[ADMB Users] glmmADMB-Function maximizer failed Poisson, ZIPoisson, NB1, ZINB1

[Ben Bolker]

On 11-12-14 07:29 PM, Rocio Jana wrote:
> Hello AMDB people
> 
> I am working on selecting the best model for my data comparing the
> AIC of 6 models of the same data, constructed with different family
> distributions and Zero and non-Zero inflation. I am having the
> Function Maximizer Failed error, and tried to fix it with BBolker's
> recommendation, didn't work. I am using R 2.14.0 and glmmADMB 0.6.7.1
> in Windows.
> 
> My response variable is mataicl (number of seeds of a particular
> species called matai), it is over dispersed and most-likely, zero
> inflated (ZI). The predictors are the dominance indexes
> (non-integers) for the canopy of 6 plant species that can be present
> above the seed traps that collect the seeds. I have data for 55 seed
> traps (trapno as grouping factor), in 12 years (year as random
> effects). Some of the seed traps have no data for some of the years
> (4 seed traps in each of 5 years). The data of the predictors is
> repeated across the years, as the seed traps and the trees are the
> same.
> 
> The model I am testing has the following syntax

> pel.mataicl.NBalt<-glmmadmb(mataicl ~ rimuover + kahikover +
> mataiover + miroover + hinauover +
> tawaover+(year|trapno),zeroInflation=FALSE,data=pel.3.dat,family =
> "nbinom1")
> 
> The distributions I am testing, and errors are: Poisson (transformed
> the values of the fixed factors to integers x*100 but still same
> error) function maximizer failed ZIPoisson function maximizer failed 
> NB (default)  NO PROBLEM ZINB (default) NO PROBLEM NBalt (type 1)
> function maximizer failed ZINBalt (type 1) function maximizer failed
> 
> The error is:
> 
> The function maximizer failed In addition: Warning messages: 1:
> running command 'C:\WINDOWS\system32\cmd.exe /c "C:/Program
> Files/R/R-2.14.0/library/glmmADMB/bin/windows32/glmmadmb.exe" -maxfn
> 500 -noinit -shess' had status 1 2: In shell(cmd, invisible = TRUE)
> : '"C:/Program
> Files/R/R-2.14.0/library/glmmADMB/bin/windows32/glmmadmb.exe" -maxfn
> 500 -noinit -shess' execution failed with error code 1
> 
> I know this has been asked in the forum before, and the answer was to
> modify the arguments. Well, I did it,
> 
> pel.mataicl.NBalt<-glmmadmb(mataicl ~ rimuover + kahikover +
> mataiover + miroover + hinauover +
> tawaover+(year|trapno),zeroInflation=FALSE,data=pel.3.dat,family =
> "nbinom1",admb.opts = admbControl(shess = FALSE, noinit = FALSE))
> 
> But still
> 
> The function maximizer failed In addition: Warning messages: 1:
> running command 'C:\WINDOWS\system32\cmd.exe /c "C:/Program
> Files/R/R-2.14.0/library/glmmADMB/bin/windows32/glmmadmb.exe" -maxfn
> 500' had status 1 2: In shell(cmd, invisible = TRUE) : '"C:/Program
> Files/R/R-2.14.0/library/glmmADMB/bin/windows32/glmmadmb.exe" -maxfn
> 500' execution failed with error code 1
> 
> Dave fournier suggested to run it from outside R. I am sorry but I
> don't know how to do that. And I don't know either how to include
> ./glmmadmb -shess Dave suggested.

  The -shess would be added in the most recent version by  using the
argument

admb.opts=admbControl(shess=FALSE)

  You might also try

admb.opts=admbControl(shess=FALSE,noinit=FALSE)


  I'm working on instructions for setting things up to try out the
optimizer outside of R.

  Ben Bolker

> 
> This is only for one of the response variable I am testing in the
> example, but I have seeds from the 6 species, so 6 groups of models
> (one response variable at the time).And for each group I am aiming to
> select the best distribution.
> 
> Sorry about the extremely long email.
> 
> Sincerely Rocio Jana
> 
> 
> 
> 
> 
> 
> Rocio C. Jaña Prado PhD candidate School of Biological Sciences 
> University of Canterbury Te Whare Wananga o Waitaha Private Bag 4800,
> Christchurch 8020 New Zealand Phone: +64 3 3643055 or ext 3055 Fax:
> +64 3 3642590
> 
> This email may be confidential and subject to legal privilege, it
> may not reflect the views of the University of Canterbury, and it is
> not guaranteed to be virus free. If you are not an intended
> recipient, please notify the sender immediately and erase all copies
> of the message and any attachments.
> 
> Please refer to http://www.canterbury.ac.nz/emaildisclaimer for more 
> information. _______________________________________________ Users
> mailing list Users at admb-project.org 
> http://lists.admb-project.org/mailman/listinfo/users