[R-sig-ME] lme4 dev github version

[Ben Bolker]

On 13-09-06 10:33 AM, Gustaf Granath wrote:
> Thanks Ben. After a few re-start/install I got the gitHub version to
> work. However, refit() is not working.
> 
>> fm1 <- lmer(Reaction ~ Days + (Days | Subject), sleepstudy)
>> dat <- simulate(fm1)
>> fm3 <- refit(fm1, newresp = dat)
> Error in stopifnot(length(newresp <- as.numeric(as.vector(newresp))) ==  :
>   (list) object cannot be coerced to type 'double'

  Ah.

  simulate() returns a *list* of simulations, even if nsim=1.

  Therefore you just need refit(fm1, newresp=dat[[1]])

  We should add an example of this to the documentation.


>> sessionInfo()
> R version 3.0.1 (2013-05-16)
> Platform: i386-w64-mingw32/i386 (32-bit)
> 
> attached base packages:
> [1] stats     graphics  grDevices utils     datasets  methods base
> 
> other attached packages:
> [1] lme4_1.1-0      Matrix_1.0-12   lattice_0.20-15 devtools_1.3
> 
> loaded via a namespace (and not attached):
>  [1] digest_0.6.3   evaluate_0.4.7 grid_3.0.1     httr_0.2 MASS_7.3-26
>  [6] memoise_0.1    minqa_1.2.1    nlme_3.1-109   parallel_3.0.1
> RCurl_1.95-4.1
> [11] splines_3.0.1  stringr_0.6.2  tools_3.0.1    whisker_0.3-2
> 
>>> Hi,
>>> Is the development version on github working properly at the moment? I
>>> cannot get it to work.
>>> (I havent done this on a windows machine before so maybe im missing
>>> something)
>>    
>>> fm1 <- lmer(Reaction ~ Days + (Days|Subject), sleepstudy)
>>> Error in .Call(merPredDupdateDecomp, Ptr, NULL) :
>>>     Incorrect number of arguments (2), expecting 1 for
>>> 'merPredDupdateDecomp'
>>> In addition: Warning message:
>>> In fetch(key) : internal error -3 in R_decompress1
>>    In my experience this sort of error usually indicates that
>> something is out of sync: I would try closing out any running
>> R sessions, reinstalling lme4, and restarting R/reloading
>> (or, restart R and run library(devtools);
>> install_packages("lme4","lme4"))
>> PLease let us know if the problem persists.
>>
>>> I mainly need the warning/conv slot available in the dev version. The
>>> version on r-forge repository (99999-8) has this function as well but
>>> the refit() function doesnt seem to work in that version. I get the
>>> error:
>>> Error in stopifnot(length(newresp <- as.numeric(as.vector(newresp)))
>>> ==  :
>>>     (list) object cannot be coerced to type 'double'
>>    Hmm, I may take a look at that, or just put the latest
>> release lme4 version (I think we're up to 1.0-3) on the R-forge
>> repository ...
>>
>>> R version 3.0.1 (2013-05-16)
>>> Platform: i386-w64-mingw32/i386 (32-bit)
>>>
>>> other attached packages:
>>> [1] lme4_1.1-0          RcppEigen_0.3.1.2.1 Rcpp_0.10.4
>>> [4] Matrix_1.0-12       lattice_0.20-15     devtools_1.3
>