import streamlit as st
import pandas as pd
import numpy as np
import joblib
import json
import folium
from streamlit_folium import st_folium
import plotly.graph_objects as go
import plotly.express as px
from datetime import datetime, timedelta
import requests

# Page configuration
st.set_page_config(
    page_title="Charsadda Flood Prediction System",
    page_icon="🌊",
    layout="wide",
    initial_sidebar_state="expanded"
)

# Custom CSS
st.markdown("""<style>
.main-header {
    font-size: 2.5rem;
    font-weight: bold;
    color: #1f77b4;
    text-align: center;
    margin-bottom: 2rem;
}
.warning-box {
    background-color: #fff3cd;
    border-left: 5px solid #ffc107;
    padding: 1rem;
    margin: 1rem 0;
}
.danger-box {
    background-color: #f8d7da;
    border-left: 5px solid #dc3545;
    padding: 1rem;
    margin: 1rem 0;
}
.safe-box {
    background-color: #d4edda;
    border-left: 5px solid #28a745;
    padding: 1rem;
    margin: 1rem 0;
}
</style>""", unsafe_allow_html=True)

# Load model and scaler
@st.cache_resource
def load_models():
    try:
        rf_model = joblib.load('random_forest_model.pkl')
        scaler = joblib.load('scaler.pkl')
        with open('model_metadata.json', 'r') as f:
            metadata = json.load(f)
        return rf_model, scaler, metadata
    except Exception as e:
        st.error(f"Error loading models: {e}")
        return None, None, None

rf_model, scaler, metadata = load_models()

# Header
st.markdown('<div class="main-header">🌊 AI-Powered Flood Prediction System</div>', unsafe_allow_html=True)
st.markdown("### Charsadda, Khyber Pakhtunkhwa, Pakistan")

# Sidebar
with st.sidebar:
    st.image("https://upload.wikimedia.org/wikipedia/commons/thumb/3/32/Flag_of_Pakistan.svg/320px-Flag_of_Pakistan.svg.png", width=150)
    st.title("Navigation")
    page = st.radio("Select Page", ["🏠 Home", "🔮 Flood Prediction", "🗺️ Interactive Map", "📊 Model Info", "ℹ️ About"])
    st.markdown("---")
    st.markdown("### Study Area")
    st.markdown("**Location:** Charsadda District")
    st.markdown("**Province:** Khyber Pakhtunkhwa")
    st.markdown("**Coordinates:** 34.15°N, 71.72°E")

# HOME PAGE
if page == "🏠 Home":
    col1, col2 = st.columns(2)
    with col1:
        st.markdown("### 📋 Project Overview")
        st.write("""This AI-powered system predicts flood risks in Charsadda district using:
        - **Satellite imagery** (Sentinel-1 SAR & Sentinel-2 optical)
        - **Weather data** (temperature, humidity, rainfall)
        - **Topographic data** (elevation, slope)
        - **Machine Learning** (Random Forest, CNN, LSTM models)""")
        st.markdown("### 🎯 Key Features")
        st.markdown("""- ✅ Real-time flood risk assessment
        - ✅ Multi-source data integration
        - ✅ Interactive flood mapping
        - ✅ 1-3 day advance predictions
        - ✅ High-accuracy ML models""")
    with col2:
        st.markdown("### 📈 Historical Floods")
        st.info("""**2022 Flood (August 25-31)**\n- Affected thousands of people\n- Major infrastructure damage\n- Rivers Swat and Kabul overflowed\n\n**2025 Floods (June & August)**\n- Multiple flood events\n- ~35 km² area affected\n- ~41,000 people impacted""")
        st.markdown("### 🚨 Current Status")
        st.success("System operational and monitoring ongoing")

# FLOOD PREDICTION PAGE
elif page == "🔮 Flood Prediction":
    st.markdown("## Flood Risk Prediction")
    tab1, tab2 = st.tabs(["📤 Upload Data", "⌨️ Manual Input"])
    
    with tab1:
        st.markdown("### Upload Weather Forecast CSV")
        st.info("Upload a CSV file with columns: temperature, humidity, rainfall, pressure, wind_speed")
        uploaded_file = st.file_uploader("Choose a CSV file", type=['csv'])
        
        if uploaded_file is not None:
            df = pd.read_csv(uploaded_file)
            st.write("**Preview of uploaded data:**")
            st.dataframe(df.head())
            
            if st.button("Predict Flood Risk", key="predict_upload"):
                df['VV'] = -10
                df['B2'] = 800
                df['B3'] = 900
                df['B4'] = 600
                df['B8'] = 2000
                df['B11'] = 1500
                df['NDWI'] = 0.0
                df['MNDWI'] = 0.0
                df['NDVI'] = 0.4
                df['elevation'] = 300
                df['slope'] = 2
                
                feature_order = metadata['features']
                df = df[feature_order]
                X_scaled = scaler.transform(df)
                predictions = rf_model.predict(X_scaled)
                probabilities = rf_model.predict_proba(X_scaled)[:, 1]
                
                df['Flood_Risk'] = predictions
                df['Confidence'] = probabilities * 100
                
                st.markdown("### Prediction Results")
                flood_count = predictions.sum()
                col1, col2, col3 = st.columns(3)
                col1.metric("Total Predictions", len(predictions))
                col2.metric("Flood Risk Days", int(flood_count))
                col3.metric("Risk Percentage", f"{flood_count/len(predictions)*100:.1f}%")
                st.dataframe(df[['temperature', 'humidity', 'rainfall', 'Flood_Risk', 'Confidence']])
    
    with tab2:
        st.markdown("### Manual Weather Input")
        col1, col2 = st.columns(2)
        
        with col1:
            temperature = st.slider("Temperature (°C)", 15.0, 45.0, 30.0, 0.5)
            humidity = st.slider("Humidity (%)", 30.0, 100.0, 70.0, 1.0)
            rainfall = st.slider("Rainfall (mm)", 0.0, 300.0, 50.0, 5.0)
        with col2:
            pressure = st.slider("Pressure (hPa)", 950.0, 1050.0, 1010.0, 1.0)
            wind_speed = st.slider("Wind Speed (m/s)", 0.0, 30.0, 5.0, 0.5)
        
        if st.button("Predict Flood Risk", key="predict_manual"):
            input_data = pd.DataFrame([{
                'VV': -10, 'B2': 800, 'B3': 900, 'B4': 600, 'B8': 2000, 'B11': 1500,
                'NDWI': 0.0, 'MNDWI': 0.0, 'NDVI': 0.4, 'elevation': 300, 'slope': 2,
                'temperature': temperature, 'humidity': humidity, 'rainfall': rainfall,
                'pressure': pressure, 'wind_speed': wind_speed
            }])
            
            input_data = input_data[metadata['features']]
            X_scaled = scaler.transform(input_data)
            prediction = rf_model.predict(X_scaled)[0]
            probability = rf_model.predict_proba(X_scaled)[0, 1] * 100
            
            st.markdown("---")
            st.markdown("### Prediction Result")
            
            if prediction == 1:
                if probability > 75:
                    st.markdown(f'<div class="danger-box"><h3>🚨 HIGH FLOOD RISK</h3><p>Confidence: {probability:.1f}%</p><p><strong>Action Required:</strong> Immediate evacuation recommended.</p></div>', unsafe_allow_html=True)
                else:
                    st.markdown(f'<div class="warning-box"><h3>⚠️ MODERATE FLOOD RISK</h3><p>Confidence: {probability:.1f}%</p><p><strong>Action Required:</strong> Stay alert.</p></div>', unsafe_allow_html=True)
            else:
                st.markdown(f'<div class="safe-box"><h3>✅ LOW FLOOD RISK</h3><p>Confidence: {100-probability:.1f}%</p><p>Conditions are favorable.</p></div>', unsafe_allow_html=True)
            
            fig = go.Figure(go.Indicator(
                mode="gauge+number", value=probability, title={'text': "Flood Risk Level"},
                gauge={'axis': {'range': [0, 100]}, 'bar': {'color': "darkblue"},
                       'steps': [{'range': [0, 33], 'color': "lightgreen"},
                                 {'range': [33, 66], 'color': "yellow"},
                                 {'range': [66, 100], 'color': "red"}],
                       'threshold': {'line': {'color': "red", 'width': 4}, 'thickness': 0.75, 'value': 50}}
            ))
            fig.update_layout(height=300)
            st.plotly_chart(fig, use_container_width=True)

# INTERACTIVE MAP PAGE
elif page == "🗺️ Interactive Map":
    st.markdown("## Interactive Flood Risk Map")
    st.info("This map shows the Charsadda study area with historical flood zones.")
    
    m = folium.Map(location=[34.15, 71.72], zoom_start=11)
    boundary_coords = [[34.05, 71.55], [34.05, 71.90], [34.25, 71.90], [34.25, 71.55], [34.05, 71.55]]
    folium.Polygon(locations=boundary_coords, color='red', weight=3, fill=False, popup='Charsadda Study Area').add_to(m)
    folium.Marker([34.15, 71.72], popup='Charsadda Center', tooltip='Charsadda District', icon=folium.Icon(color='blue', icon='info-sign')).add_to(m)
    
    flood_zones = [([34.10, 71.65], '2022 Flood Zone', 'red'),
                   ([34.18, 71.75], '2025 June Flood', 'orange'),
                   ([34.12, 71.80], '2025 August Flood', 'darkred')]
    for coords, name, color in flood_zones:
        folium.Circle(location=coords, radius=2000, popup=name, color=color, fill=True, fillColor=color, fillOpacity=0.4).add_to(m)
    
    st_folium(m, width=1000, height=600)
    st.markdown("### Map Legend")
    col1, col2, col3 = st.columns(3)
    col1.markdown("🔴 **2022 Flood Zone**")
    col2.markdown("🟠 **2025 June Flood**")
    col3.markdown("🔴 **2025 August Flood**")

# MODEL INFO PAGE
elif page == "📊 Model Info":
    st.markdown("## Model Information")
    if metadata:
        col1, col2 = st.columns(2)
        with col1:
            st.markdown("### Model Details")
            st.info(f"""**Best Model:** {metadata['best_model']}\n**Training Date:** {metadata['training_date'][:10]}\n**Features:** {metadata['n_features']}\n**Training Samples:** {metadata['training_samples']:,}\n**Test Samples:** {metadata['test_samples']:,}""")
        with col2:
            st.markdown("### Performance Metrics")
            metrics_df = pd.DataFrame(metadata['performance_metrics'])
            st.dataframe(metrics_df, use_container_width=True)
        st.markdown("### Features Used")
        st.write(metadata['features'])
        fig = px.bar(metrics_df, x='Model', y=['Accuracy', 'Precision', 'Recall', 'F1-Score', 'ROC-AUC'], barmode='group', title='Model Performance Comparison')
        st.plotly_chart(fig, use_container_width=True)

# ABOUT PAGE
elif page == "ℹ️ About":
    st.markdown("## About This Project")
    st.markdown("""### AI-Powered Flood Prediction and Emergency Response System
    
#### Objective
To develop an AI-based flood prediction system that provides early warnings and supports efficient coordination of emergency services for Charsadda district, KPK, Pakistan.

#### Methodology
- **Multi-Source Data Fusion:** Combines Sentinel-1 SAR, Sentinel-2 optical imagery, SRTM elevation data, and weather forecasts
- **Machine Learning Models:** Random Forest, CNN, and LSTM models trained on historical flood events (2022 and 2025)
- **Real-time Monitoring:** Integration with OpenWeatherMap API for current conditions
- **Prediction Timeframe:** 1-3 days advance flood risk assessment

#### Data Sources
- **Satellite:** Sentinel-1, Sentinel-2 via Google Earth Engine
- **Elevation:** SRTM Digital Elevation Model
- **Weather:** OpenWeatherMap API
- **Historical Events:** 2022 and 2025 flood data

#### Technical Stack
- **Backend:** Python, scikit-learn, TensorFlow, Google Earth Engine
- **Frontend:** Streamlit
- **Deployment:** Hugging Face Spaces
- **Geospatial:** Folium, Geopandas, GDAL

#### Impact
- Early warning system for flood-prone communities
- Optimized emergency resource allocation
- Reduced flood-related casualties and economic losses
- Enhanced disaster preparedness and response""")
    st.markdown("---")
    st.markdown("### 📧 Contact & Support")
    st.info("For questions or support, please contact the development team.")

# Footer
st.markdown("---")
st.markdown('<div style="text-align: center; color: gray;"><p>© 2025 Charsadda Flood Prediction System | Powered by AI & Satellite Data</p></div>', unsafe_allow_html=True)