = result.get("model_obj")
+            
+                # ✅ Fix: ensure model is safe to access before fitting
+                if hasattr(model_obj, "estimators_"):
+                    delattr(model_obj, "estimators_")  # clear stale ref if any
+                result["model_obj"] = model_obj
+                tuned_results.append(result)
+
             # --- Leaderboard
             lb = pd.DataFrame([{"family": r["family"], "cv_r2": r["cv_score"], "params": r["best_params"]} for r in tuned_results])
             lb = lb.sort_values("cv_r2", ascending=False).reset_index(drop=True)
@@ -675,6 +683,13 @@ with tabs[4]:
             kf = KFold(n_splits=5, shuffle=True, random_state=42)
             base_models, oof_preds = [], pd.DataFrame(index=X_sel.index)
 
+            #  Prevent premature __len__() access on unfitted ensemble models
+            for r in tuned_results:
+                m = r.get("model_obj")
+                if m and hasattr(m, "__len__") and not hasattr(m, "estimators_"):
+                    m.__len__ = lambda self=m: 0
+
+
             for fam, entry in [(r["family"], r) for r in tuned_results if r.get("model_obj")]:
                 model_obj = entry["model_obj"]
                 oof = np.zeros(X_sel.shape[0])