Upload app.py

app.py

@@ -74,17 +74,16 @@ def build_faiss_index(dataset: pd.DataFrame) -> Tuple[faiss.IndexFlatIP, np.ndar
     return index
 
 
-def compute_correlations_faiss(index, book_titles: List[str],
-                               target_book, num_recommendations) -> pd.DataFrame:
+def compute_correlations_faiss(index: faiss.IndexFlatIP, book_titles: List[str],
+                               target_book, ) -> pd.DataFrame:
     print(target_book, type(target_book))
-    emb = create_embedding([target_book])
+    emb = create_embedding([target_book[0]])
     # target_vector = book_titles.index(emb)
 
 
     # Perform the search
-    k = num_recommendations
+    k = len(book_titles)  # Search for all books
     similarities, I = index.search(emb.astype('float16'), k)
-    print(similarities, I)
 
     # # Reduce database and query vectors to 2D for visualization
     # pca = PCA(n_components=2)
@@ -132,7 +131,7 @@ def load_and_prepare_data():
     book_titles = dataset["Book-Title"]
 
 
-def recommend_books(target_book: str, num_recommendations: int = 10):
+def recommend_books(target_book: str, num_recommendations: int = 10) -> str:
     global dataset, faiss_index, normalized_data, book_titles
 
     if dataset is None or faiss_index is None or normalized_data is None or book_titles is None:
@@ -140,23 +139,22 @@ def recommend_books(target_book: str, num_recommendations: int = 10):
 
     target_book = target_book.lower()
     # Fuzzy match the input to the closest book title
-    closest_match, score, _ = process.extractOne(target_book, book_titles)
+    closest_match = process.extractOne(target_book, book_titles)
 
-    if score < 50:  # You can adjust this threshold
-        return f"No close match found for '{target_book}'. Please try a different title."
 
-    correlations = compute_correlations_faiss(faiss_index, list(dataset["Book-Title"]), closest_match, num_recommendations)
+    correlations = compute_correlations_faiss(faiss_index, book_titles, closest_match)
 
-    recommendations = dataset[dataset["Book-Title"][correlations['book']] != target_book].head(num_recommendations)
+    recommendations = correlations[correlations['book'] != target_book].head(num_recommendations)
 
+    print(recommendations['book'])
     result = f"Top {num_recommendations} recommendations for '{target_book}':\n\n"
     for i, (_, row) in enumerate(recommendations.iterrows(), 1):
         result += f"{i}. {row['book']} (Correlation: {row['corr']:.2f})\n"
-    return result
 
+    return result
 
-import gradio as gr
 
+# Create Gradio interface
 iface = gr.Interface(
     fn=recommend_books,
     inputs=[
@@ -165,8 +163,9 @@ iface = gr.Interface(
     ],
     outputs=gr.Textbox(label="Recommendations"),
     title="Book Recommender",
-    description="Enter a book title to get recommendations based on user ratings and book similarities.",
-    theme="light"  # Force light mode
+    description="Enter a book title to get recommendations based on user ratings and book similarities."
+    , theme=gr.themes.Base()
 )
 
-iface.launch()
+# Launch the app
+iface.launch()