[R-sig-ME] lme : Error in y[revOrder] - Fitted : non-conformablearrays

[David Duffy]

On Thu, 24 Feb 2011, Martha Ellis wrote:

> I've run across this error using lme().  Seems like it's got to be something
> small and format-y.  In trying to set up some reproducible data, I found a
> slight difference that changes whether or not you get the error message.
>
> So, originally the dataframe that I was working with has these attributes:
>> str(tmp)
> 'data.frame':   22 obs. of  3 variables:
> $ V1: int  2 2 2 2 2 3 3 3 3 3 ...
> $ V2: num [1:22(1d)] 1.01 1.04 1.07 1.02 1.03 ...
> $ V3: num  4.21 3.68 7.12 3.32 1.82 ...
>
> And if I run [an lme] model, I get the non-conformable arrays
> error:
>
> But, if you copy and paste the data back in, it drops the dimnames
> attribute. Here's the output of str():
>> str(tmp)
> 'data.frame':   22 obs. of  3 variables:
> $ species: int  2 2 2 2 2 3 3 3 3 3 ...
> $ hold   : num  1.01 1.04 1.07 1.02 1.03 ...
> $ react  : num  4.21 3.68 7.12 3.32 1.82 ...
>
> And now it runs, no problem:

Why does str(tmp) give the length only of V2 on the first occasion?  Can 
you dump("tmp") from that version for us to see?

Cheers, David Duffy.
-- 
| David Duffy (MBBS PhD)                                         ,-_|\
| email: davidD at qimr.edu.au  ph: INT+61+7+3362-0217 fax: -0101  /     *
| Epidemiology Unit, Queensland Institute of Medical Research   \_,-._/
| 300 Herston Rd, Brisbane, Queensland 4029, Australia  GPG 4D0B994A v