app.py

import streamlit as st
import pandas as pd
import joblib
import numpy as np

# Load the trained model
@st.cache_resource
def load_model():
    return joblib.load("superkart_sales_prediction_model_v1_0.joblib")

model = load_model()

# Streamlit UI for Sales Prediction
st.title("SuperKart Product Store Sales Prediction App")
st.write("This app predicts the Total revenue generated by the sale of a particular product in a particular store.")

st.subheader("Enter the product and store details:")

# Collect user input for SuperKart features
product_weight = st.number_input("Product Weight", min_value=4.0, max_value=22.0, value=12.0)
product_allocated_area = st.number_input("Product Allocated Area", min_value=0.004, max_value=0.298, value=0.05)
product_mrp = st.number_input("Product MRP", min_value=31.0, max_value=266.0, value=150.0)
store_age = st.number_input("Store Age", min_value=16, max_value=38, value=20) # Derived from Store_Establishment_Year

product_identifier = st.selectbox("Product Identifier", ["FD", "NC", "DR"])
product_sugar_content = st.selectbox("Product Sugar Content", ["Low Sugar", "Regular", "No Sugar"])
store_id = st.selectbox("Store ID", ['OUT002', 'OUT003', 'OUT004', 'OUT010', 'OUT013', 'OUT017', 'OUT018', 'OUT019', 'OUT027']) # Assuming these are the store IDs after one-hot encoding
store_size = st.selectbox("Store Size", ["High", "Medium", "Small"])
store_location_city_type = st.selectbox("Store Location City Type", ["Tier 1", "Tier 2", "Tier 3"])
store_type = st.selectbox("Store Type", ["Departmental Store", "Supermarket Type1", "Supermarket Type2", "Food Mart"])


# Create a dictionary from user input for prediction
user_input_data = {
    'Product_Weight': product_weight,
    'Product_Allocated_Area': product_allocated_area,
    'Product_MRP': product_mrp,
    'Store_Age': store_age,
    'Product_Identifier': product_identifier,
    'Product_Sugar_Content_No Sugar': product_sugar_content == 'No Sugar',
    'Product_Sugar_Content_Regular': product_sugar_content == 'Regular',
    'Product_Sugar_Content_reg': product_sugar_content == 'reg', # Check if 'reg' is a valid category
    'Product_Type_Breads': False, # Placeholder - need to handle Product Type encoding based on user selection
    'Product_Type_Breakfast': False, # Placeholder
    'Product_Type_Canned': False, # Placeholder
    'Product_Type_Dairy': False, # Placeholder
    'Product_Type_Frozen Foods': False, # Placeholder
    'Product_Type_Fruits and Vegetables': False, # Placeholder
    'Product_Type_Hard Drinks': False, # Placeholder
    'Product_Type_Health and Hygiene': False, # Placeholder
    'Product_Type_Household': False, # Placeholder
    'Product_Type_Meat': False, # Placeholder
    'Product_Type_Others': False, # Placeholder
    'Product_Type_Seafood': False, # Placeholder
    'Product_Type_Snack Foods': False, # Placeholder
    'Product_Type_Soft Drinks': False, # Placeholder
    'Product_Type_Starchy Foods': False, # Placeholder
    'Store_Id_OUT002': store_id == 'OUT002',
    'Store_Id_OUT003': store_id == 'OUT003',
    'Store_Id_OUT004': store_id == 'OUT004',
    'Store_Id_OUT010': store_id == 'OUT010', # Assuming OUT010, OUT013, OUT017, OUT018, OUT019, OUT027 are also present in the training data
    'Store_Id_OUT013': store_id == 'OUT013',
    'Store_Id_OUT017': store_id == 'OUT017',
    'Store_Id_OUT018': store_id == 'OUT018',
    'Store_Id_OUT019': store_id == 'OUT019',
    'Store_Id_OUT027': store_id == 'OUT027',
    'Store_Size_Medium': store_size == 'Medium',
    'Store_Size_Small': store_size == 'Small',
    'Store_Location_City_Type_Tier 2': store_location_city_type == 'Tier 2',
    'Store_Location_City_Type_Tier 3': store_location_city_type == 'Tier 3',
    'Store_Type_Food Mart': store_type == 'Food Mart',
    'Store_Type_Supermarket Type1': store_type == 'Supermarket Type1',
    'Store_Type_Supermarket Type2': store_type == 'Supermarket Type2'
}

# Convert the dictionary to a DataFrame with a single row
input_data_df = pd.DataFrame([user_input_data])

# Need to handle Product Type encoding based on user selection and add all expected Product Type columns
# This requires knowing all possible Product Types from the training data

# Predict button
if st.button("Predict"):
    # Perform prediction
    prediction = model.predict(input_data_df)

    # Display the prediction
    st.write(f"The predicted Product Store Sales Total is: {prediction[0]:.2f}")