GAMs are so flexible that supporting the build-in ModelArray.gam is a lot of work and probably doesn't do everything users want anyways. We're thinking ModelArray.wrap is more useful in general and we can have detailed examples of how to get extra hypothesis testing or other features out of the gam fits.
GAMs are so flexible that supporting the build-in ModelArray.gam is a lot of work and probably doesn't do everything users want anyways. We're thinking ModelArray.wrap is more useful in general and we can have detailed examples of how to get extra hypothesis testing or other features out of the gam fits.