[R-sig-ME] false convergence of lmer model

[Ben Bolker]

Tara Woud <t.y.woud at ...> writes:

> 
> Dear all,
> 
> To determine the effect of weather parameters on the amount of core areas
> of mallards (71 different mallards with measurements on several days), I'm
> trying to fit the following model to my data:
> 
> model1 <- lmer(Core_areas  ~ c_MaxTemp + I(c_MaxTemp^2) + c_MinTemp + I(
> c_MinTemp^2) + c_GemTemp + I(c_GemTemp^2) + c_D.rain + c_S.rain + c_Gemwind +
> c_Hwind + c_Lwind + (1 | Mallard)   # random slope | random factor (here
> individual is random factor), family = poisson)
> 
> But then I get the following warning message:
> 
> In mer_finalize(ans) : false convergence (8)
> What consequences does this have? And had anybody an idea how I can solve
> this??
> 
> Best,
> 
> Tara
> 

  [snip]

  Are you willing to try out the development version of lme4
(which will, hopefully, be released and become the CRAN version
in about a week anyway)?  See the Installation section at
https://github.com/lme4/lme4/blob/master/README.md  ...

  The new version tends to have fewer problems with false convergence
and more flexibility for exploring and fixing convergence problems
if they do occur.

  Ben Bolker