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

Groups > linux.debian.bugs.dist > #937856 > unrolled thread

Bug#918819: Backport fix for DeprecationWarnings with Numpy 1.16

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

Contents

  Bug#918819: Backport fix for DeprecationWarnings with Numpy 1.16 "E. Madison Bray" <erik.m.bray@gmail.com> - 2019-01-09 16:30 +0100
    Bug#918819: Backport fix for DeprecationWarnings with Numpy 1.16 Sandro Tosi <morph@debian.org> - 2019-01-09 16:40 +0100
      Bug#918819: Backport fix for DeprecationWarnings with Numpy 1.16 "E. Madison Bray" <erik.m.bray@gmail.com> - 2019-01-09 16:50 +0100
        Bug#918819: Backport fix for DeprecationWarnings with Numpy 1.16 Sandro Tosi <morph@debian.org> - 2019-01-09 17:10 +0100
          Bug#918819: Backport fix for DeprecationWarnings with Numpy 1.16 "E. Madison Bray" <erik.m.bray@gmail.com> - 2019-01-09 17:20 +0100

#937856 — Bug#918819: Backport fix for DeprecationWarnings with Numpy 1.16

Package: matplotlib2
Version: 2.2.3-5
Severity: normal

Dear Maintainer,

There is a known incompatibility between the current release of
matplotlib 2.2.3 and Numpy 1.16, which results in some
DeprecationWarnings.  In turn, this issue is affecting tests in sagemath
in experimental, which use matplotlib, causing them to fail due to the
unexpected DeprecationWarnings.

The issue is already fixed upstream and backported for matplotlib
v2.2.4, but it would be nice to have this patch soon for reducing test
failures in sagemath:

https://github.com/matplotlib/matplotlib/pull/12508

Thanks in advance!

-- System Information:
Debian Release: buster/sid
  APT prefers unstable
  APT policy: (500, 'unstable'), (1, 'experimental')
Architecture: amd64 (x86_64)

Kernel: Linux 3.13.0-123-generic (SMP w/16 CPU cores)
Locale: LANG=C, LC_CTYPE=C.UTF-8 (charmap=UTF-8), LANGUAGE=C (charmap=UTF-8)
Shell: /bin/sh linked to /usr/bin/dash
Init: unable to detect

[toc] | [next] | [standalone]

#937858

On Wed, Jan 9, 2019 at 10:24 AM E. Madison Bray <erik.m.bray@gmail.com> wrote:
>
> Package: matplotlib2
> Version: 2.2.3-5
> Severity: normal
>
> Dear Maintainer,
>
> There is a known incompatibility between the current release of
> matplotlib 2.2.3 and Numpy 1.16, which results in some
> DeprecationWarnings.  In turn, this issue is affecting tests in sagemath
> in experimental, which use matplotlib, causing them to fail due to the
> unexpected DeprecationWarnings.

DeprecationWarning are, by definition, warnings. if your code/your
package code fails because of them, you need to relax whatever check
is been executed. warnings are there to let you know things will
change in the future, they are NOT an indication of anything wrong.

it looks to be a common practice in the "python science" packages to
threat DepWarnings as errors, this is not the correct usage.

> The issue is already fixed upstream and backported for matplotlib
> v2.2.4, but it would be nice to have this patch soon for reducing test
> failures in sagemath:
>
> https://github.com/matplotlib/matplotlib/pull/12508

that being said, i'll have a look at it

-- 
Sandro "morph" Tosi
My website: http://sandrotosi.me/
Me at Debian: http://wiki.debian.org/SandroTosi
G+: https://plus.google.com/u/0/+SandroTosi

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

#937868

[Multipart message — attachments visible in raw view] — view raw

On Wed, Jan 9, 2019, 16:32 Sandro Tosi <morph@debian.org wrote:

> On Wed, Jan 9, 2019 at 10:24 AM E. Madison Bray <erik.m.bray@gmail.com>
> wrote:
> >
> > Package: matplotlib2
> > Version: 2.2.3-5
> > Severity: normal
> >
> > Dear Maintainer,
> >
> > There is a known incompatibility between the current release of
> > matplotlib 2.2.3 and Numpy 1.16, which results in some
> > DeprecationWarnings.  In turn, this issue is affecting tests in sagemath
> > in experimental, which use matplotlib, causing them to fail due to the
> > unexpected DeprecationWarnings.
>
> DeprecationWarning are, by definition, warnings. if your code/your
> package code fails because of them, you need to relax whatever check
> is been executed. warnings are there to let you know things will
> change in the future, they are NOT an indication of anything wrong.
>
> it looks to be a common practice in the "python science" packages to
> threat DepWarnings as errors, this is not the correct usage.
>

That's not the issue here. It's not that it's considered an error. Rather,
the problem for sagemath in particular is that it has an extensive suite of
doctests, and so unexpected deprecation warnings results in many failing
doctests.

The alternative is to silence the DeprecationWarning but that is also not
desirable as we do want to know about such issues when they occur.

Thanks.

>

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

#937873

> That's not the issue here. It's not that it's considered an error. Rather, the problem for sagemath in particular is that it has an extensive suite of doctests, and so unexpected deprecation warnings results in many failing doctests.

doctest shouldnt fail when DeprecationWarnings are generated. are you
sure sagemath doesnt treat them as errors or trying to do something
fancy with them?

-- 
Sandro "morph" Tosi
My website: http://sandrotosi.me/
Me at Debian: http://wiki.debian.org/SandroTosi
G+: https://plus.google.com/u/0/+SandroTosi

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

#937878

On Wed, Jan 9, 2019 at 5:05 PM Sandro Tosi <morph@debian.org> wrote:
>
> > That's not the issue here. It's not that it's considered an error. Rather, the problem for sagemath in particular is that it has an extensive suite of doctests, and so unexpected deprecation warnings results in many failing doctests.
>
> doctest shouldnt fail when DeprecationWarnings are generated. are you
> sure sagemath doesnt treat them as errors or trying to do something
> fancy with them?

Yes, I'm sure.  This is actually perfectly normal and expected
behavior with Sage's doctest runner and has been forever (it also
tests *expected* warnings, for example).

You can see examples of this like:
https://buildd.debian.org/status/fetch.php?pkg=sagemath&arch=amd64&ver=8.6~rc0-1&stamp=1546789550&raw=1

(grep for `np.asscalar` or something)

This issue is responsible for the majority of noise in current builds :)

[toc] | [prev] | [standalone]

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

csiph-web