Hi, dear all. when running the model "mymodel<-glmmadmb(response~var1+var2+var3,random=~1|var3,family="nbinom",data=mydata)" I'm getting the successives following error messages: step1: error message: cannot change the working directory then, I've set save.dir to "tmp step2: I run again mymodel with the argument save.dir set to "tmp", I get the error message "Couldn't find STD file ...... maxfn 500 maxph 5 noinit had statut1 then, I navigate to admbControl page and rewrite mymodel as mymodel<-glmmadmb(response~var1+var2+var3,random=~1|var3,family="nbinom",data=mydata, admb.opts=admbControl(impSamp=0,maxfn=500,imaxfn=500,maxph=500,noinit=FALSE,shess=FALSE),data=transmpics_ext,save.dir="tmp") after this, I still getting STD file not found message error like " Error in glmmadmb(response~var1+var2+var3,random=~1| : The function maximizer failed (couldn't find STD file) Troubleshooting steps include (1) run with 'save.dir' set and inspect output files; (2) change run parameters: see '?admbControl' In addition: Warning message: running command 'C:\Windows\system32\cmd.exe /c "C:/Users/Coliasso/Documents/R/win-library/3.4/glmmADMB/bin/windows64/glmmadmb.exe" -maxfn 500 -maxph 500' had status 1 " Additional informations 1) I'm using R version 3.4.3 2) I've installed glmmADMB package from a Package archive File on PC In advance, many thanks for your helps. Kind regards,
Problems-running-glmmadmb
3 messages · C. AMAL D. GLELE, Ben Bolker, Amal Dahounto
It's hard to know without a reproducible example: the most likely situation is that your data are a little more complex than your model can handle, and that the optimizer is failing (or, more mildly, it's finding a solution but the variance-covariance matrix is not positive definite, meaning it can't reliably compute standard errors for the parameters). Your two choices: (1) provide a lot more detail about your problem and see if you can find someone with the time and energy to reproduce the problem and see what's going wrong (2) try using the glmmTMB package instead; it's faster, more stable, and more or less does all the things that glmmADMB does. For the model you have written above, glmer.nb from the lme4 package should also work ...
On Wed, Jul 25, 2018 at 1:18 PM C. AMAL D. GLELE <altessedac2 at gmail.com> wrote:
Hi, dear all.
when running the model
"mymodel<-glmmadmb(response~var1+var2+var3,random=~1|var3,family="nbinom",data=mydata)"
I'm getting the successives following error messages:
step1:
error message: cannot change the working directory
then, I've set save.dir to "tmp
step2:
I run again mymodel with the argument save.dir set to "tmp", I get the
error message
"Couldn't find STD file
...... maxfn 500 maxph 5 noinit had statut1
then, I navigate to admbControl page and rewrite mymodel as
mymodel<-glmmadmb(response~var1+var2+var3,random=~1|var3,family="nbinom",data=mydata,
admb.opts=admbControl(impSamp=0,maxfn=500,imaxfn=500,maxph=500,noinit=FALSE,shess=FALSE),data=transmpics_ext,save.dir="tmp")
after this, I still getting STD file not found message error like
"
Error in glmmadmb(response~var1+var2+var3,random=~1| :
The function maximizer failed (couldn't find STD file) Troubleshooting
steps include (1) run with 'save.dir' set and inspect output files; (2)
change run parameters:
see '?admbControl'
In addition: Warning message:
running command 'C:\Windows\system32\cmd.exe /c
"C:/Users/Coliasso/Documents/R/win-library/3.4/glmmADMB/bin/windows64/glmmadmb.exe"
-maxfn 500 -maxph 500' had status 1
"
Additional informations
1) I'm using R version 3.4.3
2) I've installed glmmADMB package from a Package archive File on PC
In advance, many thanks for your helps.
Kind regards,
[[alternative HTML version deleted]]
_______________________________________________ R-sig-mixed-models at r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
Dear Ben, many thanks for your reply. Kind regards, 2018-07-25 19:41 GMT+02:00 Ben Bolker <bbolker at gmail.com>:
It's hard to know without a reproducible example: the most likely situation is that your data are a little more complex than your model can handle, and that the optimizer is failing (or, more mildly, it's finding a solution but the variance-covariance matrix is not positive definite, meaning it can't reliably compute standard errors for the parameters). Your two choices: (1) provide a lot more detail about your problem and see if you can find someone with the time and energy to reproduce the problem and see what's going wrong (2) try using the glmmTMB package instead; it's faster, more stable, and more or less does all the things that glmmADMB does. For the model you have written above, glmer.nb from the lme4 package should also work ... On Wed, Jul 25, 2018 at 1:18 PM C. AMAL D. GLELE <altessedac2 at gmail.com> wrote:
Hi, dear all. when running the model "mymodel<-glmmadmb(response~var1+var2+var3,random=~1|var3,
family="nbinom",data=mydata)"
I'm getting the successives following error messages: step1: error message: cannot change the working directory then, I've set save.dir to "tmp step2: I run again mymodel with the argument save.dir set to "tmp", I get the error message "Couldn't find STD file ...... maxfn 500 maxph 5 noinit had statut1 then, I navigate to admbControl page and rewrite mymodel as mymodel<-glmmadmb(response~var1+var2+var3,random=~1|var3,
family="nbinom",data=mydata,
admb.opts=admbControl(impSamp=0,maxfn=500,imaxfn=500,maxph=
500,noinit=FALSE,shess=FALSE),data=transmpics_ext,save.dir="tmp")
after this, I still getting STD file not found message error like " Error in glmmadmb(response~var1+var2+var3,random=~1| : The function maximizer failed (couldn't find STD file) Troubleshooting steps include (1) run with 'save.dir' set and inspect output files; (2) change run parameters: see '?admbControl' In addition: Warning message: running command 'C:\Windows\system32\cmd.exe /c "C:/Users/Coliasso/Documents/R/win-library/3.4/glmmADMB/
bin/windows64/glmmadmb.exe"
-maxfn 500 -maxph 500' had status 1
"
Additional informations
1) I'm using R version 3.4.3
2) I've installed glmmADMB package from a Package archive File on PC
In advance, many thanks for your helps.
Kind regards,
[[alternative HTML version deleted]]
_______________________________________________ R-sig-mixed-models at r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
Amal DAHOUNTO GLELE C. *Statistician, Data ManagerElectronicDataCapture(EDC) Systems Manager*Institut de Recherche pour le D?veloppement (IRD) MIVEGEC (UMR IRD 224 - CNRS 5290) IRD-CREC (Cotonou, BENIN) IRD-IRSS (Bobo, BURKINA FASO) *T?l:(+229) 97 33 83 44 ; (+226) 60 96 23 74* *Skype: coliasso* [[alternative HTML version deleted]]