Update app.py

app.py

@@ -1,63 +1,63 @@
-import gradio as gr
-import pickle
-import pandas as pd
-import os
-
-# Paths
-MODEL_DIR = "models"
-MOVIE_DATA_PATH = "data/movies.csv"  # adjust to your actual metadata file
-
-# Load models (choose what you want to demo)
-with open(os.path.join(MODEL_DIR, "recommender_svd_mf.pkl"), "rb") as f:
-    svd_model = pickle.load(f)
-
-# Load movie metadata
-movies_df = pd.read_csv(MOVIE_DATA_PATH)  # should include [movieId, title, poster_url, actors]
-
-def recommend(user_id, top_k=5):
-    """Generate top-k recommendations using SVD model."""
-    # Predict scores for all movies for this user
-    all_movie_ids = movies_df["movieId"].unique()
-    predictions = []
-    for mid in all_movie_ids:
-        try:
-            est = svd_model.predict(str(user_id), str(mid)).est
-            predictions.append((mid, est))
-        except Exception:
-            continue
-
-    # Sort and pick top_k
-    top_movies = sorted(predictions, key=lambda x: x[1], reverse=True)[:top_k]
-
-    # Build output
-    results = []
-    for mid, score in top_movies:
-        row = movies_df[movies_df["movieId"] == mid].iloc[0]
-        explanation = f"Because you liked movies with {row.get('actors', 'similar style')}."
-        results.append((row["title"], row.get("poster_url", None), explanation))
-    
-    return results
-
-def format_output(results):
-    titles = [r[0] for r in results]
-    posters = [r[1] for r in results if r[1] is not None]
-    explanations = [r[2] for r in results]
-    return titles, posters, explanations
-
-demo = gr.Interface(
-    fn=lambda user_id, k: format_output(recommend(user_id, k)),
-    inputs=[
-        gr.Number(label="User ID"),
-        gr.Slider(1, 10, value=5, step=1, label="Top-K")
-    ],
-    outputs=[
-        gr.Textbox(label="Recommended Movies"),
-        gr.Gallery(label="Posters").style(grid=[3], height="auto"),
-        gr.Textbox(label="Explanations")
-    ],
-    title="Movie Recommender System",
-    description="Enter your User ID to get top-K movie recommendations with posters and explanations."
-)
-
-if __name__ == "__main__":
-    demo.launch()
+import gradio as gr
+import pickle
+import pandas as pd
+import os
+
+# Paths
+MODEL_DIR = "models"
+MOVIE_DATA_PATH = "data/movies_metadata.csv"  # adjust to your actual metadata file
+
+# Load models (choose what you want to demo)
+with open(os.path.join(MODEL_DIR, "recommender_svd_mf.pkl"), "rb") as f:
+    svd_model = pickle.load(f)
+
+# Load movie metadata
+movies_df = pd.read_csv(MOVIE_DATA_PATH)  # should include [movieId, title, poster_url, actors]
+
+def recommend(user_id, top_k=5):
+    """Generate top-k recommendations using SVD model."""
+    # Predict scores for all movies for this user
+    all_movie_ids = movies_df["movieId"].unique()
+    predictions = []
+    for mid in all_movie_ids:
+        try:
+            est = svd_model.predict(str(user_id), str(mid)).est
+            predictions.append((mid, est))
+        except Exception:
+            continue
+
+    # Sort and pick top_k
+    top_movies = sorted(predictions, key=lambda x: x[1], reverse=True)[:top_k]
+
+    # Build output
+    results = []
+    for mid, score in top_movies:
+        row = movies_df[movies_df["movieId"] == mid].iloc[0]
+        explanation = f"Because you liked movies with {row.get('actors', 'similar style')}."
+        results.append((row["title"], row.get("poster_url", None), explanation))
+    
+    return results
+
+def format_output(results):
+    titles = [r[0] for r in results]
+    posters = [r[1] for r in results if r[1] is not None]
+    explanations = [r[2] for r in results]
+    return titles, posters, explanations
+
+demo = gr.Interface(
+    fn=lambda user_id, k: format_output(recommend(user_id, k)),
+    inputs=[
+        gr.Number(label="User ID"),
+        gr.Slider(1, 10, value=5, step=1, label="Top-K")
+    ],
+    outputs=[
+        gr.Textbox(label="Recommended Movies"),
+        gr.Gallery(label="Posters").style(grid=[3], height="auto"),
+        gr.Textbox(label="Explanations")
+    ],
+    title="Movie Recommender System",
+    description="Enter your User ID to get top-K movie recommendations with posters and explanations."
+)
+
+if __name__ == "__main__":
+    demo.launch()