Update app.py

app.py

@@ -174,8 +174,30 @@ def recommend_books(user_id, top_n):
     user_id = int(user_id)
 
     if user_id not in ratings_df['user_id'].unique():
-        return pd.DataFrame([["Error", f"User {user_id} not found.", ""]], 
-                            columns=["Book Cover", "Book Title", "Predicted Rating"])
+        
+        # Cold start fallback → Top-N highest-rated books
+        top_books = (
+            ratings_df.groupby('book_id')['rating']
+            .mean()
+            .reset_index()
+            .sort_values(by='rating', ascending=False)
+            .head(top_n)
+        )
+
+        recommendations = []
+        for _, row in top_books.iterrows():
+            book_id = row['book_id']
+            avg_rating = row['rating']
+            book_info = features_df.loc[features_df['book_id'] == book_id]
+            if book_info.empty:
+                continue
+            title = book_info['title'].values[0]
+            image_url = book_info['image_url'].values[0] if 'image_url' in book_info else ""
+            recommendations.append(
+                (f"<img src='{image_url}' width='60'>", title, round(avg_rating, 3))
+            )
+
+        return pd.DataFrame(recommendations, columns=["Book Cover", "Book Title", "Predicted
 
     rated_books = set(ratings_df[ratings_df['user_id'] == user_id]['book_id'])
     unseen_books = set(features_df['book_id']) - rated_books