I think I've finally uncovered my central frustration with pandas.
-
-
I'd just write SQL if I wanted SQL.
-
What I want is to be able to do basic filter-map-apply on data without having to worry whether it's going to magically change type
-
doesnt help that frames/series/multiindex-frames act mostly the same so you dont realize something's changed type until oops, broken
-
Yep--and nonsensical inconsistence with method names
End of conversation
New conversation -
-
-
I thought this would be about real pandas.
-
My frustration with pandas/numpy/scipy is where pip install doesn't work with them in a requirements.txt because maintainers blame pip...
-
You need numpy installed before any of the rest or shit breaks and as an ops eng it's frustrating.
End of conversation
New conversation -
-
-
Ah. The magic word. "academics". Designed for people who don't live in the real world.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Ive definitely been bit by pandas types, but why not work in straight numpy if ur more concerned about mapping functions over the whole set?
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Oh yeah, I am 100% with you. I have similar feelings about data.table in R. OTOH, I do like how dplyr works with RDBMSs.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.