[R-sig-ME] glmmADMB zero-inflated poisson model. How to solve
Thierry.ONKELINX at inbo.be
Mon May 5 20:58:08 CEST 2014
You have 29 non-zero observations. Assuming that all variables are continuous (or discrete with only two levels), your full model estimates 9 parameters for the poisson part (and 1 for the zero inflated part). 29 observations for 9 parameters is just not enough. As a rule of thumb, you need about 10 observations per parameter.
This leaves you with 3 options: 1) gather more non-zero data (aim for at least 300 more non-zero observations). 2) rethink your full model so that is only uses 3 parameters. Note that you need 2 parameters for your random effects. Hence one (1!) left for the fixed effects. 3) Put this dataset in the garbage bin and start a new experiment from scratch. Consult a local statistician while designing your experiment.
I'm afraid the quotes in my signature apply to your problem.
ir. Thierry Onkelinx
Instituut voor natuur- en bosonderzoek / Research Institute for Nature and Forest
team Biometrie & Kwaliteitszorg / team Biometrics & Quality Assurance
+ 32 2 525 02 51
+ 32 54 43 61 85
Thierry.Onkelinx op inbo.be
To call in the statistician after the experiment is done may be no more than asking him to perform a post-mortem examination: he may be able to say what the experiment died of.
~ Sir Ronald Aylmer Fisher
The plural of anecdote is not data.
~ Roger Brinner
The combination of some data and an aching desire for an answer does not ensure that a reasonable answer can be extracted from a given body of data.
~ John Tukey
Van: r-sig-mixed-models-bounces op r-project.org [mailto:r-sig-mixed-models-bounces op r-project.org] Namens laura laura
Verzonden: maandag 5 mei 2014 18:20
Aan: r-sig-mixed-models op r-project.org
Onderwerp: [R-sig-ME] glmmADMB zero-inflated poisson model. How to solve
I'm trying to run a ZIGLMM with the package glmmADMB. My data set contains 37 zero-valued of 66 observations in the dependent variable (number of attaks), so I consider that a zero-inflated model would be appropriated. I checked the dataset and, apparently, there are enough variation for all combinations.
I centered and rescaled continuos input variables before performing the model. I run several models and they looked OK, but an error message appeared when trying some combinations, usually with only one explanatory variable (but not in all cases!!) and with the null model (only random effects).
This is the full model:
model_full<-glmmadmb (Nat~ coldecoy+ colmale*colfem+sfecha+ snnd+ colvecino+ (1|Year) +(1|Nido),data=data, zeroInflation=TRUE,family="poisson")
and this is the error message:
model_7<-glmmadmb (Nat~ colfem+ (1|Year) +(1|Nido),data=data, zeroInflation=TRUE,family="poisson")
Parameters were estimated, but not standard errors were not: the most likely problem is that the curvature at MLE was zero or negative Error en glmmadmb(Nat ~ colfem + (1 | Year) + (1 | Nido), data = data, :
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'
Además: Mensajes de aviso perdidos
comando ejecutado './glmmadmb -maxfn 500 -maxph 5 -noinit -shess' tiene estatus 1
I found similar an almost identical question in the list (sorry, I was unable to find it again and I cannot give more details). I checked the ?admbControl and further explanations about this provided in a previous post, but it was unclear to me what should I do. Sorry if I'm repeating an already "solved" topic, but I was unable to find an answer. I would be very grateful if someone can give me some help.
[[alternative HTML version deleted]]
* * * * * * * * * * * * * D I S C L A I M E R * * * * * * * * * * * * *
Dit bericht en eventuele bijlagen geven enkel de visie van de schrijver weer en binden het INBO onder geen enkel beding, zolang dit bericht niet bevestigd is door een geldig ondertekend document.
The views expressed in this message and any annex are purely those of the writer and may not be regarded as stating an official position of INBO, as long as the message is not confirmed by a duly signed document.
More information about the R-sig-mixed-models