Thanks for your extended reply. Application of the splines seems to be very
plausible.
I have now added random effects to the GAMM-model:
random= list( objectID= ~1|doy )
(is that defined correct?)
But I am wondering how to include both time and space in one te()-function?
Or would it be better to not use the by= factor(region), but a special
spatial autocorrelation with the x and y-coordinates per objectID? Thus
something like:
correlation= corAR1( form= ~ objectX + objectY )
Thus resulting in the GAMM-model:
model.formula <- formula( tau ~ te( x1, doy, bs= c('cr','cc' ) ) +
... + te( x4, doy, bs= c('cr', 'cc' ) ) )
model <- gamm( formula= model.formula,
random= list( farmID=
~1|dayOfTheYear ),
correlation= corAR1( form= ~
farmX+farmY ),
control= ctrl,
na.action= na.omit )
Concerning the memory, yes this will be an issue. I have a 16 GB server
available with 6 processors. Maybe it would be wise to run 4 seperate
GAMM-models, i.e. with x1, x2, x3, and x4 seperated.
Thanks in advance.
Jeroen
--
View this message in context:
http://r.789695.n4.nabble.com/Multiple-interaction-terms-in-GAMM-model-tp4672297p4672577.html
Sent from the R help mailing list archive at Nabble.com.
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
