Update app.py

app.py

@@ -29,7 +29,7 @@ selected_days = date_range[1]
 # --- NEW: Subreddit input ---
 subreddit = st.text_input("Specify a subreddit (optional, e.g., 'Military' or 'worldnews'). Leave blank for all.", value="")
 
-query = st.text_input("Enter your topic or query:", value="ENTER VALUE HERE")
+query = st.text_input("Enter your topic or query:", value="US Army INDOPACOM")
 max_articles = st.slider("Number of news articles:", 5, 25, 12)
 
 # --- CLEANING FUNCTION ---
@@ -80,16 +80,23 @@ if st.button("Search"):
         # Optional: Show data table
         st.dataframe(results_df)
 
-        # Robust sentiment plot (handles missing/dirty data)
+        # Robust and crash-proof sentiment plot!
         sentiment_counts = results_df['sentiment'].value_counts(dropna=True)
-        # Drop NaN or empty-string sentiment labels
         sentiment_counts = sentiment_counts[~sentiment_counts.index.isna() & (sentiment_counts.index != '')]
+        sentiment_counts = sentiment_counts.rename(str)
+        sentiment_counts = sentiment_counts[sentiment_counts.index.str.strip() != '']
 
-        if not sentiment_counts.empty and len(sentiment_counts.index) == len(sentiment_counts.values):
+        sentiment_df = pd.DataFrame({
+            'Sentiment': sentiment_counts.index,
+            'Count': sentiment_counts.values
+        })
+
+        if not sentiment_df.empty and sentiment_df['Sentiment'].nunique() > 0:
             fig = px.bar(
-                x=sentiment_counts.index.tolist(),
-                y=sentiment_counts.values.tolist(),
-                labels={'x': 'Sentiment', 'y': 'Count'},
+                sentiment_df,
+                x='Sentiment',
+                y='Count',
+                labels={'Sentiment': 'Sentiment', 'Count': 'Count'},
                 title='Sentiment Distribution'
             )
             st.plotly_chart(fig, use_container_width=True)