Python Packages – with their popularity and how they're commonly used together (opens in new tab)

(package.wiki)

12 pointsrobertlacok5y ago3 comments

3 comments

Hey, I found a cool BigQuery dataset with files from GitHub and got really curious about which packages are often used together. I joined it with download counts to estimate popularity.

You can start with a cool package like PyTorch: https://package.wiki/torch or Spacy: https://package.wiki/spacy and explore what people use them with.

Let me know what you think and if you stumble upon some curious insights!

Equiet5y ago

Looks really good, but how is this different from PyPI? Is it just the design or are there more features planned?

robertlacokOP5y ago

PyPI is the ground truth for a lot of the official information about packages, and package.wiki reuses some of it (like short descriptions) from the psf BigQuery dataset.

The initial idea was to do a fun data science project which builds some extra layers on top of that information - such as easier discovery, and showcase how popular the packages are.

It provides an extra context, similar to what https://bundlephobia.com/ does to https://www.npmjs.com

j / k navigate · click thread line to collapse

3 comments

robertlacokOP5y ago

Hey, I found a cool BigQuery dataset with files from GitHub and got really curious about which packages are often used together. I joined it with download counts to estimate popularity.

You can start with a cool package like PyTorch: https://package.wiki/torch or Spacy: https://package.wiki/spacy and explore what people use them with.

Let me know what you think and if you stumble upon some curious insights!

Equiet5y ago

Looks really good, but how is this different from PyPI? Is it just the design or are there more features planned?

robertlacokOP5y ago

PyPI is the ground truth for a lot of the official information about packages, and package.wiki reuses some of it (like short descriptions) from the psf BigQuery dataset.

The initial idea was to do a fun data science project which builds some extra layers on top of that information - such as easier discovery, and showcase how popular the packages are.

It provides an extra context, similar to what https://bundlephobia.com/ does to https://www.npmjs.com

j / k navigate · click thread line to collapse