Comment by skylurk

Luckily, polars has .to_pandas() so you can still pass pandas dataframes to the libraries that really are still stuck on that interface.

I maintain one of those libraries and everything is polars internally.