Feature request: Skyride tree prior

[richelbilderbeek]

Here is an email I got:

Inquiry about the possibility of including support for the Skyride tree prior for beautier

Dear Dr. Richèl Bilderbeek,

I am Jianhui Wang, a loyal user of your R package, beautier, which I find incredibly useful for preparing BEAST2 XML files.

I am reaching out to inquire about the possibility of including support for the Skyride tree prior. Currently, the package supports several tree priors such as Bayesian Skyline and Coalescent Exponential, but I would like to explore using Skyride for my analyses, as it is particularly well-suited for my dataset.

I greatly appreciate your efforts in replying my questions on GitHub and I hope beautier would be more widely used. Because I really like using R language instead of linux command line :)

Best regards,
Jianhui Wang

[richelbilderbeek]

Dear dr./prof. Wang,

Thanks for contacting me for this feature request!

I can add this tree prior, but only if (1) I get your help, (2) I get time, (3) It works with the current BEAST2 version

• 1. Will you volunteer in helping me do this?
  • If yes: I'll need to know which tests to write and you to check me. We'll meet. This will be fun!
  • If not,:we'll abandon this issue.
• 2. It will take some time, think weeks/months here!
  • If that is too slow, than I we'll abandon this issue
  • If that's OK, we go!
• 3. Currently, beautier uses version 2.6.0. If the Skyride tree prior does not work with this version of BEAST2, this will be too result in too much untested code. Does it run with BEAST2/beautier v.2.6.0?
  • If yes: let's go!
  • If no: I'll will not do it, unless there is funding for one full-time month of development. I don't want to cause (even more) headaches to my users

> beastier::get_beast2_version()
[1] "2.6.0"

I hope that help! In short: 'Yes, I'd be happy to help, but I need help' :-)

Cheers, Richel

[WJH58]

@richelbilderbeek
I would definitely like to help. However, as I stated in the previous question (ropensci/babette#108), I solved the problem by using beast v2.5. Is it possible that we get to this question after you solve that? Otherwise I could only test the new function with beast v2.5, instead of v2.6.
Thank you!

[richelbilderbeek]

Great! Yes, I could use some time to work on the earlier error, then add this feature for v2.5.0. I will work on it this Saturday and will report before midnight.

As that version has the least amount of tests, it may be harder for me, but I'll signal when it is like that; it may just as well be quite easy ...

[richelbilderbeek]

Waiting for user to respond to the issue that needed to be solved first.