feature request: supply lme4 style model formula

[braza2]

Hi, I have a feature request where one could specify the fixed and random effects via lme4 style formulas. Cheers, braza2

[fabsig]

Thank you for the request.

I agree that this might be a convenient feature. However, I currently do not have the time to work on this. Here is demo code that showcases how one can do easily the same thing without formula notation: https://github.com/fabsig/GPBoost/blob/master/R-package/demo/compare_usage_lme4_gpboost.R

Contributions are welcome. Note that GPBoost is a C++ library with Python and R interfaces. This means that the interface from the Python package should be the same as from R. I.e., any changes in the R package, such as lme4 style formula support, must also be implemented equivalently in the Python package. Further, GPBoost can do more than just GLMMs (GPs, boosting, and combinations of the former) and one should also think how these features fit into a formula notation.