Why not turn this into a package?

[JosephLalli]

The statannot package is a fantastic, much needed contribution to the python data science community, but it doesn't seem to have the community around it that it needs to stay updated.

As it happens I was just about to propose my own multiple corrections addition as a pull request, but yours is far better documented/tested/organized.

I'm just learning how to contribute to open-source projects and properly use git, so my commit history is poor and my code is not as organized as I would like. However, I encountered several other bugs in statannot that I monkey-patched over the course of a year:

1. Perform work in axes coordinates instead of data coordinates to allow for consistent annotation with log axes
2. Add bootstrapping/paired bootstrapping statistical tests (I wrote these myself, but if you know of a well supported 'standard' module I'm happy to use that.
3. Handful of minor bugs (integer category names no longer raise errors, other things I'm forgetting).

Anyway, its 3am in my time zone right now, so I will refactor my code as a fork of this package over the weekend. I think you should consider making this more widely available! (Unless I'm getting open source etiquette totally wrong.)

[trevismd]

Hi @JosephLalli,
Thank you for the positive comments! I'm looking forward to going over your PR !

As to your question, I had just been waiting to see

• if statannotations got any attention and would keep developing, with the community
• if work would resume on statannot, since there is some activity even though the author/maintainer doesn't seem to respond anymore. I would have been happy to continue contributing to the original package same as you.

You are the first to post here, while several others kept raising issues and submitting PRs to statannot so I wanted to see how it played out. I know it is a vicious circle not to reach out as long as it is not known, but I didn't want to promote it too much while my contribution was still moderate in comparison with past work.
If this package does get traction and grow (and again, you are showing the first sign of that), we could surely increase its visibility!

[trevismd]

The package is now on PyPi too, including your contribution on reference coordinates.