Update app.py

app.py

@@ -17,7 +17,6 @@ nltk.download('punkt')
 nltk.download('stopwords')
 stop_words = set(stopwords.words('english'))
 
-# Global DataFrame to hold embeddings and metadata
 embed_df = pd.DataFrame()
 
 def analyze_bigrams(zip_file, perplexity):
@@ -25,20 +24,16 @@ def analyze_bigrams(zip_file, perplexity):
     if zip_file is None:
         return "Please upload a ZIP file containing .txt files.", None
 
-    # Wrap the binary zip file with BytesIO
     zip_stream = io.BytesIO(zip_file)
 
-    # Extract files into a temporary directory
     with tempfile.TemporaryDirectory() as tmpdir:
         with zipfile.ZipFile(zip_stream, 'r') as zip_ref:
             zip_ref.extractall(tmpdir)
 
-        # Find all .txt files
         txt_files = [os.path.join(tmpdir, f) for f in os.listdir(tmpdir) if f.endswith(".txt")]
         if not txt_files:
-            return "No .txt files found in the ZIP file.", None
+            return "No .txt files found.", None
 
-        # Read and tokenize texts
         all_texts = []
         for path in txt_files:
             with open(path, "r", encoding="utf-8") as f:
@@ -73,29 +68,29 @@ def analyze_bigrams(zip_file, perplexity):
                          size="count", template="plotly_white", title="Bigram t-SNE Projection")
         fig.update_layout(dragmode="lasso")
 
-        return "✅ Bigram analysis complete. Select points on the plot.", fig
+        return "✅ Bigram analysis complete. Use lasso to select points.", fig
 
-def generate_bar_plot(selected_points):
+def generate_bar_plot(events):
     global embed_df
-    if not selected_points or embed_df.empty:
+    if not events or embed_df.empty:
         return None
 
-    selected_indices = [point["pointIndex"] for point in selected_points["points"]]
+    selected_indices = [pt["pointIndex"] for pt in events]
     selected_df = embed_df.iloc[selected_indices]
 
-    bar_fig = px.bar(selected_df.sort_values("count", ascending=False),
-                     x="count", y="bigram", orientation="h",
-                     title="Selected Bigram Frequencies")
-    return bar_fig
+    fig = px.bar(selected_df.sort_values("count", ascending=False),
+                 x="count", y="bigram", orientation="h",
+                 title="Selected Bigram Frequencies")
+    return fig
 
 # Gradio UI
 with gr.Blocks() as demo:
-    gr.Markdown("## 📦 Upload a ZIP of .txt files to Analyze Bigrams with t-SNE and Select to Compare Frequencies")
+    gr.Markdown("## 📦 Upload ZIP of .txt files to Analyze Bigrams")
 
     zip_input = gr.File(label="Upload ZIP File of .txt Files", type="binary")
     perplexity_input = gr.Number(label="t-SNE Perplexity", value=30)
 
-    analyze_btn = gr.Button("Analyze Bigrams")
+    analyze_btn = gr.Button("Analyze")
     status = gr.Label()
     scatter_plot = gr.Plot()
     bar_plot = gr.Plot()
@@ -104,9 +99,8 @@ with gr.Blocks() as demo:
                       inputs=[zip_input, perplexity_input],
                       outputs=[status, scatter_plot])
 
-    scatter_plot.select(generate_bar_plot,
-                        inputs=[],
-                        outputs=bar_plot)
+    gr.Plot.update(scatter_plot, interactive=True)
+
+    gr.plotly_events(scatter_plot, select_event=True)(generate_bar_plot, inputs=None, outputs=bar_plot)
 
-# Enable public sharing (optional for Hugging Face Spaces)
 demo.launch(share=True)