app.py

import streamlit as st
import pandas as pd
import joblib

# Load model and encoders
@st.cache_resource
def load_model_and_data():
    model = joblib.load("deployment_files/Quarterly_Sales_Revenue_model_v1_0.joblib")
    data = pd.read_csv("/content/drive/MyDrive/SuperKart.csv")  # Update with your actual data file path
    return model, data
   # SuperKart =  pd.read_csv("/content/drive/MyDrive/SuperKart.csv")

# Prediction function
def predict_total_sales_by_store(store_ids, df, rf_tuned):
    results = []

    feature_cols = [
        'Product_Id', 'Store_Id', 'Product_Type', 'Product_Sugar_Content',
        'Store_Type', 'Store_Location_City_Type', 'Store_Size',
        'Product_Weight', 'Product_Allocated_Area', 'Product_MRP',
        'Store_Establishment_Year'
    ]

    for store_id in store_ids:
        store_df = df[df['Store_Id'] == store_id].copy()

        if store_df.empty:
            results.append((store_id, 0.0))
            continue

        X_input = store_df[feature_cols]
        predicted_sales = rf_tuned.predict(X_input).sum()
        results.append((store_id, round(predicted_sales, 2)))

    return results

# Load
rf_tuned, df = load_model_and_data()

# Streamlit UI
st.title("🧾 Predict Total Quarterly Sales per Store")

# Store ID selection
store_ids = st.multiselect("Select Store ID(s)", sorted(df["Store_Id"].unique()))

if st.button("Predict Sales"):
    if not store_ids:
        st.warning("Please select at least one Store ID.")
    else:
        results = predict_total_sales_by_store(store_ids, df, rf_tuned)
        st.subheader("📊 Predicted Sales:")
        for store_id, sales in results:
            st.write(f"🛒 Store {store_id}: ${sales:,.2f}")