Re:Modern Pandas (Part 2): Method Chaining
by riyadparvez on Sunday May 01, 2022 @11:51AM
(#31226022)
The issue is there's a huge ecosystem developed on top of Pandas. So if it's different API, now you lose access to all the ecosystem.
Should not be a real blocker. There is good interop with arrow, numpy and pandas. Where arrow and numpy mostly is zero-copy.
So you are one `df.to_numpy()/df.to_pandas()` away to `X` libary you want to use.
It has a rather different api, and is significantly faster. Highly recommend it.