For #metasynth (:python: pkg for generating synthetic tables from sensitive data) we're using #polars (https://pola.rs) as our data frame library.

I had to get used to it (it's not #tidyverse ) but I'm really liking how consistent (and fast!) it is. Also good thought went into data types. For this, I like it much better than #pandas!

The only minor problem is that it is not yet easy to find answers to some basic questions, because there are not that many users.

Use polars everyone!