Enhance VayuChat with retry functionality and improved UX

- Set DeepSeek-R1 as default model with GPT-OSS fallback
- Add retry button for regenerating responses with current model
- Expand question visibility from 20 to all 30 questions organized in categories:
* NCAP Funding & Policy Analysis
* Meteorology & Environmental Factors
* Population & Demographics
* Multi-Pollutant Analysis
* Other Analysis Questions
- Improve chat UI: increase message width from 70% to 85%, fix input width
- Update questions.txt with concrete policy-relevant queries
- Add seaborn to available libraries in system prompt

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <[email protected]>

app.py

@@ -158,7 +158,7 @@ st.markdown("""
     color: white;
     padding: 0.75rem 1rem;
     border-radius: 12px;
-    max-width: 70%;
+    max-width: 85%;
 }
 
 .user-info {
@@ -173,7 +173,7 @@ st.markdown("""
     color: #334155;
     padding: 0.75rem 1rem;
     border-radius: 12px;
-    max-width: 70%;
+    max-width: 85%;
 }
 
 .assistant-info {
@@ -235,6 +235,8 @@ st.markdown("""
     background: #ffffff;
     padding: 0.75rem 1rem;
     font-size: 1rem;
+    width: 100% !important;
+    max-width: none !important;
 }
 
 .stChatInput:focus {
@@ -563,9 +565,11 @@ if not available_models:
     st.error("No API keys available! Please set up your API keys in the .env file")
     st.stop()
 
-# Set GPT-OSS-120B as default if available
+# Set DeepSeek-R1 as default if available
 default_index = 0
-if "gpt-oss-120b" in available_models:
+if "deepseek-R1" in available_models:
+    default_index = available_models.index("deepseek-R1")
+elif "gpt-oss-120b" in available_models:
     default_index = available_models.index("gpt-oss-120b")
 
 # Simple header - just title and model selector
@@ -644,12 +648,45 @@ with st.sidebar:
     selected_prompt = None
     
     
-    for i, question in enumerate(questions[:20]):  # Show more questions including policy-focused ones
-        # Simple left-aligned buttons without icons for cleaner look
-        if st.button(question, key=f"sidebar_prompt_{i}", use_container_width=True, help=f"Click to analyze: {question}"):
-            if question != st.session_state.get("last_selected_prompt"):
-                selected_prompt = question
-                st.session_state.last_selected_prompt = question
+    # Show all questions but in a scrollable format
+    if len(questions) > 0:
+        st.markdown("**Select a question to analyze:**")
+        # Create expandable sections for better organization
+        with st.expander("📊 NCAP Funding & Policy Analysis", expanded=True):
+            for i, q in enumerate([q for q in questions if any(word in q.lower() for word in ['ncap', 'funding', 'investment', 'rupee'])]):
+                if st.button(q, key=f"ncap_q_{i}", use_container_width=True, help=f"Analyze: {q}"):
+                    selected_prompt = q
+                    st.session_state.last_selected_prompt = q
+        
+        with st.expander("🌬️ Meteorology & Environmental Factors", expanded=False):
+            for i, q in enumerate([q for q in questions if any(word in q.lower() for word in ['wind', 'temperature', 'humidity', 'rainfall', 'meteorological', 'monsoon', 'barometric'])]):
+                if st.button(q, key=f"met_q_{i}", use_container_width=True, help=f"Analyze: {q}"):
+                    selected_prompt = q
+                    st.session_state.last_selected_prompt = q
+        
+        with st.expander("👥 Population & Demographics", expanded=False):
+            for i, q in enumerate([q for q in questions if any(word in q.lower() for word in ['population', 'capita', 'density', 'exposure'])]):
+                if st.button(q, key=f"pop_q_{i}", use_container_width=True, help=f"Analyze: {q}"):
+                    selected_prompt = q
+                    st.session_state.last_selected_prompt = q
+        
+        with st.expander("🏭 Multi-Pollutant Analysis", expanded=False):
+            for i, q in enumerate([q for q in questions if any(word in q.lower() for word in ['ozone', 'no2', 'correlation', 'multi-pollutant', 'interaction'])]):
+                if st.button(q, key=f"multi_q_{i}", use_container_width=True, help=f"Analyze: {q}"):
+                    selected_prompt = q
+                    st.session_state.last_selected_prompt = q
+        
+        with st.expander("📈 Other Analysis Questions", expanded=False):
+            remaining_questions = [q for q in questions if not any(any(word in q.lower() for word in category) for category in [
+                ['ncap', 'funding', 'investment', 'rupee'],
+                ['wind', 'temperature', 'humidity', 'rainfall', 'meteorological', 'monsoon', 'barometric'],
+                ['population', 'capita', 'density', 'exposure'],
+                ['ozone', 'no2', 'correlation', 'multi-pollutant', 'interaction']
+            ])]
+            for i, q in enumerate(remaining_questions):
+                if st.button(q, key=f"other_q_{i}", use_container_width=True, help=f"Analyze: {q}"):
+                    selected_prompt = q
+                    st.session_state.last_selected_prompt = q
     
     st.markdown("---")
     
@@ -815,6 +852,32 @@ for response_id, response in enumerate(st.session_state.responses):
                 st.session_state.responses[response_id]["feedback"] = feedback
                 st.success("Thanks for your feedback!")
                 st.rerun()
+        
+        # Add retry button for each assistant response
+        col1, col2, col3 = st.columns([1, 1, 2])
+        with col1:
+            if st.button("🔄 Retry", key=f"retry_{response_id}", help="Regenerate this response with current model"):
+                # Get the last user prompt that led to this response
+                user_prompt = ""
+                if response_id > 0:
+                    user_prompt = st.session_state.responses[response_id-1].get("content", "")
+                
+                if user_prompt:
+                    # Remove this response and the user message before it, then re-add the user message
+                    if response_id > 0:
+                        # Store the user prompt
+                        retry_prompt = st.session_state.responses[response_id-1].get("content", "")
+                        # Remove both user message and assistant response
+                        del st.session_state.responses[response_id]
+                        del st.session_state.responses[response_id-1]
+                        # Re-add user message and trigger new response
+                        st.session_state.follow_up_prompt = retry_prompt
+                        st.rerun()
+        
+        with col2:
+            if st.button("💬 Follow-up", key=f"followup_{response_id}", help="Ask a follow-up question"):
+                st.session_state.follow_up_mode = True
+                st.rerun()
 
 # Chat input with better guidance
 prompt = st.chat_input("💬 Ask about air quality trends, compare cities, or request visualizations...", key="main_chat")

questions.txt

@@ -1,31 +1,30 @@
-What percentage of India’s population lives in states where the average PM2.5 concentration exceeds 60 µg/m³?
-What is the correlation between wind speed and PM2.5 levels in Ahmedabad during July 2023?
-Which NCAP-funded city showed the greatest improvement in PM2.5 levels from 2018 to 2019?
-Which month in 2023 had the highest average PM2.5 pollution levels?
-Which city has the worst air quality based on average PM2.5 levels?
-Calculate the overall annual average PM2.5 concentration across all cities in 2023
-Plot monthly average PM2.5 trends for 2023 across all cities
-Rank all cities by average PM2.5 levels from highest to lowest pollution
-Compare average PM2.5 levels: winter months (Dec-Feb) vs summer months (Apr-Jun)
-Show seasonal PM2.5 patterns: which season has highest pollution levels?
-Which cities have annual average PM2.5 exceeding WHO guideline of 15 μg/m³?
-Identify the top 10 most polluted cities based on PM2.5 in 2023
-Compare PM10 vs PM2.5 correlation strength across different cities
-Which station in Ahmedabad shows highest PM2.5 variability in winter 2023?
-Calculate PM2.5 improvement rate from 2022 to 2023 for Mumbai vs Delhi
-Identify Gujarat cities with PM2.5 >50 μg/m³ for 6+ consecutive months in 2023
-Compare PM2.5 vs PM10 correlation: winter vs summer across top 5 polluted cities
-Compare average pollution levels: weekdays (Mon-Fri) vs weekends (Sat-Sun)
-Plot year-over-year PM2.5 trend from 2022 to 2023 for major cities
-Show PM2.5 distribution histogram across all cities and time periods
-Create scatter plot showing PM2.5 vs PM10 correlation with trend line
-Which cities need immediate emergency intervention with PM2.5 >100 μg/m³?
-Identify cities showing consistent pollution improvement for policy replication
-Which months require targeted interventions based on highest pollution spikes?
-Compare pollution levels between industrial vs non-industrial cities for policy focus
-Rank states by average air quality to prioritize national-level resource allocation
-Which cities have PM2.5 levels in 'hazardous' category (>250 μg/m³) requiring urgent action?
-Calculate population-weighted pollution exposure to identify areas affecting most people
-Show cities with worsening pollution trends that need immediate policy intervention
-Compare pollution reduction success: which cities improved most and how?
-Identify seasonal emergency periods when public health advisories should be issued
+How much NCAP funding did Delhi receive vs its PM2.5 improvement from 2018-2023?
+Which NCAP cities achieved the best PM2.5 reduction per rupee invested?
+Does wind speed above 5 m/s significantly reduce PM2.5 levels in Delhi?
+Show correlation between rainfall and PM2.5 reduction in Mumbai during monsoon
+Which cities with high population have dangerously high PM2.5 exposure levels?
+Compare winter PM2.5 levels: high-funded vs low-funded NCAP cities
+Does temperature increase correlate with ozone levels in Chennai during summer?
+Plot wind direction vs PM2.5 concentration rose diagram for Delhi in November
+Which meteorological factor most influences PM2.5 reduction in Ahmedabad?
+Rank NCAP cities by pollution improvement efficiency per capita funding
+Show monthly PM2.5 trends for top 5 most populated Indian cities
+Does humidity above 70% help reduce PM10 levels in coastal cities?
+Compare NO2 vs PM2.5 correlation in traffic-heavy vs residential areas
+Which NCAP-funded cities still exceed WHO PM2.5 guidelines despite investment?
+Plot seasonal wind patterns vs pollution levels for North Indian cities
+Show population-weighted pollution exposure map across Indian states
+Does solar radiation intensity affect ground-level ozone formation patterns?
+Compare NCAP investment effectiveness: Tier-1 vs Tier-2 cities
+Which high-population cities need emergency NCAP funding based on current PM2.5?
+Show correlation between barometric pressure and pollution accumulation
+Does monsoon season consistently reduce all pollutant levels nationwide?
+Compare multi-pollutant exposure: children vs adults in high-density cities
+Which cities show pollution improvement correlated with NCAP timeline?
+Plot wind speed threshold for effective pollution dispersion by region
+Show relationship between city population density and average PM2.5 exposure
+Compare Ozone-PM2.5-NO2 interaction patterns in Delhi vs Mumbai
+Does vector wind speed predict pollution episodes better than average wind speed?
+Which NCAP phases (1,2,3) showed maximum pollution reduction per investment?
+Show real-time impact: do meteorological alerts help predict pollution spikes?
+Create pollution risk index combining PM2.5, population, and meteorology data

system_prompt.txt

@@ -3,6 +3,7 @@ I have a pandas dataframe data of PM2.5 and PM10.
 * Frequency of data is daily.
 * `pollution` generally means `PM2.5`.
 * You already have df, so don't read the csv file 
+* Available libraries: pandas, matplotlib, numpy, seaborn, plotly, geopandas, statsmodels, scikit-learn
 * Don't print anything, but save result in a variable `answer` and make it global.
 * Unless explicitly mentioned, don't consider the result as a plot.
 * PM2.5 guidelines: India: 60, WHO: 15.