[R] errors with lme4

[ONKELINX, Thierry]

Dear Alessio,

A few remarks.

- R-sig-mixed models is a better list for this kind of questions
- use the glmer() function if you want logistic or poisson regression
- the error you are getting is an indication that the model is too complex for the data
- watch for colinearity in the covariates

Best regards,

Thierry

> -----Oorspronkelijk bericht-----
> Van: r-help-bounces op r-project.org [mailto:r-help-bounces op r-project.org]
> Namens Alessio Unisi
> Verzonden: maandag 21 november 2011 18:20
> Aan: r-help op r-project.org
> Onderwerp: [R] errors with lme4
> Urgentie: Hoog
> 
> Dear list,
> i'm a new R user, so I apologize if the topic is already being addressed by some
> other user.
> 
> I'm trying to determine if the reproductive success of a species of bird is related
> to a list of covariates.
> 
> These are the covariates:
> §    elev: elevation of nest (meters)
> §    seadist: distance from the sea (meters)
> §    meanterranova: records of temperature
> §    minpengS1: records of temperature
> §    wchillpengS1: records of temperature
> §    minpengS2: records of temperature
> §    wchillpengS2: records of temperature
> §    nnd: nearest neighbour distance
> §    npd: nearest penguin distance
> §    eggs: numbers of eggs
> §    lay: laying date (julian calendar)
> §    hatch: hatching date (julian calendar)
> I have some NAs in the data.
> 
> I want to test the model with all the variable then i want to remove some, but
> the ideal model:
> GLM.1 <-lmer(fledgesucc ~ +lay +hatch +elev +seadist +nnd +npd
> +meanterranova +minpengS1 +minpengS2 +wchillpengS1 +wchillpengS2
> +(1|territory), family=binomial(logit), data=fledge)
> 
> doesn't work because of these errors:
> 'Warning message: In mer_finalize(ans) : gr cannot be computed at initial par
> (65)'.
> "matrix is not symmetric [1,2]"
> 
> If i delete one or more of the T records (i.e. minpengS2 +wchillpengS2) the
> model works...below and example:
> 
>  GLM.16 <-lmer(fledgesucc ~ lay +hatch +elev +seadist +nnd +npd
> +meanterranova +minpengS1 +(1|territory), family=binomial(logit),
> data=fledge)
> 
>  > summary(GLM.16)
> Generalized linear mixed model fit by the Laplace approximation
> Formula: fledgesucc ~ lay + hatch + elev + seadist + nnd + npd +
> meanterranova +      minpengS1 + (1 | territory)
>    Data: fledge
>  AIC   BIC logLik deviance
>  174 204.2    -77      154
> Random effects:
>  Groups    Name        Variance Std.Dev.
>  territory (Intercept) 0.54308  0.73694
> Number of obs: 152, groups: territory, 96
> 
> Fixed effects:
>                Estimate Std. Error z value Pr(>|z|)
> (Intercept)   14.136846  14.510089   0.974    0.330
> lay           -0.007642   0.267913  -0.028    0.977
> hatch         -0.025947   0.267318  -0.097    0.923
> elev           0.007481   0.027765   0.270    0.788
> seadist       -0.004277   0.004550  -0.940    0.347
> nnd           -0.035535   0.026504  -1.341    0.180
> npd            0.003788   0.005521   0.686    0.493
> meanterranova  1.242570   1.426158   0.871    0.384
> minpengS1     -0.399852   0.418722  -0.955    0.340
> 
> Correlation of Fixed Effects:
>             (Intr) lay    hatch  elev   seadst nnd    npd    mntrrn
> lay          0.411
> hatch       -0.515 -0.993
> elev        -0.015  0.141 -0.135
> seadist     -0.003 -0.023  0.019 -0.440
> nnd         -0.061  0.066 -0.059 -0.020  0.231
> npd          0.033 -0.108  0.100  0.298 -0.498 -0.338
> meanterranv  0.459 -0.118  0.075 -0.061  0.014 -0.048  0.130
> minpengS1   -0.540  0.015  0.035  0.032  0.000  0.039 -0.086 -0.970
> 
> 
> I've attached an example of my dataset only 15 rows just to see the
> dataset. Let me know if you need more informations.
> 
> Thanks in advance for your help and advices!
> regards
> 
> --
> Alessio Franceschi
> Phd student
> Dipartimento di Scienze Ambientali "G. Sarfatti"
> Università di Siena
> Via P.A. Mattioli, 8  - 53100 Siena (Italy)
> Cell. +393384431806
> email: franceschi6 op unisi.it; alfranceschi op alice.it