[ENH]: data kwarg support for mplot3d #20912

[jyjoshi]

PR Summary

Fix #20912

Hi, I have started working on this issue. This will be my first open source contribution, so I will need some help and guidance. I have gone through the contributing guide and have setup the development environment and Matplotlib in editable mode on my system.

I am currently working on understanding the Matplotlib internals.

[github-actions]

Thank you for opening your first PR into Matplotlib!

If you have not heard from us in a while, please feel free to ping @matplotlib/developers or anyone who has commented on the PR. 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

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

[story645]

Hi, thanks for the commit but you accidentally included a whole bunch of files. Please remedy that by

1. stash any current changes https://git-scm.com/docs/git-stash
2. checkout master
3. create a new branch off master
4. add your changes to the new branch
5. just add and commit the files you changed
6. push.

It's fine if you open a new PR to do this as I figure starting over with a new branch will be easier than cleaning up this one.
If you need help, please ask on gitter, especially the incubator channel https://matplotlib.org/devdocs/devel/contributing.html#contributor-incubator

[jyjoshi]

Hi, I noticed that I might have added my virtual environment files to the commit, in a hurry. I will try to cleanup this branch first and in case that does not work out, I will create a new PR and close this one.
Thanks :)

[timhoffm]

You can force-push over this PR.

1. git checkout 20912-data-kwarg-support-mplot3d
2. cleanup your branch
3. git push origin 20912-data-kwarg-support-mplot3d --force

There are multiple ways to cleanup a branch, with more or less git magic.
If that's too complicated, a semi-manual way ist:

2.1) Backup the changes you still want.
2.2) git reset --hard master
2.3) Apply your changes again.

Anyway, the important message is: you can git push ... --force a branch for a PR to replace everything in there.

[jyjoshi]

Hi, I have added the @_preprocess_data decorator to the scatter method in mplot3d/axes3d.py and it seems to work properly.

Code:

import matplotlib.pyplot as plt
>>> plt.gcf().add_subplot(projection="3d").scatter("a", "b", "c", data={"a": [0], "b": [1], "c": [2]})
<mpl_toolkits.mplot3d.art3d.Path3DCollection object at 0x7f80872b9400>
>>> plt.show()

It gives the desired output.

I have a couple of queries.

1. What labels should be included by default in replace names and label_namer.
2. How to resolve the checks that are currently failing.

[timhoffm]

How to resolve the checks that are currently failing.

Place this before the **kwargs parameter entry in the docstring.

        data : indexable object, optional
            DATA_PARAMETER_PLACEHOLDER

We auto-generate the content for the description.

[jyjoshi]

Hi,
I have added the @_preprocess_data decorator to relevant methods in mplot3d.py. All the tests are passing except Codecov/project/tests. Is it safe to merge?

[story645]

Is it ok that this isn't tested?

[timhoffm]

I think that's okish not to have tests. preprocess_data() itself is tested extensively. So tests would only add integration testing between the decorator and the 3d plotting functions. We also don't explicitly test combininations of positional and keyword arguments.

Systematically adding new tests for data kwarg an all functions seems overkill. Maybe one could add one or two image comparison tests and/or change some existing tests to use data. But given this is a first time contribution, I want to keep the barrier low.

[story645]

Congrats on your first pr @jayjoshi112711! 🥳
We hope to see you around!

[jyjoshi]

Thanks, @story645 and @timhoffm for helping me with my first PR, looking forward to making more contributions :)
Cheers!

[QuLogic]

@meeseeksdev backport to v3.5.x