Stargazers intersections for most popular GitHub projects in Venn diagrams

twitter.com

31 points·ashvardanian··12 comments

https://nitter.net/ashvardanian/status/1723393012728758565

> There should be more overlap between NetworkX and NumPy, than with Pandas, esp among Spectral Graph Theory practicioners

Sounds partly like sampling bias to me. I do not at all think it's safe to assume that GitHub users that star libraries actively is a representative random sample of all users of those libraries.

But it also sounds like a misunderstanding of just how many people think of "Numpy + Pandas" together as forming the foundations of the data analysis and data processing toolkit in Python. Almost every project where I use the latter, I end up importing the former for some reason or other, even if I'm not doing anything heavily numerical or computational.

Another interesting thing - find friends on GitHub using the intersection between stars you've gave: https://ghe.clickhouse.tech/#finding-friends-through-countin...

Super cool! Can I run this on a public instance somewhere?

It shouldn’t be hard to implement: https://ghe.clickhouse.tech/#how-to-download-the-data

It is already the public instance: https://play.clickhouse.com/play?user=play

I would assume this might been inspired by a talk given by Anthony Fu on ViteConf '23 - "Anthony's Road to Open Source: The Set Theory" https://www.youtube.com/watch?v=NJbCfAKtxUI ?

Haven’t seen that, thanks for sharing! I was thinking about implementing a set-search and clustering of similar nature here: https://ashvardanian.com/posts/abusing-vector-search/

Very cool! Where is the script? Can't find it

This: https://play.clickhouse.com/play?user=play#U0VMRUNUIGJpbihzZ...

It’s on my GitHub, mentioned in a preceding tweet: https://gist.github.com/ashvardanian/76bae6147fe558117652939...

Great. Now what?

Reply on news.ycombinator.com