Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.debian.bugs.dist > #1080901 > unrolled thread

Bug#999415: transition: pandas 1.1 -> 1.3

Back to article view | Back to linux.debian.bugs.dist

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  Bug#999415: transition: pandas 1.1 -> 1.3 "Rebecca N. Palmer" <rebecca_palmer@zoho.com> - 2021-11-10 20:30 +0100
    Bug#999415: transition: pandas 1.1 -> 1.3 Graham Inggs <ginggs@debian.org> - 2021-11-21 07:40 +0100
      Bug#999415: transition: pandas 1.1 -> 1.3 "Rebecca N. Palmer" <rebecca_palmer@zoho.com> - 2021-11-21 12:10 +0100
        Bug#999415: transition: pandas 1.1 -> 1.3 "Rebecca N. Palmer" <rebecca_palmer@zoho.com> - 2021-11-21 17:00 +0100
          Bug#999415: transition: pandas 1.1 -> 1.3 "Rebecca N. Palmer" <rebecca_palmer@zoho.com> - 2021-11-22 23:20 +0100
            Bug#999415: transition: pandas 1.1 -> 1.3 "Rebecca N. Palmer" <rebecca_palmer@zoho.com> - 2021-11-26 21:20 +0100
              Bug#999415: transition: pandas 1.1 -> 1.3 - to unstable now or not? "Rebecca N. Palmer" <rebecca_palmer@zoho.com> - 2021-11-28 20:20 +0100
                Bug#999415: transition: pandas 1.1 -> 1.3 - to unstable now or not? Scott Talbert <swt@techie.net> - 2021-11-28 23:10 +0100
                Bug#999415: transition: pandas 1.1 -> 1.3 "Rebecca N. Palmer" <rebecca_palmer@zoho.com> - 2021-11-29 22:00 +0100
                  Bug#999415: transition: pandas 1.1 -> 1.3 "Rebecca N. Palmer" <rebecca_palmer@zoho.com> - 2021-12-09 23:00 +0100

#1080901 — Bug#999415: transition: pandas 1.1 -> 1.3

Source: pandas
Version: 1.1.5+dfsg-2
Severity: wishlist

On 10/11/2021 17:14, Andreas Tille wrote:
> pandas is lagging behind upstream by several versions.  I guess we
> should try to get in sync with upstream a bit more.

Yes, but please don't upload this yet: it's common for a pandas upgrade 
to break reverse dependencies.

> The package does not build yet since the documentation needs sphinx_panels.
On 10/11/2021 18:32, Timo Röhling wrote> Maybe, as a compromise, we can 
cut out all the notebooks^H bells and
> whistles and limit the offline documentation to the API reference
> itself, which is arguably the most useful part? I don't know
> how easy it is to trim down the documentation, though.

The pandas (and probably statsmodels) documentation already is modified 
to build with fewer dependencies - see sphinx_no_pandas_theme.patch.

[toc] | [next] | [standalone]

#1082514

It might be easier to upgrade to 1.2.4 first, then 1,3,x later.

[toc] | [prev] | [next] | [standalone]

#1082527

Control: block 996584 by -1

On 21/11/2021 06:30, Graham Inggs wrote:
> It might be easier to upgrade to 1.2.4 first, then 1,3,x later.

Probably not, given that we're now trying to add Python 3.10 support 
(#996584), and pandas upstream's work on that is recent (maybe still 
unfinished): https://github.com/pandas-dev/pandas/issues/41940

The 1.3.x currently in Salsa built last time I tried (before Python 
3.10), but is (intentionally, to avoid FTBFS) missing some 
documentation, and I haven't tested the reverse dependencies yet.

[toc] | [prev] | [next] | [standalone]

#1082547

I have now tried 1.3.x with Python 3.10: 5 test failures due to 
matplotlib 3.5 with an upstream fix, 1 that might be python 3.10 or 
numpy 1.21 but is trivially fixable either way.

[toc] | [prev] | [next] | [standalone]

#1082742

1.3.4 is now in experimental.

Lots of tests failed on non-amd64 (the build "succeeded" because these 
were intentionally ignored), but it looks plausible that they're only a 
few easily-fixable problems.

Reverse dependencies testing is in progress.  (At least some of the 
failures in 
https://release.debian.org/britney/pseudo-excuses-experimental.html#pandas 
look like attempts to import something in Python 3.10 that hasn't been 
built for 3.10 yet, so probably just require waiting for that.)

[toc] | [prev] | [next] | [standalone]

#1083201

The pandas package currently in Salsa should fix the test failures on 
all release architectures (except mips* which haven't attempted a build 
yet), but hasn't yet been tested (anywhere).  It's still missing one 
page of the documentation, but I'm willing to accept that for now to 
unblock Python 3.10.

I mostly wasted my first rdeps testing run by accidentally using 
stable's source packages (causing most of them to fail).  I have started 
another run.

Of the autopkgtest failures in 
https://release.debian.org/britney/pseudo-excuses-experimental.html#pandas 
, python-feather-format and python-geotiepoints look like "not built for 
Python 3.10 yet", python-skbio looks like a real bug.

[toc] | [prev] | [next] | [standalone]

#1083398 — Bug#999415: transition: pandas 1.1 -> 1.3 - to unstable now or not?

After build-testing about half of the reverse dependencies, failures 
that look new-pandas-related are cfgrib #1000726, joypy #1000727, 
python-skbio #1000752, and maybe hyperspy (not filed yet).

python-skbio and hyperspy already FTBFS for unrelated reasons (but fail 
more tests with new pandas), and joypy looks trivially fixable.

Given this and expecting to find a similar number in the other half, 
against pandas 1.3 working on python3.10 while 1.1 doesn't (#1000422), 
would you prefer to have pandas 1.3 in unstable now, or not?

[toc] | [prev] | [next] | [standalone]

#1083430 — Bug#999415: transition: pandas 1.1 -> 1.3 - to unstable now or not?

On Sun, 28 Nov 2021, Rebecca N. Palmer wrote:

> After build-testing about half of the reverse dependencies, failures that 
> look new-pandas-related are cfgrib #1000726, joypy #1000727, python-skbio 
> #1000752, and maybe hyperspy (not filed yet).
>
> python-skbio and hyperspy already FTBFS for unrelated reasons (but fail more 
> tests with new pandas), and joypy looks trivially fixable.
>
> Given this and expecting to find a similar number in the other half, against 
> pandas 1.3 working on python3.10 while 1.1 doesn't (#1000422), would you 
> prefer to have pandas 1.3 in unstable now, or not?

My vote would be: go for it, but then again, I don't maintain any of 
pandas BDs.

Thanks,
Scott

[toc] | [prev] | [next] | [standalone]

#1083528

Testing the other half added only python-cooler (#1000814) and maybe 
rdkit (not tested for lack of time).

Also (see e.g. #1000812), if we want build-time tests run in Python 
3.10, pandas has to add Python 3.10 before its rdeps can be built.

Hence, I now intend to upload pandas 1.3 to unstable.

[toc] | [prev] | [next] | [standalone]

#1084866

The autopkgtest found 2 more, dask #1001348 (32 bit only) and 
python-anndata #1001349 (hidden in my test by an unrelated bug). 
python-skbio and python-cooler have been fixed, and I consider hyperspy 
to probably be a non-pandas-related bug (and it isn't in testing 
anyway).  I have now tested rdkit, which worked.

This leaves cfgrib, dask, joypy and python-anndata.  All except cfgrib 
have what I think are patches, though I haven't tested any of these.

[toc] | [prev] | [standalone]

Back to top | Article view | linux.debian.bugs.dist

csiph-web