app.py

import joblib
import pandas as pd
from flask import Flask, request, jsonify

# Initialize Flask app with a name
SuperKart_predictor_api = Flask ("SuperKart Predictor")

# Load the trained churn prediction model
model = joblib.load ("SuperKart_prediction_model_v1_0.joblib")

# Define a route for the home page
@SuperKart_predictor_api.get ('/')
def home ():
    return "Welcome to the Super Kart Sales Prediction!"

# Define an endpoint to predict sales for Super Kart
@SuperKart_predictor_api.post ('/v1/SuperKartSales')
def predict_sales ():
    # Get JSON data from the request
    sales_data = request.get_json ()

    import datetime

    current_year = datetime.datetime.now ().year   # dynamic current year

    # Extract relevant features from the input data
    data_info = {
        'Product_Weight'            : sales_data ['Product_Weight'],
        'Product_Sugar_Content'     : sales_data ['Product_Sugar_Content'],
        'Product_Allocated_Area'    : sales_data ['Product_Allocated_Area'],
        'Product_Type'              : sales_data ['Product_Type'],
        'Product_MRP'               : sales_data ['Product_MRP'],
        'Store_Id'                  : sales_data ['Store_Id'],
        'Store_Size'                : sales_data ['Store_Size'],
        'Store_Location_City_Type'  : sales_data ['Store_Location_City_Type'],
        'Store_Type'                : sales_data ['Store_Type']
    }

    # Convert the extracted data into a DataFrame
    input_data = pd.DataFrame ([data_info])

    # Enforce types
    #input_data["Store_Establishment_Year"] = input_data["Store_Establishment_Year"].astype (int)
    numeric_cols = ["Product_Weight", "Product_Allocated_Area", "Product_MRP"]
    input_data[numeric_cols] = input_data[numeric_cols].astype (float)

    # Make a churn prediction using the trained model
    #prediction = model.predict (input_data)
    predicted_sales = model.predict (input_data).tolist ()[0]

    # Return the prediction as a JSON response
    return jsonify ({'SalesPrediction': predicted_sales})
    #return jsonify ({'SalesPrediction': 6789.56})


# Define an endpoint to predict sales for a batch of data
# here we assume the data to conatain same columns as per the data provided for this project
@SuperKart_predictor_api.post ('/v1/SuperKartBatchSales')
def predict_churn_batch ():
    # Get the uploaded CSV file from the request
    file = request.files ['file']

    # Read the file into a DataFrame
    input_data = pd.read_csv (file)

    # Handle Product_Id if present
    if "Product_Id" in input_data.columns:
        product_ids = input_data["Product_Id"].copy ()
        X = input_data.drop(columns=["Product_Id"])
    else:
        product_ids = None
        X = input_data

    # Make predictions
    predictions = model.predict (X).tolist ()

    # Prepare response
    if product_ids is not None:
        # Return mapping of Product_Id to predictions
        output_dict = dict(zip(product_ids.tolist (), predictions))
    else:
        # If no Product_Id, just return index → predictions
        output_dict = dict(zip(input_data.index.tolist (), predictions))

    return jsonify (output_dict)

# Run the Flask app
if __name__ == "__main__":
    import os
    port = int (os.environ.get("PORT", 7860))
    SuperKart_predictor_api.run(host="0.0.0.0", port=port)