Update app.py

app.py

@@ -98,18 +98,28 @@ def explain(batch: Optional[BatchInputData] = None, limit: int = 100):
     try:
         if batch:
             X = pd.DataFrame([item.dict() for item in batch.data])
+            source = "client batch"
         else:
             X = fetch_test_data(limit=limit)
+            source = f"NoCoDB (limit={limit})"
+
+        print(f"[DEBUG] SHAP explain called using {source} | shape={X.shape} | cols={list(X.columns)}")
+
+        # Ensure model is a pipeline
+        if hasattr(model, "named_steps"):
+            explainer = shap.Explainer(model.named_steps["classifier"], model.named_steps["preprocessor"].transform(X))
+        else:
+            explainer = shap.Explainer(model, X)
 
-        explainer = shap.Explainer(model, X)
         shap_values = explainer(X)
 
-        # Aggregate mean absolute SHAP value per feature
         shap_summary = pd.DataFrame({
             "feature": X.columns,
             "mean_abs_shap": np.abs(shap_values.values).mean(axis=0)
         }).sort_values("mean_abs_shap", ascending=False)
 
+        print(f"[DEBUG] SHAP summary created successfully with {len(shap_summary)} features.")
+
         return {
             "n_samples": len(X),
             "shap_summary": shap_summary.to_dict(orient="records")
@@ -117,6 +127,8 @@ def explain(batch: Optional[BatchInputData] = None, limit: int = 100):
 
     except Exception as e:
         import traceback
+        print("[ERROR] SHAP explain failed:", e)
+        print(traceback.format_exc())
         return {"error": str(e), "trace": traceback.format_exc()}
 
 # =====================================================