Remove (some) features deprecated in mpl2.2 #12245

anntzer · 2018-09-24T09:23:11Z

PR Summary

@dstansby gets to do mlab in #12165 :)

PR Checklist

Has Pytest style unit tests
Code is Flake 8 compliant
New features are documented, with examples if plot related
Documentation is sphinx and numpydoc compliant
Added an entry to doc/users/next_whats_new/ if major new feature (follow instructions in README.rst there)
Documented in doc/api/api_changes.rst if API changed in a backward-incompatible way

dstansby · 2018-09-25T10:47:37Z

doc/api/next_api_changes/2018-09-18-AL-removals.rst

+  ``GridSpecFromSubplotSpec.get_subplot_params``,
+- svgfont support (in :rc:`svg.fonttype`),
+- passing 'box-forced' to `axes.Axes.set_adjustable`,
+- support for the strings 'on'/'true'/'off'/'false' to mean True/False.


I think the methods this affects should be listed here

dstansby

Having just spend ages trying to work out how to replace deprecated code in another project, I think all the deprecations here should have clear alternatives in the api_changes file where appropriate; I'm happy to work on this and push to this branch @anntzer?

anntzer · 2018-09-27T11:42:38Z

Go for it.

dstansby · 2018-10-06T15:16:30Z

lib/matplotlib/axes/_axes.py

@@ -6255,8 +6255,7 @@ def table(self, **kwargs):
    def hist(self, x, bins=None, range=None, density=None, weights=None,
             cumulative=False, bottom=None, histtype='bar', align='mid',
             orientation='vertical', rwidth=None, log=False,
-             color=None, label=None, stacked=False, normed=None,
-             **kwargs):
+             color=None, label=None, stacked=False, **kwargs):


Will this now just silently suck up normed into **kwargs, and not do anything? If so I think this removal needs a bit more thought (ie. maybe 1 or 2 releases that error when normed is passed).

I unremoved the normed kwargs. Feel free to handle its removal as you see fit.

timhoffm · 2018-10-07T12:24:58Z

I took the liberty to push more API-change notes. Each change should have an alternative given or stated that there is no alternative.

I'm ok with all the changes, but of course, cannot approve formally myself anymore because I've now code in this.

anntzer · 2018-11-13T13:13:50Z

Rebased. @dstansby can you re-review?

Edit: Hmmmm sorry failing the tests now.

WeatherGod · 2018-11-13T15:23:53Z

lib/matplotlib/legend.py

-        The update parameter control which parameter of the legend changes
-        when dragged. If update is "loc", the *loc* parameter of the legend
-        is changed. If "bbox", the *bbox_to_anchor* parameter is changed.
-        """


Just a comment here... this docstring was never updated to note the deprecation, which means one would not know of the deprecation until they actually use the method. That seems sub-optimal.

But if someone is not actually using the method, how are they bothered by the deprecation and removal?

WeatherGod · 2018-11-13T15:29:12Z

doc/api/next_api_changes/2018-09-18-AL-removals.rst

+  ``backend_bases.FigureManagerBase.show_popup`` (this never did anything)
+- ``backend_wx.SubplotToolWx`` (no replacement)
+- ``backend_wx.Toolbar`` (use ``backend_wx.NavigationToolbar2Wx`` instead)
+- ``cbook.align_iterators`` (no replacment)


Perhaps say "use pandas instead" or something?

If you have a specific replacement in mind, please push it (or I can do it), but otherwise that seems a bit vague (at least I don't know what replacement you're thinking about).

WeatherGod · 2018-11-13T15:34:03Z

I am not particularly enthused by the number of "no replacement" we are putting in the documentation. I feel like we need to at least throw a line to users for similar functionality, or a hint towards how to implement the missing feature themselves.

In addition, for future deprecations, I think we might want to have this sort of guidance documented at the time of deprecation, not at the time of removal. That way, people can know how to update their code at a normal cadence, rather than later (not saying that will always happen, but one can hope!)

anntzer · 2018-11-13T15:38:46Z

a hint towards how to implement the missing feature themselves.

They can copy-paste the code from an old version of matplotlib.

anntzer · 2018-11-15T11:09:32Z

test failures should be fixed now.

Don't have time to look at this, but my earlier concerns have been fixed

anntzer · 2019-01-01T15:38:49Z

rebased

dstansby · 2019-01-04T13:02:17Z

Looks good to me, I just added a few words to the note about svg support to make it clearer 👍

timhoffm · 2019-01-05T21:39:47Z

Needs rebase to unbreak the doc builds (c.f. #13080).

anntzer · 2019-01-05T21:41:19Z

done

timhoffm

I'm bold and approve even though I pushed a commit. But that commit only documented some alternatives and added removed one now-unnecessary argument case.

dstansby · 2019-01-07T11:26:25Z

Going to merge on the basis of two reviews and it's probably better to get this in sooner rather than later so people using the master branch can find any problems.

anntzer force-pushed the kill22 branch from aef9fd4 to ad1d14e Compare September 24, 2018 10:14

dstansby added this to the v3.1 milestone Sep 24, 2018

dstansby reviewed Sep 25, 2018

View reviewed changes

anntzer force-pushed the kill22 branch from ad1d14e to 818bc89 Compare September 25, 2018 11:54

dstansby requested changes Sep 27, 2018

View reviewed changes

dstansby self-assigned this Sep 27, 2018

dstansby previously requested changes Oct 6, 2018

View reviewed changes

dstansby force-pushed the kill22 branch from 9205fad to 877d777 Compare October 6, 2018 15:27

anntzer force-pushed the kill22 branch from 5886d14 to 6ca5ce6 Compare November 13, 2018 13:13

WeatherGod reviewed Nov 13, 2018

View reviewed changes

anntzer force-pushed the kill22 branch from 6ca5ce6 to 51bbffc Compare November 15, 2018 11:09

dstansby added status: needs rebase Release critical For bugs that make the library unusable (segfaults, incorrect plots, etc) and major regressions. labels Jan 1, 2019

anntzer force-pushed the kill22 branch from 51bbffc to e2c7b34 Compare January 1, 2019 15:38

anntzer added Release critical For bugs that make the library unusable (segfaults, incorrect plots, etc) and major regressions. and removed Release critical For bugs that make the library unusable (segfaults, incorrect plots, etc) and major regressions. status: needs rebase labels Jan 1, 2019

anntzer force-pushed the kill22 branch from e2c7b34 to 8f95256 Compare January 1, 2019 17:37

dstansby approved these changes Jan 4, 2019

View reviewed changes

timhoffm added the status: needs rebase label Jan 5, 2019

anntzer and others added 6 commits January 5, 2019 22:41

Remove (most) APIs deprecated in mpl2.2.

04aea74

Kill _string_to_bool.

08a2502

Update API changes

5258291

Unremove the normed kwarg to hist().

4271616

More alternatives in API changes

8b99e41

Make svg note clearer

893a0a3

anntzer force-pushed the kill22 branch from 5bc04d1 to 893a0a3 Compare January 5, 2019 21:41

anntzer removed the status: needs rebase label Jan 5, 2019

timhoffm approved these changes Jan 5, 2019

View reviewed changes

dstansby approved these changes Jan 7, 2019

View reviewed changes

dstansby merged commit 4614a26 into matplotlib:master Jan 7, 2019

anntzer deleted the kill22 branch January 7, 2019 12:47

jklymak mentioned this pull request Jan 8, 2019

ENH: add a user-friendly verbose interface #13129

Closed

6 tasks

anntzer mentioned this pull request Aug 13, 2019

Remove mention of now-removed feature in docstring. #15055

Merged

6 tasks

QuLogic mentioned this pull request Apr 28, 2023

Deprecate svgfont support, possibly replacing with woff? #10063

Closed

Uh oh!

Remove (some) features deprecated in mpl2.2 #12245

Remove (some) features deprecated in mpl2.2 #12245

Uh oh!

Conversation

anntzer commented Sep 24, 2018

PR Summary

PR Checklist

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

dstansby left a comment

Choose a reason for hiding this comment

Uh oh!

anntzer commented Sep 27, 2018

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

timhoffm commented Oct 7, 2018

Uh oh!

anntzer commented Nov 13, 2018 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

WeatherGod commented Nov 13, 2018

Uh oh!

anntzer commented Nov 13, 2018

Uh oh!

anntzer commented Nov 15, 2018

Uh oh!

anntzer commented Jan 1, 2019

Uh oh!

dstansby commented Jan 4, 2019

Uh oh!

timhoffm commented Jan 5, 2019

Uh oh!

anntzer commented Jan 5, 2019

Uh oh!

timhoffm left a comment

Choose a reason for hiding this comment

Uh oh!

dstansby commented Jan 7, 2019

Uh oh!

Uh oh!

anntzer commented Nov 13, 2018 •

edited

Loading