app.py

import gradio as gr
import requests
import pandas as pd
from sklearn.linear_model import LogisticRegression
import os

# Use environment variables for API keys
API_KEY = os.environ.get("OPENWEATHER_API_KEY")
BASE_URL = 'https://api.openweathermap.org/data/2.5/'

def get_weather_data(location):
    current_weather_url = f'{BASE_URL}weather?q={location}&appid={API_KEY}&units=metric'
    current_response = requests.get(current_weather_url)
    current_data = current_response.json()

    current_weather = {
        'temperature': current_data['main']['temp'],
        'feels_like': current_data['main']['feels_like'],
        'description': current_data['weather'][0]['description'],
        'wind_speed': current_data['wind']['speed'],
        'pressure': current_data['main']['pressure'],
        'humidity': current_data['main']['humidity'],
        'visibility': current_data['visibility'] / 1000,
        'dew_point': current_data['main']['temp'] - ((100 - current_data['main']['humidity']) / 5.0)
    }

    return current_weather

def train_fog_model():
    df = pd.read_csv('fog_weather_data.csv')
    df = pd.get_dummies(df, columns=['Description'], drop_first=True)
    X = df.drop('Fog', axis=1)
    y = df['Fog']
    model = LogisticRegression()
    model.fit(X, y)
    return model, X.columns

def predict_fog(model, feature_columns, weather_data):
    new_data = pd.DataFrame({
        'Temperature': [weather_data['temperature']],
        'Feels like': [weather_data['feels_like']],
        'Wind speed': [weather_data['wind_speed']],
        'Pressure': [weather_data['pressure']],
        'Humidity': [weather_data['humidity']],
        'Dew point': [weather_data['dew_point']],
        'Visibility': [weather_data['visibility']]
    })

    for col in feature_columns:
        if col.startswith('Description_'):
            new_data[col] = 0

    description_column = f"Description_{weather_data['description'].replace(' ', '_')}"
    if description_column in feature_columns:
        new_data[description_column] = 1

    prediction = model.predict(new_data)
    return "Foggy weather" if prediction[0] == 1 else "Clear weather"

# Load the model once when the app starts
fog_model, feature_columns = train_fog_model()

def predict_current_weather(location):
    try:
        current_weather = get_weather_data(location)
        fog_prediction = predict_fog(fog_model, feature_columns, current_weather)
        
        result = f"Current weather in {location}:\n"
        result += f"Temperature: {current_weather['temperature']}°C\n"
        result += f"Feels like: {current_weather['feels_like']}°C\n"
        result += f"Description: {current_weather['description']}\n"
        result += f"Wind speed: {current_weather['wind_speed']} m/s\n"
        result += f"Pressure: {current_weather['pressure']} hPa\n"
        result += f"Humidity: {current_weather['humidity']}%\n"
        result += f"Dew point: {current_weather['dew_point']}°C\n"
        result += f"Visibility: {current_weather['visibility']} km\n"
        result += f"\nFog Prediction: {fog_prediction}"
        
        return result
    except Exception as e:
        return f"Error: {str(e)}"

def predict_custom_weather(temperature, feels_like, wind_speed, pressure, humidity, visibility, description):
    try:
        weather_data = {
            'temperature': temperature,
            'feels_like': feels_like,
            'wind_speed': wind_speed,
            'pressure': pressure,
            'humidity': humidity,
            'visibility': visibility,
            'description': description,
            'dew_point': temperature - ((100 - humidity) / 5.0)
        }
        
        fog_prediction = predict_fog(fog_model, feature_columns, weather_data)
        
        result = "Custom weather prediction:\n"
        result += f"Fog Prediction: {fog_prediction}"
        
        return result
    except Exception as e:
        return f"Error: {str(e)}"

# Gradio interface
with gr.Blocks() as demo:
    gr.Markdown("# Weather and Fog Prediction")
    
    with gr.Tab("Current Weather Prediction"):
        location_input = gr.Textbox(label="Enter Location")
        predict_button = gr.Button("Predict Weather")
        output = gr.Textbox(label="Prediction Result")
        
        predict_button.click(predict_current_weather, inputs=location_input, outputs=output)
    
    with gr.Tab("Custom Weather Prediction"):
        with gr.Row():
            temperature = gr.Number(label="Temperature (°C)")
            feels_like = gr.Number(label="Feels Like (°C)")
            wind_speed = gr.Number(label="Wind Speed (m/s)")
            pressure = gr.Number(label="Pressure (hPa)")
        with gr.Row():
            humidity = gr.Number(label="Humidity (%)")
            visibility = gr.Number(label="Visibility (km)")
            description = gr.Dropdown(label="Weather Description", choices=["clear sky", "few clouds", "scattered clouds", "broken clouds", "shower rain", "rain", "thunderstorm", "snow", "mist"])
        
        custom_predict_button = gr.Button("Predict Fog")
        custom_output = gr.Textbox(label="Prediction Result")
        
        custom_predict_button.click(
            predict_custom_weather,
            inputs=[temperature, feels_like, wind_speed, pressure, humidity, visibility, description],
            outputs=custom_output
        )

demo.launch()