converting mcplates to pymc3

[duserzym]

Hi Ian @ian-r-rose ,

Hope you are doing well and safe!

I am Yiming Zhang, a first year PhD student with Nick Swanson-Hysell at Berkeley EPS. I am currently working on a paleomagnetism compilation project on Keweenawan intrusive igneous rocks, similar to this https://doi.org/10.1130/B31944.1 where mcplates project played an important role. I am a newbie in Bayesian statistics, and I was trying to learn about your mcplates code through reproducing it using pymc3. I noticed that the customized fisher and watson logp functions in pymc2 can take multiple variables, but in pymc3 a logp function can only take in one variable, and this causes some difficulty in subsequent testings.

I am very interested in this project, and I am wondering if I may have your contact info for a discussion. I have a repo here (https://github.com/duserzym/2020_mcplates3) showing my current progress.

Sincerely,
Yiming