These are wrapper that fits multple QGAM models using qgam::mqgam and
converts it to a mgamViz object using the getViz function.
It is essentially a shortcut.
mqgamV(form, data, qu, lsig = NULL, err = NULL, aQgam = list(), aViz = list())
Arguments
| form, data, qu, lsig, err | same arguments as in qgam::mqgam. |
|---|---|
| aQgam | list of further arguments to be passed to qgam::mqgam. |
| aViz | list of arguments to be passed to getViz. |
Value
An object of class "mgamViz" which can, for instance, be plotted using plot.mgamViz.
Examples
library(mgcViz) set.seed(2) ## simulate some data... dat <- gamSim(2,n=500,dist="normal",scale=0.25)$data#> Bivariate smoothing example# Fit GAM and get gamViz object b <- mqgamV(y~s(x) + s(z) + I(x*z), data = dat, qu = c(0.25, 0.5, 0.75), aQgam = list(argGam = list(select = TRUE)), aViz = list("nsim" = 0))#> Estimating learning rate. Each dot corresponds to a loss evaluation. #> qu = 0.5..........done #> qu = 0.25.......done #> qu = 0.75...........done# This is equivalent to doing # 1. Fit QGAM # b <- mqgam(y~s(x) + s(z) + I(x*z), data=dat, # qu = c(0.25, 0.5, 0.75), argGam = list(select = TRUE)) # 2. Convert to gamViz object # b <- getViz(b, nsim = 0) # Either way, we all effects by doing print(plot(b, allTerms = TRUE), pages = 1)
