Update app.py

app.py

@@ -188,25 +188,29 @@ def extract_feature_timeseries(video_path:str, frame_skip:int=FRAME_SKIP, debug:
 # Predictor & event logic
 # =====================================================
 def predict_scores(df):
-    feats=["red_ratio","green_ratio","red_diff","green_diff","z_red","z_green"]
-    X=df[feats].copy()
-    ag=ag_predictor()
-    available_models = ag.get_model_names()
+    feats = ["red_ratio", "green_ratio", "red_diff", "green_diff", "z_red", "z_green"]
+    X = df[feats].copy()
+    ag = ag_predictor()
+
+    # List available models and remove any FastAI learners that might break
+    available_models = ag.model_names
     safe_models = [m for m in available_models if "fastai" not in m.lower()]
-    print(f"[INFO] Using models: {safe_models}")
+    print(f"[INFO] Using safe models: {safe_models}")
+
     try:
-        proba=ag.predict_proba(X, models=safe_models)
-        if isinstance(proba,pd.DataFrame) and (1 in proba.columns):
+        proba = ag.predict_proba(X, models=safe_models)
+        if isinstance(proba, pd.DataFrame) and (1 in proba.columns):
             return proba[1]
     except Exception as e:
-        print("[WARN] AutoGluon fastai model failed, falling back to safe models only:", e)
-        proba=ag.predict_proba(X, models=safe_models)
-        if isinstance(proba,pd.DataFrame) and (1 in proba.columns):
+        print("[WARN] AutoGluon fastai model failed; retrying with safe models only:", e)
+        proba = ag.predict_proba(X, models=safe_models)
+        if isinstance(proba, pd.DataFrame) and (1 in proba.columns):
             return proba[1]
-    s=pd.Series(ag.predict(X, models=safe_models)).astype(float)
-    rng=(s.quantile(0.95)-s.quantile(0.05)) or 1.0
-    return ((s-s.quantile(0.05))/rng).clip(0,1)
 
+    # Fallback in case probability data isn't available
+    s = pd.Series(ag.predict(X, models=safe_models)).astype(float)
+    rng = (s.quantile(0.95) - s.quantile(0.05)) or 1.0
+    return ((s - s.quantile(0.05)) / rng).clip(0, 1)
 
 def pick_events(df,score,fps):
     z=rolling_z(score,45); strong=(z>4.0); keep=strong.rolling(3,min_periods=1).sum()>=2