GeezerVison

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.

I would like some advice regarding implementing tests for the added function val_in_range and formatting the docstrings and documentations. Thank you.

I've made the changes which I think should fix the type ambiguity I caused earlier. I will need a little time to implement the tests as I have never written tests before and will need to study the current tests before I can implement mine and fix all the linting error with it as well.

While working on this I noticed a couple of behaviors that I wanted to clarify.

For LinearScale, calling limit_range_for_scale with NaN values returns NaN unchanged:

Example code:

from matplotlib.scale import LinearScale

s = LinearScale(axis=None)

vmin, vmax = s.limit_range_for_scale(float('nan'), float('nan'), minpos=1e-100)
print(vmin, vmax)

which results in nan nan. This seems to indicate that limit_range_for_scale does not attempt to reject or normalize NaN for the linear scale. Is this intentional?

Similarly for LogScale, limit_range_for_scale allows +inf to pass through unchanged. As far as I know, the logarithm is defined for 0 < x < +inf, but +inf itself is not part of the domain. I just needed a small clarification on whether +inf will be allowed in val_in_range or not.

This is a difficult question. Without having looked into this, I suspect, one of the two is correct:

• it's unintentional, but does not matter for the current use cases limit_range_for_scale
• it's intentional, because we want non-finite values to pass through for the current use cases

Which one it is needs to be found out by looking at the current usage contexts. In either case, the behavior should be specified.

Thinking about it, the primary use for val_in_range is to determine whether a value is "compatible" with the scale, e.g. whether i can draw it (under the boundary condition of properly chosen limits.
In that sense, NaN and Inf values are not compatible. I therefore propose to return False for these cases. For example LinearScale would return math.isfinite(val).

I have added the finite bound check for the ScaleBase as well as LinearScale, LogScale and LogitScale and also edited the tests to match the new requirements.

Thanks @null-dreams! Congratulations on your first PR to Matplotlib 🎉 We hope to hear from you again.