Small pyspark project

I have a small project. Just sample data. The goal is to calculate which top 2 group fluctuated the most in the last 48 hours of orders comparing to the historical data of the last 20 days. But needs to be done using pspark using python, kafka and MinIO. Everything is already setup in a server as docker containers. Docker compose file is also available (spark, kafka, minio, jupitar notebook as docker containers). Will provide access to the server. Let me know you have Questions, comments or for more information. Will provide attachment for full description upon request