Skip to content

A Streamlit application that provides sales data analysis and utilizes XGBoost to generate 90-day sales forecasts for individual products or product families.

Notifications You must be signed in to change notification settings

karol-pa/Automatic_Sales_Forecast_App

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

19 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Sales Forecast Project for SPICED Academy

This is an application for analyzing and forecasting future product sales based on historical sales numbers within the context of e-commerce. The app uses the XGBoost model.

Once you have installed all required packages

pip install -r requirements.txt

run the following command from the terminal to open the app in your default web browser:

streamlit run Home.py

Or try the cloud version (Please note that the XGBoost model used for forecasting will probably run slower on the (free) cloud server than on your local machine.)

At the moment, two datasets are already preloaded: Product families A and B.

  • Home: Select one of the product families.
  • Time series data: View sales numbers and Warehouse stock over time. Select either all products or individual products. Select Sales/stock by product or by country (via the tab above the chart). You can also apply corrections to the data (look at the tooltips to see how the corrections impact the sales data).
  • Aggregated data: View the seasonalities hidden in the sales data.
  • Map: See the sales number associated with different European countries.
  • Sales forecast: Here you can make 90-day sales forecasts. Select one or more products, corrections and the date from which you want to start the prediction. Everything before that date is used to train the forecasting model. So, if you set the Start date for prediction to August, 1'st, 2023 (or before), you can test the model performance against the actual sales.

About

A Streamlit application that provides sales data analysis and utilizes XGBoost to generate 90-day sales forecasts for individual products or product families.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages