As a pet project, I wanted to explore Steam data.

The acquisition and cleaning are more documented here: https://github.com/Duerkos/steam_analysis. In this dataset you can find the cleaned data, as well as some minor files that I extracted to support the analysis. I mainly extracted data from Steam API, Steam Spy (which at the end was not much useful), Steam Reviews API and the Steam web with a web-scraping tool ( https://scrapy.org/ )

A preliminary version (unclean, raw data and a bit more outdated) can be found here: https://www.kaggle.com/datasets/vicentearce/steam-and-steam-spy-raw-datasets

Although the process has diverged much from my original version, I could not have began to do it without following the work from Nik Davis, with minor diferences in the gathering code. See https://www.kaggle.com/nikdavis/steam-store-raw

If I had to do it again, I think I would use only scrapy and forget about the Steam API. Why? Well, the Steam API seems a bit outdated and there is data that is not being stored, such as Tags,Early Access, VR compatibility and now Steam Deck compatibility. It also allows a faster pulling rate than the Steam API which only allows 200 entries every 5 minutes (or something similar). However, it is true that it might contain some information which is not visible in the store.