Dear R-people,
It's me again with, maybe, one more silly question for you. As a remainder, I am running version 2.7.1 on Windows Vista. I have small dataset which consists of:
# NestID: nest indicator for each chicken. Siblings sharing the same nest have the same nest indicator.
# Chick: chick indicator consisting of a unique ID for each single chick.
# Year: 1, 2.
# ClutchSize: 1-, 2- , 3-eggs.
# HO: hatching order within each clutch (1, 2, 3?[first, second and third-hatched chick]).
# SibComp: sibling competence: present/ absent (0, 1)
# Death10: death at ten days post-hatch (0, 1)
In order to account for lack of independence at the nest level (many chicks are nested in nest), I'd like to run a GLMM with random slopes and intercepts for nests.
Using lmer, whenever I try to model two-way interaction like specified below:
model1 <- lmer(Death10~HO*ClutchSize+(1|NestID),family=binomial,1)
.... the following error message pops up:
In mer_finalize(ans, verbose) : gr cannot be computed at initial par (65)
1. What does this error mean?
I look forward to hearing from you soon!
Best, Luciano
____________________________________________________________________________________
?Busc? desde tu celular!
Yahoo! oneSEARCH ahora est? en Claro
http://ar.mobile.yahoo.com/onesearch
Commonly observed error message in lmer
2 messages · Luciano La Sala, Daniel Ezra Johnson
This error came up earlier in the year and Doug Bates wrote:
This call is returning a warning about evaluation of the gradient at the initial values of the parameters. I'm not sure if it then goes on to optimize the approximated deviance.
If the approximated deviance is not being minimized for this model you may want to start with a simpler model, omitting some of the terms in the fixed effects.
Your fixed effects don't seem too complicated but there is a certain inherent non-independence between ClutchSize and HO (hatching order). For example if ClutchSize is 1 then HO must be 1. I wonder if that's what's causing the problem? Maybe you could try fitting separate models for each ClutchSize and observe the HO effect separately (when ClutchSize > 1)? You could also add a random slope as you said you wanted, that would be done with a term (HO|NestID). Hope this helps, Dan On Sat, Dec 27, 2008 at 2:39 AM, Luciano La Sala
<lucianolasala at yahoo.com.ar> wrote:
Dear R-people, It's me again with, maybe, one more silly question for you. As a remainder, I am running version 2.7.1 on Windows Vista. I have small dataset which consists of: # NestID: nest indicator for each chicken. Siblings sharing the same nest have the same nest indicator. # Chick: chick indicator consisting of a unique ID for each single chick. # Year: 1, 2. # ClutchSize: 1-, 2- , 3-eggs. # HO: hatching order within each clutch (1, 2, 3 [first, second and third-hatched chick]). # SibComp: sibling competence: present/ absent (0, 1) # Death10: death at ten days post-hatch (0, 1) In order to account for lack of independence at the nest level (many chicks are nested in nest), I'd like to run a GLMM with random slopes and intercepts for nests. Using lmer, whenever I try to model two-way interaction like specified below: model1 <- lmer(Death10~HO*ClutchSize+(1|NestID),family=binomial,1) .... the following error message pops up: In mer_finalize(ans, verbose) : gr cannot be computed at initial par (65) 1. What does this error mean? I look forward to hearing from you soon! Best, Luciano
____________________________________________________________________________________ ?Busc? desde tu celular! Yahoo! oneSEARCH ahora est? en Claro http://ar.mobile.yahoo.com/onesearch _______________________________________________ R-sig-mixed-models at r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models