final update with severity

app.py

@@ -98,21 +98,20 @@ class ModelPredictor:
                 probs.append(np.full(len(user_input), np.nan))
         return preds, probs
 
-    def evaluate_severity(self, count_twos: int) -> str:
+    def evaluate_severity(self, count_ones: int) -> str:
         """
-        Evaluate severity based on total # of '2' predictions across all labels.
-        In other words, if '2' is the 'Yes' or 'Has it' category, 
-        we can interpret how many 2's are present.
+        Evaluate severity based on how many labels predicted = 1.
+        The bigger the number of 1’s, the more severe the condition.
         """
-        # If you're treating "2" as the 'positive' class, let's rename "count_ones" to "count_twos".
-        if count_twos >= 13:
+        if count_ones >= 13:
             return "Mental Health Severity: Severe"
-        elif count_twos >= 9:
+        elif count_ones >= 9:
             return "Mental Health Severity: Moderate"
-        elif count_twos >= 5:
+        elif count_ones >= 5:
             return "Mental Health Severity: Low"
         else:
             return "Mental Health Severity: Very Low"
+    
 
 predictor = ModelPredictor(model_path, model_filenames)
 
@@ -333,9 +332,14 @@ def predict(
 
     # Flatten predictions into a single array
     all_preds = np.concatenate(preds)
-    # Count how many are "2" (assuming "2" is the 'positive' or 'yes' condition)
-    count_twos = np.sum(all_preds == 2)
-    severity_msg = predictor.evaluate_severity(count_twos)
+    # =====================================
+    # Count how many are "1" (the 'Yes' or 
+    #  more severe category in your new mapping)
+    # =====================================
+    count_ones = np.sum(all_preds == 1)
+    
+    # Evaluate severity using count_ones
+    severity_msg = predictor.evaluate_severity(count_ones)
 
     # 4) Summarize predictions (with probabilities)
     # Build label -> (pred_value, prob_value)