Use buildmer to fit big generalized additive models using bam from package mgcv
Use buildmer to fit big generalized additive models using bam from package mgcv
buildbam( formula, data = NULL, family = gaussian(), buildmerControl = buildmerControl(), ... )
formula |
See the general documentation under |
data |
See the general documentation under |
family |
See the general documentation under |
buildmerControl |
Control arguments for buildmer — see the general documentation under |
... |
Additional options to be passed to |
To work around an issue in bam(), you must make sure that your data do not contain a variable named 'intercept'.
lme4 random effects are supported: they will be automatically converted using re2mgcv.
As bam uses PQL, only crit='deviance' is supported for non-Gaussian errors.
library(buildmer)
model <- buildbam(f1 ~ s(timepoint,by=following) + s(participant,by=following,bs='re') +
s(participant,timepoint,by=following,bs='fs'),data=vowels)Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.