app.py

import numpy as np
import joblib
import pandas as pd
from flask import Flask, request, jsonify
from datetime import datetime

# Initialize the Flask application
sales_predictor_api = Flask("Super Kart Predictor")

# Load the trained machine learning model
model = joblib.load("SuperKart_model_v1_0.joblib")

# Define a route for the home page (GET request)
@sales_predictor_api.get('/')
def home():
    """
    Welcome endpoint for the Retail Sales Prediction API
    """
    return "Welcome to the Retail Product Sales Prediction API!"

# Define an endpoint for single product prediction (POST request)
@sales_predictor_api.post('/v1/sales')
def predict_sales():
    """
    Endpoint for predicting sales for a single product-store combination
    Expects JSON payload with product and store details
    """
    try:
        # Get the JSON data from the request body
        data = request.get_json()
        
        # Prepare input data with only the features used in training
        input_data = pd.DataFrame([{
            'Product_Weight': data['product_weight'],
            'Product_Allocated_Area': data['product_allocated_area'],
            'Product_MRP': data['product_mrp'],
            'Store_Establishment_Year': data['store_establishment_year'],
            'Product_Sugar_Content': data['product_sugar_content'],
            'Product_Type': data['product_type'],
            'Store_Size': data['store_size'],
            'Store_Location_City_Type': data['store_location_city_type'],
            'Store_Type': data['store_type']
        }])
        
        # Make prediction
        predicted_sales = model.predict(input_data)[0]
        
        # Return the predicted sales
        return jsonify({
            'predicted_sales': round(float(predicted_sales), 2),
            'features_used': list(input_data.columns)
        })
        
    except Exception as e:
        return jsonify({'error': str(e)}), 400

# Define an endpoint for batch prediction (POST request)
@sales_predictor_api.post('/v1/salesbatch')
def predict_sales_batch():
    """
    Endpoint for batch predictions from CSV file
    Expects CSV file with multiple product-store combinations
    """
    try:
        # Check if file was uploaded
        if 'file' not in request.files:
            return jsonify({'error': 'No file uploaded'}), 400
            
        file = request.files['file']
        
        # Read CSV file
        input_data = pd.read_csv(file)
        
        # Ensure we only keep the columns used in training
        required_columns = [
            'Product_Weight',
            'Product_Allocated_Area',
            'Product_MRP',
            'Store_Establishment_Year',
            'Product_Sugar_Content',
            'Product_Type',
            'Store_Size',
            'Store_Location_City_Type',
            'Store_Type'
        ]
        
        # Verify all required columns are present
        missing_cols = [col for col in required_columns if col not in input_data.columns]
        if missing_cols:
            return jsonify({'error': f'Missing required columns: {missing_cols}'}), 400
        
        # Make predictions
        predictions = model.predict(input_data[required_columns])
        
        # Prepare results
        results = []
        for i, row in input_data.iterrows():
            results.append({
                'predicted_sales': round(float(predictions[i]), 2),
                'product_type': row['Product_Type'],
                'store_type': row['Store_Type']
            })
            
        return jsonify({
            'predictions': results,
            'features_used': required_columns
        })
        
    except Exception as e:
        return jsonify({'error': str(e)}), 400

# Run the Flask application
if __name__ == '__main__':
    sales_predictor_api.run(host='0.0.0.0', port=5000, debug=True)