[R-sig-eco] GAM with a Gamma error structure

[Gavin Simpson]

On Sun, 2008-11-16 at 17:58 -0500, Sally Roman wrote:
> Hi- I'm try to run a GAM with a Gamma distribution, but am getting an
> error message that there are non positive values allowed for the gamma
> family.  I do not have any negative values for my response or explanatory
> variables.
> 
> cod.fall.gam2.kept<-gam(y~s(b,bs="cr",k=4)+s(e,bs="cr",k=4)+f+s(g,bs="cr",k=4),data=cod.fall,family=Gamma)
> Error in eval(expr, envir, enclos) :
>   non-positive values not allowed for the gamma family

Non-negative is *not* non-positive. Do you have zero values in the
response? This would cause the error you are getting.

HTH

G

> 
> Thanks Sally
> 
> 
> Sally Roman
> Fisheries Technician
> University of Massachusetts, SMAST
> 200 Mill Rd., Fairhaven, MA 02719
> sroman at umassd.edu
> 508-910-6379
> 
> _______________________________________________
> R-sig-ecology mailing list
> R-sig-ecology at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-ecology