app.py

from get_forecast import get_forecast
import gradio as gr

with gr.Blocks() as demo:

    gr.Markdown(
    """
    # AB InBev Data Science Challenge
    This is a demo built for AB InBev company to forecast **SOM** and **Volume** series. This demo allows the user to analyze the price sensibility of the mentioned series. Enjoy!
    ## How do you use this demo?
    Using this demo is easy! Simply follow the next steps:
    1. Select a Serie to forecast from the dropdown (valid values: SOM and Volumen)
    2. Choose how many months ahead do you wish your forecast to predict
    3. Pick the price variation vs last year you want to analyze. Setting this value to 0 means the future price values are exactly the same as the last year. 
    """)

    with gr.Row():
        serie = gr.Dropdown(choices = ['SOM', 'Volumen'], value = 'SOM', label = 'Serie', info = 'Choose the serie to forecast')
        periods = gr.Slider(minimum = 1, maximum = 12, step = 1, value = 3, label = 'Months')
        percent_change = gr.Slider(minimum = -100, maximum = 100, step = 5, value = -5, label = '% Change vs Last Year')

    plot = gr.Plot()

    serie.change(get_forecast, [serie, periods, percent_change], plot, queue=False)
    periods.change(get_forecast, [serie, periods, percent_change], plot, queue=False)
    percent_change.change(get_forecast, [serie, periods, percent_change], plot, queue=False)    
    plot.change(get_forecast, [serie, periods, percent_change], plot, queue=False)    
    demo.load(get_forecast, [serie, periods, percent_change], plot, queue=False)

    demo.launch(debug = False)