import gradio as gr
import pandas as pd
import pmdarima as pm
from pmdarima import auto_arima
import matplotlib.pyplot as plt

def forecast_time_series(file):
    # Load data
    data = pd.read_csv(file.name, skiprows=2)
    period_type = data.columns[0]
    data[period_type] = pd.to_datetime(data[period_type])
    data.set_index(period_type, inplace=True)

    # Fit the SARIMAX model automatically
    model = auto_arima(data)
    
    # Forecasting
    n_periods = 24  # Number of periods to forecast
    forecast, conf_int = model.predict(n_periods=n_periods, return_conf_int=True)
    
    # Create a DataFrame with the forecast and confidence intervals
    forecast_index = pd.date_range(start=data.index[-1], periods=n_periods + 1, freq=data.index.inferred_freq)[1:]
    forecast_df = pd.DataFrame(forecast, index=forecast_index, columns=['Forecast'])
    conf_int_df = pd.DataFrame(conf_int, index=forecast_index, columns=['Lower CI', 'Upper CI'])
    forecast_df = pd.concat([forecast_df, conf_int_df], axis=1)

    # Calculate the YoY change
    sum_last_12_original = data.iloc[-12:, 0].sum()
    sum_first_12_forecast = forecast_df['Forecast'].iloc[:12].sum()
    yoy_change = (sum_first_12_forecast - sum_last_12_original) / sum_last_12_original

    # Plot the original time series and forecast
    plt.figure(figsize=(12, 6))
    plt.plot(data.index, data.iloc[:, 0], label='Original Series')
    plt.plot(forecast_df.index, forecast_df['Forecast'], color='red', label='Forecast')
    plt.fill_between(forecast_df.index, 
                     forecast_df['Lower CI'], 
                     forecast_df['Upper CI'], 
                     color='pink', alpha=0.3, label='Confidence Interval')
    plt.xlabel('Date')
    plt.ylabel('Values')
    plt.title('Original Time Series and Forecast with Confidence Intervals')
    plt.legend()
    
    # Save plot to a file
    plot_file = 'forecast_plot.png'
    plt.savefig(plot_file)
    plt.close()
    
    # Return plot file path and YoY change
    return plot_file, f'Year-over-Year Change in Sum of Values: {yoy_change:.2%}'

# Create Gradio interface
interface = gr.Interface(
    theme=gr.themes.Monochrome(),
    fn=forecast_time_series,
    inputs=gr.File(label="Upload Time Series CSV"),
    outputs=[
        gr.Image(label="Time Series + Forecast Chart"),
        gr.Text(label="YoY % Change")
    ],
    title="Time Series Forecasting",
    description="Upload a CSV file with a time series to forecast the next 24 periods and see the YoY % change."
)

# Launch the interface
interface.launch()