app.py

import matplotlib
matplotlib.use('Agg')  # Set the backend to Agg
from flask import Flask, request
from flask import render_template
from components.model_prediction import Prediction
from Support_module_dir.support_function_plot import combine_plot_function
import io
import base64


# Function to load the model and make predictions
def load_and_predict_model(forecast_days):
    try:
        saved_model_path = "Saved_Model_dir/2023-12-23_00_42_22/SARIMAX_FORCAST_MODEL.joblib"

        # Call the function to load and predict
        predictions_instance, dataframe_instance = Prediction(saved_model_path).model_prediction(
            forecast_days)  # called class Prediction

        # Reset the index and move it into a new column
        dataframe_instance = dataframe_instance[-7:].reset_index()

        # Assuming predictions_instance is your DataFrame
        predictions_instance['Mean_Price'] = (predictions_instance['Lower_Bound'] + predictions_instance[
            'Upper_Bound']) / 2

        # Create an in-memory buffer
        buffer = io.BytesIO()

        # calling plot function
        combine_plot_function(dataframe_instance, predictions_instance, buffer)

        # getting image decode string
        plot_img_str = base64.b64encode(buffer.getvalue()).decode()

        return predictions_instance, dataframe_instance, plot_img_str

    except FileNotFoundError as file_error:
        raise FileNotFoundError(f"Error loading model: {file_error}")
    except Exception as e:
        raise e


# Initialize Flask app
app = Flask(__name__)


# Route for the home page
@app.route('/', methods=['GET'])
def home():
    return render_template("about.html")


# Route for the prediction page
@app.route('/predict', methods=['POST'])
def predict():
    input_data = int(request.form['forecast'])
    predictions_instance, dataframe_instance, plot_img_str = load_and_predict_model(forecast_days=input_data)
    try:

        return render_template('index.html', predictions=predictions_instance.to_dict(orient="records"),
                               dataframe=dataframe_instance.to_dict(orient="records"),
                               plot_img_str=plot_img_str)
    except Exception as e:
        return render_template('error.html', error_message=str(e))

        
# Run the app
if __name__ == '__main__':
    app.run(host='0.0.0.0', port=7860)