Paper edits for JOSS

[theorashid]

Pull request type

Please check the type of change your PR introduces:

• Bugfix
• Feature
• Code style update (formatting, renaming)
• Refactoring (no functional changes, no api changes)
• Build related changes
• Documentation content changes
• Other (please describe): paper

What is the current behavior?

Issue Number: N/A

What is the new behavior?

Other information

Hey Thomas and Daniel,

This is for openjournals/joss-reviews#4455. There was a broken reference and a typo. Also, you mentioned you wanted the title changed to GPJax: A didactic Gaussian Process framework in Jax. Now is your chance to add it (although personally I think "didactic" only makes the title confusing).

A couple of other points, which I'm happy for an open discussion:

• Are there no GPs packages in R? How do R users get by?
• Is it worth mentioning that some PPLs have submodules for GPs which are commonly used, e.g pymc. Feel free to convince me these are really limited compared to the purpose-built libraries
• Is it worth mentioning the other jax-based GP packages: tinygp and bayesnewton. bayesnewton has also been used by the Aalto team for graph GPs (in jax). tinygp interests me because it integrates with numpyro, which I find much better than tfp.

Best,

Theo

[thomaspinder]

Hi @theorashid. I'm unsure if you can see my update to your PR, but I have responded to your comments and updated the paper accordingly here.

[theorashid]

Great. Are you able to merge these to master? Then I will remake the pdf using the bot and read through it again

[thomaspinder]

Great. Are you able to merge these to master? Then I will remake the pdf using the bot and read through it again

Done!

[theorashid]

It doesn't seem to have merged your updates

[thomaspinder]

My bad. Should be merged now.