[R-sig-ME] cenexponential family in MCMCglmm

Jarrod Hadfield j.hadfield at ed.ac.uk
Fri Dec 3 11:31:41 CET 2010

Hi Achaz,

The minimum and maximum values the response could take is the  
response, essentially. As an example,

y<-rexp(1000,exp(rnorm(1000, 0.5, 1)))
dat<-data.frame(y=y, ymin=floor(y), ymax=ceiling(y))

# Interval censoring [0,1) [1,2) [2,3] and right censoring [3,Inf]

m1<-MCMCglmm(cbind(ymin, ymax)~1, family="cenexponential", data=dat)



Quoting Achaz von Hardenberg <achaz.hardenberg at gmail.com>:

> Dear all,
> I am trying to fit a MCMCglmm model with right censored exponential  
> data as a response where the minimum value the response can be is 0  
> and INF the maximum.
> However, I can not find in the documentation, nor on the web, any  
> example about how I should specify such a model.
> This is what is written in the tutorial:
> "For censored responses two data columns must be passed. The first  
> column should contain the minimum value the data could take and the  
> second column the maximum."
> But it is not clear to me how I should pass these two columns  
> together with the response variable in the model...
> Could anybody send me an example?
> Thanks a lot for your assistance!
> achaz
> Dr. Achaz von Hardenberg
> --------------------------------------------------------------------------------------------------------
> Centro Studi Fauna Alpina - Alpine Wildlife Research Centre
> Servizio Sanitario e della Ricerca Scientifica
> Parco Nazionale Gran Paradiso, Degioz, 11, 11010-Valsavarenche (Ao), Italy
> E-mail: achaz.hardenberg at pngp.it
> 	     fauna at pngp.it
> Skype: achazhardenberg
> --------------------------------------------------------------------------------------------------------
> 	[[alternative HTML version deleted]]
> _______________________________________________
> R-sig-mixed-models at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models

The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.

More information about the R-sig-mixed-models mailing list