search

hendersontrent / GAM.jl

Fit, evaluate, and visualise generalised additive models (GAMs) in native Julia
MIT License
29 stars 3 forks source link

Refactor for model formula and docs #42

Closed hendersontrent closed 2 months ago

hendersontrent commented 2 months ago

Closes #40

hendersontrent commented 2 months ago

@yahrMason GitHub won't let me add you as a reviewer, but please feel free to check out this PR and see if you like the interface. Hopefully it will make it a bit easier to play nice with non-smooth (i.e., categorical) predictors once we add them as well, however, the broad goal of the PR is to get it as close to mgcv::gam as possible.

yahrMason commented 2 months ago

@hendersontrent this looks really good. The only thing I would say is that I don't think we need to pass the dictionary into the function. We can let the family and link args be simple string inputs (like mgcv). Then in the function itself we can grab the specified family and link from their respective dictionaries.

hendersontrent commented 2 months ago

@yahrMason have made that edit in #44 which I just merged :)