[R] AIC for GAM models

[ritapatricio]

Dear all,

I am analysing growth data - response variable - using GAM and GAMM models,
and 4 covariates: mean size, mean capture year, growth interval, having
tumors vs. not
The models work fine, and fit the data well, however when I try to compare
models using AIC I cannot get an AIC value.

This is the code for the gam model:

>model1=gam(growth_rate~s(sampling_year)+s(growth_interval)+s(mean_SCL)+FP_status,family=quasi(link=identity),method="REML")
> AIC(model1)
[1] NA

or:
> model2=gam(growth_rate~s(sampling_year)+s(mean_SCL),family=quasi(link=identity),method="REML")
> AIC(model1,model2)
              df AIC
model1 13.722861  NA
model2  9.898634  NA

I have tried several models and AIC always comes as 'NA', can someone help?

many thanks.

rita patrício






--
View this message in context: http://r.789695.n4.nabble.com/AIC-for-GAM-models-tp4640972.html
Sent from the R help mailing list archive at Nabble.com.