New clear() method for Radio and Check buttons

[raamana]

PR Summary

Continuing #10924 to clear up some merge/rebase issues and have a fresh start

Fixes #10922

cc @jklymak

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

[jklymak]

Ha, I was just trying to get it moving again - I'm not actually familiar enough with the Radio Button stuff to review it.

[raamana]

Hi @jklymak, it was ALL approved in #10924 (3 approving reviews etc) except for waiting for the tests!! I hope we can merge this, so I can forget about it, and feel proud about having contributed to mighty matplotlib :)

[jklymak]

Hmmm, well, it was approved, but the tests never passed, hence it never was merged. Its still not passing the tests so its still not going to be merged. People often approve PRs assuming that the tests will catch anything unforeseen. In this case, they appear to have, so you will have to pass that hurdle first.

[raamana]

I see. What do the tests in Azure pipeline test?
https://dev.azure.com/matplotlib/matplotlib/_build/results?buildId=1835&view=ms.vss-test-web.build-test-results-tab

what about this one from travis - no language set - are they not the same tests?
https://travis-ci.org/matplotlib/matplotlib/jobs/491327354

[raamana]

Also, all the errors seem to be with backed, which my code has NOTHING to do with - see https://travis-ci.org/matplotlib/matplotlib/jobs/491327353

=================================== FAILURES ===================================
___________________ lib/matplotlib/tests/test_backend_qt.py ____________________
[gw0] linux -- Python 3.7.1 /home/travis/virtualenv/python3.7.1/bin/python
Worker 'gw0' crashed while running 'lib/matplotlib/tests/test_backend_qt.py::test_fig_close[Qt5Agg]'
___________________ lib/matplotlib/tests/test_backend_qt.py ____________________
[gw1] linux -- Python 3.7.1 /home/travis/virtualenv/python3.7.1/bin/python
Worker 'gw1' crashed while running 'lib/matplotlib/tests/test_backend_qt.py::test_fig_signals'
___________________ lib/matplotlib/tests/test_backend_qt.py ____________________
[gw3] linux -- Python 3.7.1 /home/travis/virtualenv/python3.7.1/bin/python
Worker 'gw3' crashed while running 'lib/matplotlib/tests/test_backend_qt.py::test_correct_key[Qt5Agg-shift]'
___________________ lib/matplotlib/tests/test_backend_qt.py ____________________
[gw4] linux -- Python 3.7.1 /home/travis/virtualenv/python3.7.1/bin/python
Worker 'gw4' crashed while running 'lib/matplotlib/tests/test_backend_qt.py::test_correct_key[Qt5Agg-control]'
___________________ lib/matplotlib/tests/test_backend_qt.py ____________________
[gw2] linux -- Python 3.7.1 /home/travis/virtualenv/python3.7.1/bin/python
Worker 'gw2' crashed while running 'lib/matplotlib/tests/test_backend_qt.py::test_correct_key[Qt5Agg-lower]'
___________________ lib/matplotlib/tests/test_backend_qt.py ____________________
[gw5] linux -- Python 3.7.1 /home/travis/virtualenv/python3.7.1/bin/python
Worker 'gw5' crashed while running 'lib/matplotlib/tests/test_backend_qt.py::test_correct_key[Qt5Agg-unicode_lower]'
___________________ lib/matplotlib/tests/test_backend_qt.py ____________________
[gw7] linux -- Python 3.7.1 /home/travis/virtualenv/python3.7.1/bin/python
Worker 'gw7' crashed while running 'lib/matplotlib/tests/test_backend_qt.py::test_correct_key[Qt5Agg-control_alt]'
___________________ lib/matplotlib/tests/test_backend_qt.py ____________________
[gw8] linux -- Python 3.7.1 /home/travis/virtualenv/python3.7.1/bin/python
Worker 'gw8' crashed while running 'lib/matplotlib/tests/test_backend_qt.py::test_correct_key[Qt5Agg-unicode_upper]'
___________________ lib/matplotlib/tests/test_backend_qt.py ____________________
[gw9] linux -- Python 3.7.1 /home/travis/virtualenv/python3.7.1/bin/python
Worker 'gw9' crashed while running 'lib/matplotlib/tests/test_backend_qt.py::test_correct_key[Qt5Agg-backspace]'
_______________________ test_interactive_backend[qt5agg] _______________________
[gw6] linux -- Python 3.7.1 /home/travis/virtualenv/python3.7.1/bin/python
backend = 'qt5agg'

[raamana]

So when I run the tests locally, this time ensuring there is a DISPLAY via the VNC session, most of the test failures are gone. All the errors in the interactive backend are gone (esp the webagg, which I can't imagine why it would fail due to my code changes). What is left is one test in backend_pgf whatever that is!

test_backend_bases.py ........                                                                                                                                                                              [  7%]
test_backend_cairo.py s                                                                                                                                                                                     [  8%]
test_backend_nbagg.py .                                                                                                                                                                                     [  9%]
test_backend_pdf.py s............sssss.                                                                                                                                                                     [ 26%]
test_backend_pgf.py ssssssFss                                                                                                                                                                               [ 34%]
test_backend_ps.py .ss..s.s..sss                                                                                                                                                                            [ 46%]
test_backend_qt.py s..sssssssssss..............                                                                                                                                                             [ 72%]
test_backends_interactive.py ss.s.ss.                                                                                                                                                                       [ 79%]
test_backend_svg.py .s.ss.ss.ss                                                                                                                                                                             [ 89%]
test_backend_tools.py ...........  

[jklymak]

Looks like the qt5 errors are unrelated.

[raamana]

So the two errors remaining are to do with rcparams and LaTeX - see below. I think I will quit at this point and leave it to you guys to figure the errors out - I don't have anymore time for this right now unless it's something to do with code changes I made (or DIRECTLY related to it).

E           ValueError: The following params are missing in the matplotlibrc.template file: dict_items([('boxplot.flierprops.markeredgewidth', [1.0, <function validate_float at 0x7f24739522f0>]), ('scatter.edgecolors', ['face', <function validate_string at 0x7f2473952f28>])])

test_rcparams.py:464:

E           matplotlib.backends.backend_pgf.LatexError: LaTeX returned an error, probably missing font or error in preamble:
E           b'This is pdfTeXk, Version 3.141592-1.40.3 (Web2C 7.5.6)\n %&-line parsing enabled.\n**entering extended mode\n\n*! I can\'t find file `".ogut"\'.\n<to be read again> \n                   \\let \n<*> \\documentclass\n                  {minimal}\n(Press Enter to retry, or Control-D to exit)\nPlease type another input file name: ! I can\'t find file `".ogut"\'.\n<to be read again> \n                   \\let \n<*> \\documentclass\n                  {minimal}\n(Press Enter to retry, or Control-D to exit)\nPlease type another input file name: ! I can\'t find file `".ogut"\'.\n<to be read again> \n                   \\let \n<*> \\documentclass\n                  {minimal}\n(Press Enter to retry, or Control-D to exit)\nPlease type another input file name: ! I can\'t find file `\\begin{document}\'.\n<to be read again> \n                   \\let \n<*> \\documentclass\n                  {minimal}\n(Press Enter to retry, or Control-D to exit)\nPlease type another input file name: ! I can\'t find file `text\'.\n<to be read again> \n                   \\let \n<*> \\documentclass\n                  {minimal}\n(Press Enter to retry, or Control-D to exit)\nPlease type another input file name: ! I can\'t find file `\\typeout{pgf_backend_query_start}\'.\n<to be read again> \n                   \\let \n<*> \\documentclass\n                  {minimal}\n(Press Enter to retry, or Control-D to exit)\nPlease type another input file name: ! I can\'t find file `\\makeatletter\'.\n<to be read again> \n                   \\let \n<*> \\documentclass\n                  {minimal}\n(Press Enter to retry, or Control-D to exit)\nPlease type another input file name: ! I can\'t find file `\\@@end\'.\n<to be read again> \n                   \\let \n<*> \\documentclass\n                  {minimal}\n(Press Enter to retry, or Control-D to exit)\nPlease type another input file name: \n! Emergency stop.\n<to be read again> \n                   \\let \n<*> \\documentclass\n                  {minimal}\n!  ==> Fatal error occurred, no output PDF file produced!\nTranscript written on texput.log.\n'

../backends/backend_pgf.py:308: LatexError
---------------------------------------------------------------------------------------------- Captured stdout call -----------------------------------------------------------------------------------------------
This is pdfTeXk, Version 3.141592-1.40.3 (Web2C 7.5.6)
 %&-line parsing enabled.
**entering extended mode

*! I can't find file `".ogut"'.
<to be read again> 
                   \let 
<*> \documentclass
                  {minimal}
(Press Enter to retry, or Control-D to exit)
Please type another input file name: ! I can't find file `".ogut"'.
<to be read again> 
                   \let 
<*> \documentclass
                  {minimal}
(Press Enter to retry, or Control-D to exit)
Please type another input file name: ! I can't find file `".ogut"'.
<to be read again> 
                   \let 
<*> \documentclass
                  {minimal}
(Press Enter to retry, or Control-D to exit)
Please type another input file name: ! I can't find file `\begin{document}'.
<to be read again> 
                   \let 
<*> \documentclass
                  {minimal}
(Press Enter to retry, or Control-D to exit)
Please type another input file name: ! I can't find file `text'.
<to be read again> 
                   \let 
<*> \documentclass
                  {minimal}
(Press Enter to retry, or Control-D to exit)
Please type another input file name: ! I can't find file `\typeout{pgf_backend_query_start}'.
<to be read again> 
                   \let 
<*> \documentclass
                  {minimal}
(Press Enter to retry, or Control-D to exit)
Please type another input file name: ! I can't find file `\makeatletter'.
<to be read again> 
                   \let 
<*> \documentclass
                  {minimal}
(Press Enter to retry, or Control-D to exit)
Please type another input file name: ! I can't find file `\@@end'.
<to be read again> 
                   \let 
<*> \documentclass
                  {minimal}
(Press Enter to retry, or Control-D to exit)
Please type another input file name: 
! Emergency stop.
<to be read again> 
                   \let 
<*> \documentclass
                  {minimal}
!  ==> Fatal error occurred, no output PDF file produced!
Transcript written on texput.log.

[raamana]

Yes @jklymak, these errors seem to be unrelated to my work. What's the process for accepting PRs in a situation like this here?

[jklymak]

I'll try recycling...

[anntzer]

Some (real, this-PR-related) style issues: https://travis-ci.org/matplotlib/matplotlib/jobs/492794388.

[raamana]

thanks - fixed the two flake issues:

(py36) $ 13:28:22 SQuark-3 matplotlib >>  flake8 --statistics widgets.py
(py36) $ 13:28:35 SQuark-3 matplotlib >>

[raamana]

Lovely - great to see it all green here :)

[timhoffm]

Please follow numpydoc-style and PEP-257 docstring conventions.

[raamana]

Style comments are addressed.

[raamana]

Ok, I think I am done. This has been a great learning exercise trying to contribute to a complex OS library :). I totally severely underestimated what it takes :)

[raamana]

This also closes #10924

[timhoffm]

Sorry, still some changes needed. But you're getting there.

[raamana]

Thanks Tim for detailed review and comments.

[timhoffm]

Please make sure that the PR is ready for review to the best of your knowledge and then leave a comment.

Reviewing is a time consuming activity and reviewer time is better spend than finding and commenting on formal and formatting issues. I expect from a PR author to put reasonable care into the PR.

Sorry for being a bit harsh. It's not personal. You are a new contributor, so it's clear that the PR will take more effort, but we're dancing around this single PR for quite some time, and we have a backlog of 270 open PRs in the project.

[QuLogic]

I rebased, cleaned up the conflicts, and squashed to one commit. The clear method on RadioButtons does allow setting no buttons selected, which is something that's no longer allowed in __init__: 3c306a3#r106438132

I wonder if instead, we want to remember the initial active index set in __init__ and make clear reset to that?

[jklymak]

I think I agree that radio buttons should always have at least one element selected. If you need a "none of the above", then that should be another button choice. https://uxplanet.org/radio-buttons-ux-design-588e5c0a50dc

[jklymak]

I'll move to draft

[QuLogic]

OK, RadioButtons.clear now resets back to the original setup.

[timhoffm]

This needs a what's new entry.

[timhoffm]

I'm wondering whether this should be named reset().

Pro reset: This is the correct terminology. You cannot clear a radio button group.
Pro clear: Consistency with other widget. There are a number of widgets with clear methods, but none with reset.

[QuLogic]

Yes, it would probably make more sense, but I left it clear for consistency.

[timhoffm]

I’ll review later today, but this looks good.

[timhoffm]

Minor docstring improvements.