Repository navigation
Support for numpy >= 2.0.2 #432
Description
I am working on upgrading our ecosystem to use numpy >= 2.0.2 which currently breaks at your package.
Afaik you don't use numpy in your code except for testing. Thus, I am wondering if you can either drop the dependency and add it as dev-only dependency or if you can be less strict in the versioning of numpy and support numpy >= 2.0.2.
Here is also a list of the numpy ecosystem compatibility with numpy 2.0:
numpy/numpy#26191
Kind regards
Activity
[-]Support for numpy >= 2.0.2[/-][+]Support for numpy >= 2.0.2 and pandas >= 2.2.2[/+][-]Support for numpy >= 2.0.2 and pandas >= 2.2.2[/-][+]Support for numpy >= 2.0.2[/+]arredond commentedon Oct 4, 2024
+1 to this! Also in general to not setting upper bounds to package versions unless strictly necessary.
I made a similar comment about this here #427 (similar case that's affecting us but for the
pyarrow < 17.0limitationsshevlyagin commentedon Oct 9, 2024
+1
jslorrma commentedon Oct 26, 2024
+1
Bendik-ml commentedon Nov 22, 2024
+1 👯♂️
tblatrille commentedon Dec 10, 2024
+1
jprakash-db commentedon Jan 31, 2025
Pandas and numpy has compatiblity issues, numpy version 2 is only supported after pandas 2.2.2 and any version before that is not supported. - https://pandas.pydata.org/docs/whatsnew/v2.2.2.html#pandas-2-2-2-is-now-compatible-with-numpy-2-0
rgommers commentedon Jan 31, 2025
Newer versions of one package only being compatible with newer versions of another package is perfectly normal, rather than a problem that is relevant to this package. This is still a problem, so there is no reason to close this issue.
That is entirely correct.