Codestin Search App

Hannan7812 · 2026-02-26T06:16:42Z

PR summary

Why is this change necessary?
As requested in [ENH]: add errorbar.capthick and errorbar.elinewidth to mplstyle #31194, this change introduces two new parameters configurable via stylesheets.
What problem does it solve?
It is an enhancement requested in the above mentioned issue.
What is the reasoning for this implementation?
It allows the user to set errorbar.capthick and errorbar.elinewidth once in the stylesheet without having to set them again and again. Closes [ENH]: add errorbar.capthick and errorbar.elinewidth to mplstyle #31194

PR checklist

"closes #0000" is in the body of the PR description to link the related issue
new and changed code is tested (Note that only test/test_axes.py was run because that was the only file changed)
[N/A] Plotting related features are demonstrated in an example
[N/A] New Features and API Changes are noted with a directive and release note
Documentation complies with general and docstring guidelines

Additional Notes

All the files required to add new rcparams were changed as mentioned in DOC: Add sections to rcParams documentation #31201.
Minimal Example with changes tested:

import matplotlib.pyplot as plt
import matplotlib as mpl
import numpy as np

# Configure via rcParams (simulating stylesheet behavior)
plt.rcParams.update({
    "errorbar.capsize": 10,
    "errorbar.capthick": 10,
    "errorbar.elinewidth": 10,
})

x = np.arange(5)
y = x
yerr = 0.5

print(mpl.rcParams["errorbar.capthick"])
print(mpl.rcParams["errorbar.elinewidth"])

fig, ax = plt.subplots()
ax.errorbar(x, y, yerr=yerr)

plt.show()

Output:

Merges with an updated upstream into the current branch for fixing the issue

github-actions · 2026-02-26T06:16:52Z

Thank you for opening your first PR into Matplotlib!

If you have not heard from us in a week or so, please leave a new comment below and that should bring it to our attention. Most of our reviewers are volunteers and sometimes things fall through the cracks.

You can also join us on gitter for real-time discussion.

For details on testing, writing docs, and our review process, please see the developer guide.
Please let us know if (and how) you use AI, it will help us give you better feedback on your PR.

We strive to be a welcoming and open project. Please follow our Code of Conduct.

Copilot

Pull request overview

This PR adds two new rcParams (errorbar.capthick and errorbar.elinewidth) so that cap thickness and errorbar line width can be configured via Matplotlib stylesheets (requested in #31194).

Changes:

Register errorbar.capthick and errorbar.elinewidth as rcParams (validation, defaults, typing support).
Add the new keys to the default matplotlibrc template.
Update Axes.errorbar to source capthick / elinewidth from rcParams when not explicitly provided.

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 5 comments.

File	Description
lib/matplotlib/typing.py	Adds the new rcParam keys to typing key lists.
lib/matplotlib/rcsetup.py	Defines validators and `_Param` metadata for the new rcParams.
lib/matplotlib/mpl-data/matplotlibrc	Adds commented defaults for the new rcParams in the template rc file.
lib/matplotlib/axes/_axes.py	Attempts to apply rcParam-backed defaults for `capthick` and `elinewidth` in `Axes.errorbar`.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-02-26T06:24:34Z

        capsize = mpl._val_or_rc(capsize, "errorbar.capsize")
+        capthick = mpl._val_or_rc(capthick, "errorbar.capthick")
+        elinewidth = mpl._val_or_rc(elinewidth, "errorbar.elinewidth")
        if capsize > 0:
            eb_cap_style['markersize'] = 2. * capsize
        if capthick is not None:
            eb_cap_style['markeredgewidth'] = capthick
+        if eb_lines_style is not None:
+            eb_lines_style['linewidth'] = elinewidth


elinewidth is already handled earlier when building eb_lines_style; the new _val_or_rc block later unconditionally overwrites eb_lines_style['linewidth'] (even when the resolved rcParam is None). This can (a) clobber an explicitly provided linewidth kwarg and (b) set linewidth=None in the collection style, which may break downstream setters. Resolve elinewidth (including rc fallback) once, apply it only when non-None, and preserve the existing precedence where explicit elinewidth/linewidth overrides the rcParam default.

Copilot · 2026-02-26T06:24:34Z

+        capthick = mpl._val_or_rc(capthick, "errorbar.capthick")
+        elinewidth = mpl._val_or_rc(elinewidth, "errorbar.elinewidth")


New behavior (rcParams-backed defaults for capthick/elinewidth) is introduced here but there are no tests exercising rcParams['errorbar.capthick'] / rcParams['errorbar.elinewidth'] or their precedence vs explicit elinewidth/linewidth kwargs. Please add a test (likely in lib/matplotlib/tests/test_axes.py near existing errorbar tests) that sets these rcParams and asserts the resulting cap markeredgewidth and errorbar LineCollection linewidth.

Fixed documentation formatting Co-authored-by: Copilot <[email protected]>

timhoffm

Please add tests to show that the rcParams are correctly used.

Note: I don't know why the AI review was triggered. Did you do this? You may or may not look at what it said. Sometimes it's useful, sometimes it isn't. For me as a maintainer, it's too much noise and I don't review AI, so I haven't looked at it.

timhoffm · 2026-02-26T16:23:06Z

        eb_cap_style = {**base_style, 'linestyle': 'none'}
        capsize = mpl._val_or_rc(capsize, "errorbar.capsize")
+        capthick = mpl._val_or_rc(capthick, "errorbar.capthick")
+        elinewidth = mpl._val_or_rc(elinewidth, "errorbar.elinewidth")


This must be moved before l.4158

timhoffm · 2026-02-26T16:23:43Z

+        if eb_lines_style is not None:
+            eb_lines_style['linewidth'] = elinewidth


Don't touch eb_lines_style again here. It should be built consistently in l.4155ff

Hannan7812 · 2026-02-26T17:18:32Z

I've added a test which checks that rcParams are correctly used. I've also made the suggested changes to the other comments. Let me know if I overlooked anything @timhoffm.

PS, My apologies, I forgot to turn off automatic copilot reviews from my settings.

Incorporated suggestion to satisfy linting rule Co-authored-by: AlbertUnruh <[email protected]>

Incorporated suggestion to pass linting test W291 Co-authored-by: AlbertUnruh <[email protected]>

timhoffm · 2026-03-03T01:25:41Z

Thanks @Hannan7812 !

QuLogic · 2026-03-03T13:23:29Z

Note, there was no second review here...

I'm wondering why the rcParam is errorbar.elinewidth, instead of errorbar.linewidth. The e prefix is in the method arguments to separate the errorbar arguments from the regular line arguments, but the rcParam is already in the errorbar group, so it seems extraneous.

timhoffm · 2026-03-03T14:30:44Z

Given the combination of

this being a relative straight forward change
low review capacity (we leave a lot of PRs unreviewed or even waiting for a second review)
A review by Albert (My bad, I thought it was the Ablert doing all the numeric fixes and counted this as a half review - there are too many Alberts around now)

I empowered myself to move this forward after waiting for 3 days whether somebody wanted to comment.

It was a mistake due to mixing up the Alberts. I shouldn't have done it.

But overall we do have a review problem. In particular, we should talk about ways to get a second review if a first approval is there. PRs with one review have a decent quality and it's typically comparably low effort to do the second review. We shouldn't leave them on the table because. (1) They increase the amout of work in progess. (2) They are rather low haning fruits for improvment. (3) Not following up discourages contibutors.

On errorbar.elinewidth vs. errorbar.linewidth: One can argument both ways. I'm inclinded to go with the structural mapping rcParams["<method>.<parameter>"]. We have a number of parameters that follow this pattern.
It's also not much effort to spend the additional e and it makes it very obvious that one can it is for the elinewidth parameter. Whereas errorbar.linewidth would leave some room for misunderstanding.

Hannan7812 added 3 commits February 26, 2026 10:24

Added 2 new rcparams: errorbar.capthick and errorbar.elinewidth

0f88d5f

Merge branch 'main' into fix-issue-31194

60fa15f

Merges with an updated upstream into the current branch for fixing the issue

Modified _axes.py to properly render the rcparams

b6cd569

Copilot AI review requested due to automatic review settings February 26, 2026 06:16

github-actions Bot added first-contribution topic: rcparams and removed first-contribution labels Feb 26, 2026

Copilot started reviewing on behalf of Hannan7812 February 26, 2026 06:17 View session

Fixed minor formatting to pass linter tests

d5ddfdb

Copilot AI reviewed Feb 26, 2026

View reviewed changes

Update lib/matplotlib/mpl-data/matplotlibrc

060e315

Fixed documentation formatting Co-authored-by: Copilot <[email protected]>

timhoffm reviewed Feb 26, 2026

View reviewed changes

Hannan7812 and others added 3 commits February 26, 2026 22:10

Adressed reviewer comments and added a test

82660d6

DOC: Add sections to rcParams documentation

384dac5

Apply suggestion from @timhoffm

9e690f2

github-actions Bot added Documentation: build building the docs status: needs rebase labels Feb 26, 2026

Merge branch 'main' into fix-issue-31194

d5232fc

github-actions Bot removed Documentation: build building the docs status: needs rebase labels Feb 26, 2026

AlbertUnruh reviewed Feb 26, 2026

View reviewed changes

Comment thread lib/matplotlib/tests/test_axes.py

AlbertUnruh reviewed Feb 26, 2026

View reviewed changes

Comment thread lib/matplotlib/tests/test_axes.py Outdated

Hannan7812 and others added 2 commits February 27, 2026 09:18

Update lib/matplotlib/tests/test_axes.py

97f6a32

Incorporated suggestion to satisfy linting rule Co-authored-by: AlbertUnruh <[email protected]>

Update lib/matplotlib/tests/test_axes.py

0c95fba

Incorporated suggestion to pass linting test W291 Co-authored-by: AlbertUnruh <[email protected]>

timhoffm reviewed Feb 27, 2026

View reviewed changes

Comment thread lib/matplotlib/axes/_axes.py Outdated

Moved capthick to line 4174

58596fb

github-actions Bot added the Documentation: devdocs files in doc/devel label Feb 27, 2026

timhoffm reviewed Feb 27, 2026

View reviewed changes

Comment thread doc/devel/contribute.rst Outdated

Hannan7812 force-pushed the fix-issue-31194 branch from 2c6b1ce to 58596fb Compare February 27, 2026 13:53

github-actions Bot removed the Documentation: devdocs files in doc/devel label Feb 27, 2026

timhoffm approved these changes Feb 27, 2026

View reviewed changes

AlbertUnruh approved these changes Feb 28, 2026

View reviewed changes

timhoffm added this to the v3.11.0 milestone Mar 3, 2026

timhoffm merged commit e3d6dfc into matplotlib:main Mar 3, 2026
38 of 40 checks passed

		capthick = mpl._val_or_rc(capthick, "errorbar.capthick")
		elinewidth = mpl._val_or_rc(elinewidth, "errorbar.elinewidth")

		if eb_lines_style is not None:
		eb_lines_style['linewidth'] = elinewidth

Uh oh!

Conversation

Hannan7812 commented Feb 26, 2026

PR summary

PR checklist

Additional Notes

Uh oh!

github-actions Bot commented Feb 26, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI Feb 26, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 26, 2026

Choose a reason for hiding this comment

Uh oh!

timhoffm left a comment

Choose a reason for hiding this comment

Uh oh!

timhoffm Feb 26, 2026

Choose a reason for hiding this comment

Uh oh!

timhoffm Feb 26, 2026

Choose a reason for hiding this comment

Uh oh!

Hannan7812 commented Feb 26, 2026

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

timhoffm commented Mar 3, 2026

Uh oh!

QuLogic commented Mar 3, 2026

Uh oh!

timhoffm commented Mar 3, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants