app.py

import streamlit as st
import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt
import plotly.express as px
from plotly.subplots import make_subplots
import plotly.graph_objects as go
st.set_page_config(
    page_title="Air Quality Analysis",
    page_icon="🌍",
    layout="wide"  # Set layout to wide
)


# Add custom CSS for background color
page_bg_color = """
<style>
/* Change the background color of the entire page */
[data-testid="stAppViewContainer"] {
    background-color: #5C7285; /* Light blue-gray color */
}

/* Optional: Adjust sidebar background for consistency */
[data-testid="stSidebar"] {
    background-color: #727D73; /* Light gray-blue for the sidebar */
}
</style>
"""

st.markdown(page_bg_color, unsafe_allow_html=True)


# Load your AQI data
@st.cache_data
def load_data():
    # Replace with your data loading logic
    return pd.read_csv(r"random.csv")

# Load data
data = load_data()

# Dashboard Title
st.markdown(
    """
    
# **Comprehensive Analysis of Air Quality and Pollutant Dynamics Across Indian Cities**
### **Completed On:** February 2025  
### **Author:** Kaustubh Yewale  

## **Introduction**  
Air pollution is one of the most critical challenges in urban environments, affecting public health, ecosystems, and overall quality of life.  
This project provides an in-depth analysis of air quality across Indian cities, focusing on pollutants like PM2.5, PM10, CO, SO₂, NO₂, and O₃.  
Through **advanced visualizations and data-driven insights**, the project uncovers pollution patterns, identifies high-risk areas, and explores the relationship between **weather conditions and pollution levels**.  
The ultimate goal is to provide **actionable recommendations** for mitigating pollution and supporting **policymakers in making informed decisions**.

    """,
    unsafe_allow_html=True
)
st.image(r"Picture1.png")
st.markdown(
    """
    
## **Problem Statement**  
Urban air pollution has emerged as a severe environmental and public health issue. Understanding air quality trends and their correlation with weather conditions is crucial for:  
- Identifying pollution hotspots  
- Evaluating health risks associated with high pollution levels  
- Formulating strategic mitigation policies  

This project **analyzes air quality trends** across Indian cities, using **data science techniques** to extract meaningful insights from pollutant and weather data.

   
    """,
    unsafe_allow_html=True
)
st.write('''
## **Objectives**  
1. **Retrieve and preprocess air quality data** from sources like `aqi.in` and **merge with weather data**.
2. **Perform Exploratory Data Analysis (EDA)** to identify pollution trends across different cities and regions.
3. **Analyze the correlation** between AQI and weather conditions (e.g., temperature, humidity, wind speed).
4. **Develop machine learning models** to predict AQI levels based on weather conditions and pollutants.
5. **Provide visual insights** using maps, correlation heatmaps, and distribution plots.
6. **Propose recommendations** for improving air quality and highlight policy implications.
''')
st.markdown(
    """
    ## **📊 Data Collection Approach**
    
    ##### 1. 🌍 Latitude and Longitude Retrieval:
    Obtained latitude and longitude of various areas using **ChatGPT** for location-based analysis.

    ##### 2. 🌫️ Air Quality Data Scraping:
    Scraped data for key pollutants (**PM2.5, PM10, CO, SO₂, NO₂, O₃**) and AQI from **aqi.in** using Python libraries.

    ##### 3. ☁️ Weather Data Integration:
    Retrieved additional weather features (**humidity, pressure, wind speed**) using the **Weatherstack API**.

    ##### 4. 🔗 Data Consolidation:
    Merged pollutant data with weather data for each location to create a comprehensive dataset.
    """,
    unsafe_allow_html=True
)


st.markdown('<h1 style="color: #FFD700;">AQI Dashboard</h1>', unsafe_allow_html=True)
st.markdown("Analyze Air Quality Index data and trends.")

# Sidebar Filters
st.subheader("Filter Options")
cities = st.multiselect("Select Cities", options=data["areas"].unique())
pollutants = st.multiselect("Select Pollutants", options=['pm2.5', 'pm10', 'co', 'so2', 'no2', 'o3'])


# Filter Data
filtered_data = data.copy()
if cities:
    filtered_data = filtered_data[filtered_data["areas"].isin(cities)]
if pollutants:
    filtered_data = filtered_data[pollutants + ["areas", "AQI"]]



# Display Filtered Data
st.subheader("Filtered Data")
st.dataframe(filtered_data)

# Plot AQI Trends


# Additional Features
st.subheader("Statistics")
if not filtered_data.empty:
    st.write(filtered_data.describe())
else:
    st.write("No data available for statistics.")




st.markdown('<h1 style="color: #FFD700;">Outlier Detection for Various Pollutants</h1>', unsafe_allow_html=True)

# Create a subplot grid
fig = make_subplots(
    rows=3, cols=2, 
    subplot_titles=[
        'Outlier Detection: PM2.5 and PM10',
        'Outlier Detection: SO2, NO2, O3',
        'Outlier Detection: CO and Total Pollution',
        'Outlier Detection: Humidity and Temperature',
        'Outlier Detection: CO',
        'Outlier Detection: Pressure'
    ]
)

# Add box plots for PM2.5 and PM10
fig.add_trace(
    go.Box(y=data['pm2.5'], name='PM2.5'),
    row=1, col=1
)
fig.add_trace(
    go.Box(y=data['pm10'], name='PM10'),
    row=1, col=1
)

# Add box plots for SO2, NO2, and O3
fig.add_trace(
    go.Box(y=data['so2'], name='SO2'),
    row=1, col=2
)
fig.add_trace(
    go.Box(y=data['no2'], name='NO2'),
    row=1, col=2
)
fig.add_trace(
    go.Box(y=data['o3'], name='O3'),
    row=1, col=2
)

# Add box plots for CO and Total Pollution
fig.add_trace(
    go.Box(y=data['co'], name='CO'),
    row=2, col=1
)
fig.add_trace(
    go.Box(y=data['Total_Pollution'], name='Total Pollution'),
    row=2, col=1
)

# Add box plots for Humidity and Temperature
fig.add_trace(
    go.Box(y=data['humdity'], name='Humidity'),
    row=2, col=2
)
fig.add_trace(
    go.Box(y=data['temperature'], name='Temperature'),
    row=2, col=2
)

# Add box plots for Pressure and CO
fig.add_trace(
    go.Box(y=data['pressure'], name='Pressure'),
    row=3, col=1
)
fig.add_trace(
    go.Box(y=data['co'], name='CO'),
    row=3, col=2
)

# Update layout
fig.update_layout(
    height=1000,
    title_text="Outlier Detection for Various Pollutants",
    showlegend=False
)

# Display the boxplot in Streamlit
st.plotly_chart(fig, use_container_width=True)
st.write('''### Insights and Action-Oriented Insights:
1. **CO exhibits the highest variability** with extreme outliers, reaching levels over **2000**, making it the most critical pollutant to address.
2. **PM10 shows noticeable outliers**, suggesting localized high emissions, likely from construction or dust-related activities.
3. **Other pollutants (PM2.5, SO₂, NO₂, O₃)** have relatively lower levels and variability but should still be monitored to avoid spikes.
4. Focus on reducing **CO emissions** through stricter vehicle and industrial emission regulations.
5. Address **PM10 outliers** by implementing dust control measures at construction sites and promoting green buffers in urban areas.
''')






st.markdown('<h1 style="color: #FFD700;">Air Quality Distribution by Category</h1>', unsafe_allow_html=True)
category_counts = data['Air quality'].value_counts().reset_index()
category_counts.columns = ['Air quality', 'Count']

# Create the bar chart using Plotly
fig1 = px.bar(category_counts, x='Air quality', y='Count',
              color='Air quality', title='Air Quality Distribution by Category',
              barmode='stack', text='Count')
fig1.update_traces(textposition='outside')

# Display the Plotly chart in Streamlit
st.plotly_chart(fig1, use_container_width=True)
st.write('''### Insights:
1. Most areas fall under **"Moderately Polluted"** and **"Satisfactory"** categories, indicating manageable but concerning pollution levels.
2. Very few areas (40) have **"Good"** air quality, highlighting the need for improvement.
3. **"Poor" and "Very Poor"** areas require urgent intervention to address harmful pollution.

---

### Action-Oriented Insights:
1. Focus on **"Poor" and "Very Poor" regions** with stricter emission controls and localized mitigation strategies.
2. Promote **urban greening** and **renewable energy adoption** to improve air quality across all regions.
''')

st.markdown('<h1 style="color: #FFD700;">Pollution Category Distribution</h1>', unsafe_allow_html=True)

# Calculate Pollution Category Counts
pollution_counts = data['Pollution_Category'].value_counts().reset_index()
pollution_counts.columns = ['Pollution Category', 'Count']

# Define pull values (explode only 'Considerable')
pollution_counts['pull'] = pollution_counts['Pollution Category'].apply(
    lambda x: 0.2 if x == 'Considerable' else 0
)

# Create the pie chart with explosion
fig = px.pie(
    pollution_counts, 
    names='Pollution Category', 
    values='Count',
    title='Distribution of Pollution Categories',
)
fig.update_traces(
    textinfo='percent+label', 
    pull=pollution_counts['pull']
)
fig.update_layout(
    title_font_size=24,  # Increase title font size
    width=800,          # Set the chart width
    height=800,         # Set the chart height
    showlegend=True,    # Ensure the legend is visible
)

# Display the pie chart
st.plotly_chart(fig, use_container_width=True)
st.write('''### Combined Insights and Action-Oriented Insights:
1. Over **50% of areas** fall into the **"Hazardous"** category, posing severe public health risks.  
2. **Unsafe (25.7%)** and **Extremely Hazardous (21.4%)** areas demand immediate interventions to reduce pollutant levels.  
3. Implement stricter emission controls and promote **renewable energy adoption** to tackle high pollution areas.  
4. Deploy **real-time air monitoring systems** to track pollution trends and take proactive measures.  
5. Raise public awareness to encourage sustainable practices like reducing vehicular emissions and waste burning.  
''')



# Scatter Plot for AQI vs Temperature
st.markdown('<h1 style="color: #FFD700;">Pollution (AQI) vs Temperature</h1>', unsafe_allow_html=True)
fig3 = px.scatter(
    data, 
    x='temperature', 
    y='AQI', 
    color='Air quality',
    size='pm2.5', 
    hover_name='areas', 
    title='Pollution (AQI) vs Temperature',
    labels={'temperature': 'Temperature (°C)', 'AQI': 'Air Quality Index'}
)

# Display the scatter plot in Streamlit
st.plotly_chart(fig3, use_container_width=True)
st.write('''### Insights and Action-Oriented Insights:
1. AQI levels are highest between **10°C and 20°C**, with "Moderately Polluted" and "Satisfactory" air quality dominating most temperature ranges.  
2. "Very Poor" air quality is primarily observed in regions with temperatures above **15°C**, indicating a potential correlation with warmer conditions.  
3. Focus on reducing **PM2.5 emissions**, as larger bubbles highlight its significant contribution to poor AQI levels.  
4. Investigate pollutant sources in regions with **10°C–20°C** temperatures and enforce stricter emission controls in warmer areas.  
''')


st.markdown('<h1 style="color: #FFD700;">Total Pollution vs Temperature</h1>', unsafe_allow_html=True)

# Create the scatter plot using Plotly
fig3 = px.scatter(
    data, x='temperature', y='Total_Pollution', 
    color='Air quality', size='pm2.5', 
    hover_name='areas', 
    title='Total Pollution vs Temperature',
    labels={'temperature': 'Temperature (°C)', 'Total_Pollution': 'Total Pollution'}
)

# Display the scatter plot in Streamlit
st.plotly_chart(fig3, use_container_width=True)

st.write('''### **Insights and Action-Oriented Recommendations**
1. **Pollution Peaks at Moderate Temperatures**:
   - Total pollution is highest around 15°C, with "Very Poor" air quality dominating. Target pollution control efforts during these conditions by limiting industrial and vehicular emissions.

2. **Decline in Pollution at Higher Temperatures**:
   - Total pollution decreases as temperatures exceed 25°C. Encourage maintaining such trends through renewable energy initiatives and green urban planning.

3. **Satisfactory Air Quality Across Wider Ranges**:
   - Areas with "Satisfactory" air quality are evenly distributed across temperatures. Expand awareness campaigns to sustain and improve such conditions.

4. **Outliers with Extreme Pollution**:
   - Outlier points with Total Pollution > 2000 demand immediate investigation and action, particularly in "Very Poor" air quality zones.

5. **Seasonal Pollution Management**:
   - Develop temperature-based pollution mitigation strategies to address seasonal patterns, focusing on cooler months when pollution spikes.
''')


st.markdown('<h1 style="color: #FFD700;">Weather Conditions vs AQI</h1>', unsafe_allow_html=True)

# Aggregate AQI by weather condition
weather_aqi = data.groupby('Weather_Condition')['AQI'].mean().reset_index()

# Create the bar chart using Plotly
fig5 = px.bar(
    weather_aqi, x='Weather_Condition', y='AQI', 
    color='Weather_Condition', text='AQI', 
    title='Weather Conditions vs AQI',
    color_discrete_sequence=px.colors.qualitative.Pastel
)

fig5.update_traces(textposition='outside')

# Display the chart in Streamlit
st.plotly_chart(fig5, use_container_width=True)
st.write('''### Insights and Action-Oriented Insights:
1. **Cool & Dry** and **Normal** weather conditions exhibit the highest AQI levels (**118.36** and **117.23** respectively), indicating significant pollution during these conditions.
2. **Hot & Humid** conditions show relatively lower AQI levels (**80.75**), possibly due to improved pollutant dispersion in warmer, humid climates.
3. Focus on **Cool & Dry** regions for pollution control measures, such as stricter emission regulations and increased monitoring during these conditions.
4. Implement **urban greenery initiatives** and promote **clean energy** solutions to mitigate AQI spikes in areas with "Cool & Dry" and "Normal" conditions.
5. Investigate local sources of pollution in these weather conditions to design targeted mitigation strategies.
''')
st.markdown('<h1 style="color: #FFD700;">Relationship Between Wind Speed and PM2.5 Levels</h1>', unsafe_allow_html=True)

# Create scatter plot using Plotly
fig_wind_pm25 = px.scatter(
    data, 
    x='wind_speed', 
    y='pm2.5', 
    color='Air quality',
    size='pm2.5',
    title='Relationship Between Wind Speed and PM2.5 Levels',
    labels={'wind_speed': 'Wind Speed (m/s)', 'pm2.5': 'PM2.5 Levels'},
    hover_name='areas'
)

fig_wind_pm25.update_layout(
    xaxis_title='Wind Speed (m/s)',
    yaxis_title='PM2.5 Levels',
    legend_title='Air Quality'
)

# Display the scatter plot in Streamlit
st.plotly_chart(fig_wind_pm25, use_container_width=True)
st.write('''### **Insights**:
🌟 Areas with **low wind speed (<10 m/s)** tend to have **higher PM2.5 levels**, indicating insufficient dispersion of particulate matter.  
🌟 **Moderately polluted and very poor air quality categories** dominate in regions with low wind speed, emphasizing the impact of stagnant air conditions.  

### **Action-Oriented Insights**:
✅ Promote **urban greening and wind corridors** in high-density areas to naturally improve air circulation.  
✅ **Install pollution control technologies** near industrial and urban hotspots to mitigate the accumulation of PM2.5.  ''')

st.markdown('<h1 style="color: #FFD700;">Relationship Between Humidity and AQI</h1>', unsafe_allow_html=True)

# Create the plot
fig, ax = plt.subplots(figsize=(10, 6))
sns.regplot(
    x=data['humdity'], y=data['AQI'], 
    scatter_kws={'alpha': 0.6}, line_kws={'color': 'red'}, ax=ax
)

ax.set_title('Relationship Between Humidity and AQI')
ax.set_xlabel('Humidity (%)')
ax.set_ylabel('Air Quality Index (AQI)')

# Display the plot in Streamlit
st.pyplot(fig)
st.write('''### Insights:
1. The scatterplot shows a **weak positive correlation** between humidity (%) and AQI, with higher humidity levels slightly associated with increased AQI values.  
2. The **wide spread** of AQI across all humidity levels suggests that humidity alone is not a strong determinant of air quality.

### Action-Oriented Insights:
1. **Focus on humid regions**: Implement targeted air quality monitoring and mitigation strategies in areas with high humidity and poor air quality.  
2. **Address other factors**: Investigate additional variables like wind speed, industrial emissions, and temperature to better understand AQI variability.
''')


st.markdown('<h1 style="color: #FFD700;">Top 10 Most Polluted Cities</h1>', unsafe_allow_html=True)

# Get the top 10 most polluted cities
top_10_most_polluted = data.nlargest(10, 'AQI')

# Create bar chart using Plotly
fig6 = px.bar(
    top_10_most_polluted.sort_values('AQI', ascending=True),
    y='areas',
    x='AQI',
    title='Top 10 Most Polluted Cities',
    color='AQI',
    hover_data=['pm2.5', 'pm10', 'co', 'so2', 'no2', 'o3', 'Total_Pollution'],
    color_continuous_scale=px.colors.sequential.Oryel
)

# Display the chart in Streamlit
st.plotly_chart(fig6, use_container_width=True)
st.write('''### Insights from the Visualization:

**Most Polluted Cities**:
   - Cities like **Serampore**, **Chinsurah**, and **Kalyani** rank highest in AQI, indicating critically poor air quality.
   - These cities may require immediate interventions due to the significant health risks posed by air pollution.

---

### Recommendations for Action:

**For Most Polluted Cities**:
   - Implement stricter industrial emission controls and traffic regulations.
   - Promote renewable energy usage and reduce dependency on fossil fuels.


''')

st.markdown('<h1 style="color: #FFD700;">Top 10 Least Polluted Cities</h1>', unsafe_allow_html=True)

# Get the top 10 least polluted cities
top_10_least_polluted = data.nsmallest(10, 'AQI')

# Create bar chart using Plotly
fig7 = px.bar(
    top_10_least_polluted.sort_values('AQI', ascending=True),
    y='areas',
    x='AQI',
    title='Top 10 Least Polluted Cities',
    color='AQI',
    hover_data=['pm2.5', 'co', 'pm10', 'Total_Pollution'],
    color_continuous_scale=px.colors.sequential.Blugrn
)

# Display the chart in Streamlit
st.plotly_chart(fig7, use_container_width=True)

st.write('''### Insights from the Visualization:
**Least Polluted Cities**:
   - Cities such as **Pathanamthitta**, **Adoor**, and **Chengannur** showcase very low AQI values, indicating excellent air quality.
   - These cities could serve as benchmarks for pollution control measures.
---
### Recommendations for Action:
**For Least Polluted Cities**:
   - Ensure continuous monitoring to maintain the air quality.
   - Encourage green practices such as afforestation and waste management.''')

st.markdown('<h1 style="color: #FFD700;">Correlation Heatmap of AQI and Pollutants</h1>', unsafe_allow_html=True)

# Select relevant data
correlation_data = data[['pm2.5', 'pm10', 'co', 'so2', 'no2', 'o3', 'AQI']]

# Compute correlation matrix
corr_matrix = correlation_data.corr()

# Create heatmap using Plotly
fig = go.Figure(
    data=go.Heatmap(
        z=corr_matrix.values,  # Correlation values
        x=corr_matrix.columns,  # Column names
        y=corr_matrix.columns,  # Row names
        colorscale='Viridis',  # Color scheme
        zmin=-1,  # Minimum correlation value
        zmax=1,   # Maximum correlation value
        texttemplate="%{z:.2f}",  # Format correlation values
        textfont={"size": 12}
    )
)

# Add title and layout settings
fig.update_layout(
    title="Correlation Heatmap of AQI and Pollutants",
    xaxis=dict(title="Variables", tickangle=45),
    yaxis=dict(title="Variables"),
    autosize=True,
    template="plotly_white"
)

# Display the heatmap in Streamlit
st.plotly_chart(fig, use_container_width=True)
st.write('''### Insights and Recommendations:

1. **PM2.5 (0.92)** and **PM10 (0.88)** have the strongest correlation with AQI, making them the primary contributors to air quality degradation.
2. **CO (0.78)** also plays a significant role in worsening AQI, often coexisting with particulate matter from combustion sources.
3. **O₃ (-0.11)** shows a weak negative correlation with AQI, suggesting minimal or inverse influence under certain conditions.
4. Focus on reducing **PM2.5**, **PM10**, and **CO** through stricter emission controls and cleaner fuel adoption.
5. Mitigate combined emissions of particulate matter and CO by targeting industrial and vehicular pollution sources.### Insights and Recommendations:

1. **PM2.5 (0.92)** and **PM10 (0.88)** have the strongest correlation with AQI, making them the primary contributors to air quality degradation.
2. **CO (0.78)** also plays a significant role in worsening AQI, often coexisting with particulate matter from combustion sources.
3. **O₃ (-0.11)** shows a weak negative correlation with AQI, suggesting minimal or inverse influence under certain conditions.
4. Focus on reducing **PM2.5**, **PM10**, and **CO** through stricter emission controls and cleaner fuel adoption.
5. Mitigate combined emissions of particulate matter and CO by targeting industrial and vehicular pollution sources.''')



st.markdown('<h1 style="color: #FFD700;">Pollutant Levels Across Pollution Categories</h1>', unsafe_allow_html=True)

# Create the bar chart using Plotly
fig11 = px.bar(
    data,
    x='Pollution_Category',
    y=['pm2.5', 'pm10', 'so2', 'no2', 'o3', 'co'],
    title='Pollutant Levels Across Pollution Categories',
    labels={'value': 'Pollutant Levels (µg/m³)', 'variable': 'Pollutants'},
    barmode='group',
    hover_name='state',
    color_discrete_sequence=px.colors.qualitative.Vivid_r  # Vibrant color palette
)

# Enhance hover details
fig11.update_traces(hovertemplate='%{x}<br>%{y:.2f} µg/m³<br>Pollutant: %{legendgroup}')

# Customize layout
fig11.update_layout(
    title_font=dict(size=20, color='Grey', family='Arial Black'),
    xaxis_title_font=dict(size=16, color='black'),
    yaxis_title_font=dict(size=16, color='black'),
    legend_title=dict(font=dict(size=14)),
    plot_bgcolor='rgba(240, 240, 240, 0.9)'  # Light background
)

# Display the bar chart in Streamlit
st.plotly_chart(fig11, use_container_width=True)
st.write('''### **Insights**
1. **Carbon Monoxide Dominance**:
   - CO is the most significant pollutant across all pollution categories, with exceptionally high levels in the "Extremely Hazardous" and "Hazardous" categories, indicating the substantial contribution of vehicle emissions and industrial processes.

2. **PM2.5 and PM10 Contribution**:
   - Particulate matter (PM2.5 and PM10) shows a strong presence in "Hazardous" and "Unsafe" categories, reflecting pollution sources like construction, road dust, and combustion activities.

3. **Minor Contribution from Ozone (O₃) and Sulfur Dioxide (SO₂)**:
   - O₃ and SO₂ have relatively lower levels across all categories, suggesting lesser emissions from natural sources or effective control measures for these pollutants in some regions.

4. **Disproportionate Distribution**:
   - The pollutant levels significantly vary across pollution categories, with "Extremely Hazardous" and "Hazardous" contributing disproportionately higher levels compared to other categories like "Unsafe" or "Considerable."

---

### **Action-Oriented Recommendations**
1. **Target CO Emissions**:
   - Implement stricter vehicle emission standards and promote the transition to electric vehicles to reduce CO levels in the most polluted regions.

2. **Control PM2.5 and PM10 Sources**:
   - Enforce regulations to curb construction dust, road dust, and combustion activities, especially in "Hazardous" and "Unsafe" areas.

3. **Monitor and Mitigate Pollutants**:
   - Deploy real-time monitoring systems in "Extremely Hazardous" zones to identify and address localized sources of pollution effectively.

4. **Urban Greening Initiatives**:
   - Increase vegetation and green buffers in urban areas to naturally absorb pollutants like PM2.5 and PM10 and improve air quality.

5. **Public Awareness Campaigns**:
   - Educate communities on reducing activities that contribute to pollutants, such as waste burning, and encourage the adoption of public transport.
''')

st.markdown('<h1 style="color: #FFD700;">Interactive Map of AQI</h1>', unsafe_allow_html=True)

# Create the interactive scatter geo map
fig9 = px.scatter_geo(
    data,
    lat='lat',
    lon='long',
    color='AQI',
    size='AQI',
    title='Interactive Map of AQI',
    hover_name='areas',
    hover_data=['wind_speed'],
    color_continuous_scale='RdYlGn_r'  # Red-Yellow-Green scale (reversed)
)

# Update layout for better visualization
fig9.update_layout(
    geo=dict(
        showland=True,
        landcolor="rgb(217, 217, 217)",
        showcoastlines=True,
        coastlinecolor="rgb(0, 0, 0)",
    ),
    margin=dict(l=10, r=10, t=50, b=10)
)

# Display the map in Streamlit
st.plotly_chart(fig9, use_container_width=True)
st.write(':red[Its an interactive map you can zoom it and hover it to get more info.]')
st.write('''### Insights and Action-Oriented Insights:
1. **Northern and eastern regions** exhibit the highest AQI levels, with hotspots visible, likely due to industrial activities and high population density.
2. **Central and southern regions** show relatively moderate AQI levels, indicating better air quality and manageable pollution levels.
3. Urban clusters with severe AQI levels require **localized pollution control measures**, including traffic management and stricter industrial regulations.
4. Deploy additional **air quality monitoring stations** in less covered areas to ensure comprehensive data collection for proactive interventions.
5. Focus on **reducing vehicular emissions** and promoting **green energy solutions** in high-AQI regions to improve air quality.
''')
st.markdown('<h1 style="color: #FFD700;">Conclusion</h1>', unsafe_allow_html=True)

st.markdown('''
The AQI EDA project has provided a **comprehensive analysis of air quality** across different regions.  
Key insights include:

1. **Air Quality Trends**:  
   - The project identified areas with critical pollution levels and categorized them into distinct AQI categories (*Good, Satisfactory, Moderate, Poor, Very Poor, Severe*).  
   - This helps pinpoint regions requiring urgent intervention.

2. **Geographic Visualization**:  
   - Interactive visualizations, such as scatter plots on satellite maps, offered an intuitive way to understand the **spatial distribution of pollution levels**, highlighting hotspots and relatively cleaner areas.

3. **Correlations and Patterns**:  
   - Analysis of parameters like **PM2.5, PM10, CO, SO₂, NO₂, and O₃** provided valuable insights into their contributions to overall pollution.  
   - Seasonal variations and weather conditions like humidity and wind speed were found to influence AQI.

4. **Health Impacts**:  
   - The classification of AQI into categories (*e.g., Hazardous, Unsafe*) serves as a tool for raising **public awareness** about the health risks associated with air pollution.
''')