first cut at showing multiple transfer functions on VR

[zingale]

PR Summary

Currently save_annotated() doesn't work correctly if we have multiple volume sources in the volume render. This is an attempt to fix that by looping over sources and showing all the transfer functions. With multiple transfer functions, we don't use the colorbar axis, but instead define a rectangle (tf_rect) in which to place the transfer functions. This is subdivided vertically to account for the number of render sources and the transfer functions are shown there. This gives some control over the location and sizing of the legends.

PR Checklist

• Code passes flake8 checker
• New features are documented, with docstrings and narrative docs
• Adds a test for any bugs fixed. Adds tests for new features.

[zingale]

Here's an example output:

[zingale]

updated output, with tf_rect manually changed to [0.8, 0.2, 0.08, 0.6]

[zingale]

I've added a cookbook receipe demonstrating this, using the standard IsolatedGalaxy dataset. The render appears as:

[zingale]

note also, in producing that example, I fixed a bug in the latex rendering -- the exponent was not places in {}

[zingale]

I just updated this with a fix for the case where not all the render sources have a transfer function (e.g., a volume source + a point source)

[zingale]

I've synced this up with master and verified that it still renders the two fields for the isolated galaxy correctly.

[zingale]

can I get some guidance on what people would like to see for this PR in order to get it merged?

[munkm]

Hi @zingale, I'm so sorry I dropped the ball on reviewing this. Would you mind merging or rebasing your PR with the current master branch so we can see how the tests look (and the conflicts are resolved)? I'm super excited about this feature and I think it will be a great addition to the project!

[neutrinoceros]

Wow it's been a while.
That said, the only 2 conflicts here are caused by auto-formatting. It should be trivial to fix.
@zingale do you still wish to get this merged ? As Madicken suggested, you'd need to update this branch somehow (rebase or merge), and then hopefully tests will stay green.

[zingale]

I'll sync this up today. Thanks for the poke

[neutrinoceros]

A couple more fixes are needed. Most of them should be automatable with pre-commit.
Typing issues (mypy), not so much, but I can help there if needed @zingale :)

[zingale]

okay, I merged and fixed some issues due to volume rendering API changes, and the example works again:

[zingale]

I don't know about the runtime question -- it takes a few minutes on my workstation. But this is not the only VR of IsolatedGalaxy, so it can't be that much more expensive than existing ones in the cookbook.

[neutrinoceros]

I don't know about the runtime question -- it takes a few minutes on my workstation. But this is not the only VR of IsolatedGalaxy, so it can't be that much more expensive than existing ones in the cookbook.

Fair enough

[zingale]

I think I've addressed all of the comments now

[neutrinoceros]

Indeed. So I think we'd need a second approval to merge this. I'll advertise it on slack. Thank you @zingale !

[neutrinoceros]

Looks like the docs build timed out. Let's hope that's not reproducible
@yt-fido test this please

[zingale]

I just noticed that I made the render big, so I'll reduce it to speed it up for the cookbook. I'll update in a sec.

[zingale]

okay, I made it reuse the existing render and made it smaller. At some point in the future, it would be good to separate the size of the rendering from the save_annotated image size, but I think that is a separate PR. I also fixed a bug from my merging that lost the ability to set the size of the transfer function labels. Now I can make them more appropriately sized.

[zingale]

thanks @matthewturk and @neutrinoceros !!