BayesMixR - an R interface to BayesMix

[TeoGiane]

This PR introduces bayesmixr, the R command line interface to bayesmix. The idea is the development of an R package which is almost identical to bayesmixpy, its Python counterpart.

As a demonstration, I reproduced the gaussian_mix_uni.Rmd notebook copy-pasting the .ipynb version.

Documentation is already available. There is still something to do, in the meantime, any suggestion is appreciated!

What to do:

• Type checks for function inputs
• Tests and CI

[TeoGiane]

OK, I guess the package is ready for review. Let me know what you think!

[mberaha]

I think this is a great addition! Thanks a lot @TeoGiane :)

[brunoguindani]

Apart from a few minor documentation chores missing, this is great! Thank you!

[TeoGiane]

I guess we are finally read to merge. Since last major review:

• We rely on testthat to provide unit tests. Once bayesmixr is installed, we can run tests via testthat::test_package("bayesmixr") in R
• CI is also available for bayesmixr. GitHub actions have been fixed and updated. Once merged, we only need to uncomment lines in resources/docker/test/Dockerfile (where we install the package) and in .github/workflows/test.yaml (where we run unit tests).
• Just like the python counterpart, run_mcmc can also return the whole MCMC chain. We interact with protobuf in R via the RProtoBuf package, an R wrapper for the C++ protobuf interface.
• Many small fixes in the code.

Enjoy!

[mberaha]

Definitely good to go!

[brunoguindani]

Awesome! Thank you! Just one very small observation below...