[vsni.co.uk] Contact us
Author Message
Post new topic

  ASReml  ~  Newbie problem: GLMM with poisson distribution in asreml-r

andrew
Posted: Wed Feb 21, 2018 3:09 am Reply with quote
Joined: 11 Aug 2011 Posts: 11
I have recently switched from coding asreml in a text editor to asreml-r. Mostly all good but I've hit a stumbling block when trying to fit a GLMM to count data. The error message is:

Invalid binomialproportion 22.000000/ 1.000000 in record xxx

The problem seems to be with the family=asreml.poisson() statement so I created a simple test data set and function:

Code:
X=c(10,9,8,7)
Z=c('a','a','b','b')
summary(asreml(X~1,random=~Z,family=asreml.poisson()))


It returns a similar error message:

Code:
Poisson; Log   Mu=exp(XB) V=Mu;
Note: The LogLik value is unsuitable for comparing GLM models
Invalid binomialproportion 10.000000/ 1.000000 in record 1
Invalid binomialproportion 9.000000/ 1.000000 in record 2
Invalid binomialproportion 8.000000/ 1.000000 in record 3
Invalid binomialproportion 7.000000/ 1.000000 in record 4
INVALID data values for GLMM
Exit status: 300001 - Program failure fitting GLM
Finished on: Wed Feb 21 10:49:55 2018


I don't see how I am signifying that the data in X are proportions. What am I doing wrong?!!
View user's profile Send private message
dave
Posted: Wed Mar 28, 2018 2:00 am Reply with quote
Joined: 13 Jul 2016 Posts: 2
Hi,

I can run it successfully, the code and result as fellows:

X=c(10,9,8,7)
Z=c('a','a','b','b')
summary(asreml(X~1,random=~Z,family=asreml.poisson()))$varcomp
# > summary(asreml(X~1,random=~Z,family=asreml.poisson()))$varcomp
# ASReml: Wed Mar 28 09:57:03 2018
#
# Poisson; Log Mu=exp(XB) V=Mu;
# Note: The LogLik value is unsuitable for comparing GLM models
# LogLik S2 DF wall cpu
# 1.8699 1.0000 3 09:57:03 0.0
# 1.8899 1.0000 3 09:57:03 0.0
# 2.1142 1.0000 3 09:57:03 0.0 (1 restrained)
# 2.2157 1.0000 3 09:57:03 0.0 (1 restrained)
# 2.2224 1.0000 3 09:57:03 0.0 (1 restrained)
# 2.2228 1.0000 3 09:57:03 0.0 (1 restrained)
# 2.2228 1.0000 3 09:57:03 0.0
# 2.2228 1.0000 3 09:57:03 0.0
# 2.2228 1.0000 3 09:57:03 0.0
#
# Deviance fron GLM fit: 0.59
# Variance heterogenity factor (Deviance/df): 0.20
# (assuming 3 degrees of freedom)
#
# Finished on: Wed Mar 28 09:57:03 2018
#
# LogLikelihood Converged
# gamma component std.error z.ratio constraint
# Z!Z.var 4.018431e-07 4.018431e-07 NA NA Boundary
# R!variance 1.000000e+00 1.000000e+00 NA NA Fixed


Cheers,
Dave
View user's profile Send private message
andrew
Posted: Fri Mar 30, 2018 1:31 pm Reply with quote
Joined: 11 Aug 2011 Posts: 11
Thanks, Dave.

I tried it again in R Studio and it still doesn't work. Same error messages.

There must be something wrong with the ASReml-R build that I'm using. I'll contact VSNi support directly.

Cheers, Andrew.
View user's profile Send private message

Display posts from previous:  

All times are GMT
Page 1 of 1
Post new topic

Jump to:  

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You can attach files in this forum
You can download files in this forum