You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Pandas 2.0.0 was released on April 3rd 2023.
The last version of Pandas 1 was Pandas 1.5.3 and it was released on January 19th 2023. That version only supports Python 3.10 and 3.11.
We are currently supporting up to Pandas 1.3.4 which was released in 2021 and only provides wheels for Python 3.10 and earlier.
I'm not following pandas development very closely, but I believe this to be fine. pandas 2 came out almost three years ago; it's not formally part of NEP29/SPEC0, but that should be more than enough
Activity
raulcd commentedon Jan 9, 2026
Maybe I am being naive here but what are your thoughts on this?
cc @pitrou @AlenkaF @WillAyd @jorisvandenbossche
WillAyd commentedon Jan 9, 2026
That seems reasonable to me
pitrou commentedon Jan 9, 2026
Unfortunately, I couldn't easily find Pandas' supported versions on the Internet.
pitrou commentedon Jan 9, 2026
The latest commits on the 1.5.x branch are from February 2023. It seems that Pandas 1.x is pining for the trees.
pitrou commentedon Jan 9, 2026
@h-vetinari Do you have any opinions on this?
h-vetinari commentedon Jan 9, 2026
I'm not following pandas development very closely, but I believe this to be fine. pandas 2 came out almost three years ago; it's not formally part of NEP29/SPEC0, but that should be more than enough
jorisvandenbossche commentedon Jan 12, 2026
Slightly related issue that was opened last month -> #48473 (but we can keep that for the numpy part).
As mentioned there as well, I think we can definitely drop support for pandas 1.x
raulcd commentedon Jun 29, 2026
this doesn't seem like a blocker, we might bump this to 26.0.0.
apacheGH-48808: [Python] Investigate what entails bumping pandas >=2
GH-48808: [Python] Drop support for Pandas < 2.0.3 (#50444)
pitrou commentedon Jul 9, 2026
Issue resolved by pull request 50444
#50444
force pyarrow<26 because 26 dropped pandas 2 support (#1768)