Update app.py

app.py

@@ -44,16 +44,20 @@ def predict():
         # Get top-20 indices (highest probabilities)
         top_20_idx = np.argsort(proba)[::-1][:20]
 
+        # Normalize top-20 probs to sum to 100
+        top_20_probs = proba[top_20_idx]
+        top_20_probs_normalized = top_20_probs / top_20_probs.sum() * 100
+
         results = []
-        for idx in top_20_idx:
+        for rank, (idx, prob) in enumerate(zip(top_20_idx, top_20_probs_normalized), start=1):
             choice_code = target_encoder.inverse_transform([idx])[0]
-            prob = proba[idx] * 100  # convert to %
             row = choice_code_map.loc[int(choice_code)]
             college_name = row['College Name']
             results.append({
+                "rank": rank,
                 "choice_code": choice_code,
                 "college_name": college_name,
-                "probability_percent": round(float(prob), 2)  # show as %
+                "probability_percent": round(float(prob), 2)
             })
 
         return jsonify({"top_20_predictions": results})