prediction.py

import streamlit as st
import pandas as pd
import pickle
import json
# from statsmodels.tsa.holtwinters import ExponentialSmoothing

# Load All Files

with open('model_forecasting.pkl', 'rb') as file_1:
    model = pickle.load(file_1)


def run():

    # create Streamlit app
    st.title('Demand Forecasting')

    # display prediction
    # st.write('Prediction:', prediction)

    with st.form(key='form_forecasting_2022'):
        start_date = st.date_input(
            'Start date', value=pd.to_datetime('2023-04-16'))
        end_date = st.date_input(
            'End date', value=pd.to_datetime('2023-04-30'))
        st.markdown('---')

        submitted = st.form_submit_button('Predict')

    if submitted:

        prediction = model.predict(start=start_date, end=end_date)
        st.write("Hasil prediksi pada tanggal ", start_date, ":", prediction)


# agar bisa di run sendiri, tanpa dari main
if __name__ == '__main__':
    run()