Update app.py

app.py

@@ -32,7 +32,7 @@ def create_embeddings(text):
     print("Embeddings created successfully.")
     return embeddings, sentences
 
-def generate_plot(query, pdf_file):
+def generate_plot_bokeh(query, pdf_file):
     logging.info("Generating plot.")
     # Generate embeddings for the query
     query_embedding = model.encode([query])[0]
@@ -79,36 +79,55 @@ def generate_plot(query, pdf_file):
     save(p)
     logging.info("Plot saved to file.")
     return temp_file.name
+import plotly.express as px
+import plotly.graph_objects as go
 
-def gradio_interface_o(pdf_file, query):
-    logging.info("Gradio interface called.")
-    plot_path = generate_plot(query, pdf_file)
-    with open(plot_path, "r") as f:
-        html_content = f.read()
-    logging.info("Returning HTML content.")
-    return html_content
-def gradio_interface(pdf_file, query):
-    logging.info("Gradio interface called.")
-    plot_path = generate_plot(query, pdf_file)
+def generate_plotly_figure(query, pdf_file):
+    logging.info("Generating plot with Plotly.")
+    # Generate embeddings for the query
+    query_embedding = model.encode([query])[0]
+    
+    # Process the PDF and create embeddings
+    text = process_pdf(pdf_file.name)
+    embeddings, sentences = create_embeddings(text)
+    
+    logging.info("Data prepared for UMAP.")
+    # Prepare the data for UMAP and visualization
+    all_embeddings = np.vstack([embeddings, query_embedding])
+    all_sentences = sentences + [query]
     
-    # Assuming 'plot_path' can be accessed via a URL 'plot_url'
-    plot_url = f"http://yourserver.com/path/to/{os.path.basename(plot_path)}"
-    logging.info(f"Plot URL: {plot_url}")
+    # UMAP transformation
+    umap_transform = umap.UMAP(n_neighbors=15, min_dist=0.0, n_components=2, random_state=42)
+    umap_embeddings = umap_transform.fit_transform(all_embeddings)
+    
+    logging.info("UMAP transformation completed.")
+    # Find the closest sentences to the query
+    distances = cosine_similarity([query_embedding], embeddings)[0]
+    closest_indices = distances.argsort()[-5:][::-1]  # Adjust the number as needed
+    
+    # Prepare data for plotting
+    colors = ['red' if i in closest_indices else 'blue' for i in range(len(sentences))]
+    fig = go.Figure()
+    fig.add_trace(go.Scatter(x=umap_embeddings[:-1, 0], y=umap_embeddings[:-1, 1], mode='markers',
+                             marker=dict(color=colors), text=all_sentences[:-1]))
+    
+    fig.update_layout(title="UMAP Projection of Sentences", xaxis_title="UMAP 1", yaxis_title="UMAP 2")
     
-    # Return an iframe pointing to the plot URL
-    iframe_html = f'<iframe src="{plot_url}" width="100%" height="700"></iframe>'
-    logging.info("Returning iframe HTML content.")
-    return iframe_html
+    logging.info("Plotly figure created successfully.")
+    return fig
 
-# Update the outputs to gr.HTML() if it's not already
+def gradio_interface(pdf_file, query):
+    logging.info("Gradio interface called.")
+    fig = generate_plotly_figure(query, pdf_file)
+    logging.info("Returning Plotly figure.")
+    return fig
 iface = gr.Interface(
     fn=gradio_interface,
     inputs=[gr.File(label="Upload PDF"), gr.Textbox(label="Query")],
-    outputs=gr.HTML(),
+    outputs=gr.Plot(),  # Updated to use gr.Plot() for Plotly figures
     title="PDF Content Visualizer",
     description="Upload a PDF and enter a query to visualize the content."
 )
 
-
 if __name__ == "__main__":
     iface.launch()