Just want to let everyone know about a new integration to
R using the
causact package. The package does not let you write
numpyro code directly, rather let's users create a programmatic model specification using a DAG. Then, given the DAG,
causact converts it to
numpyro code and returns the results back in #rstats.
If interested, check out the package here:
Also, I wrote the
causact package and an accompanying textbook "A Business Analyst's Introduction to Business Analytics (2nd Edition): Coding in R For Fast, Easy, and Visual Bayesian Inference" to gently introduce Bayesian thinking and computational Bayesian inference to business school students.
See the book here: https://www.causact.com/
I welcome feedback and encouragement. Thanks!