app.py

import streamlit as st
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from xgboost import XGBRegressor
from sklearn.model_selection import train_test_split
from statsmodels.tsa.holtwinters import ExponentialSmoothing

import streamlit as st
import pandas as pd
from sklearn.ensemble import RandomForestRegressor
from statsmodels.tsa.holtwinters import ExponentialSmoothing

st.set_page_config(page_title="Superstore Analytics", layout="wide")

st.title("📊 Superstore Data Science Project")

st.markdown("""
This web app demonstrates:
- Profit prediction using Machine Learning
- Sales forecasting using Time Series analysis
""")

# Load data
df = pd.read_csv("data/Sample - Superstore.csv", encoding="latin1")
df['Order Date'] = pd.to_datetime(df['Order Date'])

menu = st.sidebar.selectbox(
    "Select Option",
    ["Dataset Preview", "Profit Prediction", "Sales Forecast"]
)

# ================= Dataset Preview =================
if menu == "Dataset Preview":
    st.subheader("Dataset Preview")
    st.dataframe(df.head(20))

# ================= Profit Prediction =================
elif menu == "Profit Prediction":
    st.subheader("🔮 Profit Prediction")

    df_model = df.copy()
    df_model['Year'] = df_model['Order Date'].dt.year
    df_model['Month'] = df_model['Order Date'].dt.month

    X = pd.get_dummies(
        df_model[['Sales', 'Quantity', 'Discount', 'Year', 'Month', 'Category', 'Region']],
        drop_first=True
    )
    y = df_model['Profit']

    model = RandomForestRegressor(n_estimators=100, random_state=42)
    model.fit(X, y)

    sales = st.number_input("Sales", value=500.0)
    quantity = st.number_input("Quantity", value=2)
    discount = st.slider("Discount", 0.0, 0.5, 0.1)

    input_df = pd.DataFrame({
        "Sales": [sales],
        "Quantity": [quantity],
        "Discount": [discount],
        "Year": [2025],
        "Month": [1],
        "Category_Technology": [1],
        "Region_West": [1]
    })

    input_df = input_df.reindex(columns=X.columns, fill_value=0)

    if st.button("Predict Profit"):
        prediction = model.predict(input_df)[0]
        st.success(f"💰 Predicted Profit: {prediction:.2f}")

# ================= Sales Forecast =================
elif menu == "Sales Forecast":
    st.subheader("📈 Sales Forecast")

    monthly_sales = (
        df
        .set_index('Order Date')
        .resample('ME')['Sales']
        .sum()
    )

    model = ExponentialSmoothing(
        monthly_sales,
        trend='add',
        seasonal='add',
        seasonal_periods=12
    ).fit()

    forecast = model.forecast(6)

    st.line_chart(pd.concat([monthly_sales, forecast]))

    st.header("📂 Upload Your Dataset")

uploaded_file = st.file_uploader(
    "Upload a CSV file (Sales data)",
    type=["csv"]
)

if uploaded_file is not None:
    df = pd.read_csv(uploaded_file, encoding="latin1")
    st.subheader("Dataset Preview")
    st.dataframe(df.head())


    st.subheader("Select Columns")

    date_col = st.selectbox("Select Date Column", df.columns)
    sales_col = st.selectbox("Select Sales Column", df.columns)

    profit_col = st.selectbox(
        "Select Profit Column (optional)",
        ["None"] + list(df.columns)
    )

    df[date_col] = pd.to_datetime(df[date_col], errors="coerce")
    df = df.dropna(subset=[date_col, sales_col])

    df = df.sort_values(date_col)

    monthly_sales = (
        df
        .set_index(date_col)
        .resample("M")[sales_col]
        .sum()
    )

    st.subheader("📈 Sales Forecast")

    try:
        model_ts = ExponentialSmoothing(
            monthly_sales,
            trend="add",
            seasonal="add",
            seasonal_periods=12
        ).fit()

        forecast = model_ts.forecast(6)

        fig, ax = plt.subplots()
        ax.plot(monthly_sales, label="Actual Sales")
        ax.plot(forecast, label="Forecast", linestyle="--")
        ax.legend()
        st.pyplot(fig)

    except Exception as e:
        st.error("Not enough data for forecasting.")


    if profit_col != "None":
        st.subheader("💰 Profit Prediction (XGBoost)")

        df_ml = df.dropna(subset=[profit_col])
        df_ml["Year"] = df_ml[date_col].dt.year
        df_ml["Month"] = df_ml[date_col].dt.month

        X = df_ml[[sales_col, "Year", "Month"]]
        y = df_ml[profit_col]

        X_train, X_test, y_train, y_test = train_test_split(
            X, y, test_size=0.2, random_state=42
        )

        model = XGBRegressor(
            n_estimators=300,
            learning_rate=0.05,
            max_depth=5,
            random_state=42
        )
        model.fit(X_train, y_train)

        sales_input = st.number_input("Enter Sales", value=float(X[sales_col].mean()))

        input_df = pd.DataFrame({
            sales_col: [sales_input],
            "Year": [2025],
            "Month": [1]
        })

        pred_profit = model.predict(input_df)[0]
        st.success(f"Predicted Profit: {pred_profit:.2f}")