[R-sig-ME] Error en mer_finalize(ans) : Downdated X'X is not positive definite, 1. What is wrong with my model?

[PALACIO BLASCO, SARA]

Hi Emmanuel,

Yes, I have the same problem when using treatment + Bud_type.

And yes, I produced a contingency table for each species...

My conclusion is that glmer can not handle designs with nested fixed  
factors since it is not suitable to work with rank deficient designs.  
This might be the downside of using cutting-edge models!

Thanks a lot for your help,

Best wishes,

Sara
Quoting Emmanuel Curis <emmanuel.curis at parisdescartes.fr>:

> Hi Sara,
>
> If this is the case, you should have the same problem when using
> Treatment + Bud_Type instead of Treatment * Bud_Type. Is it the case ?
>
> When « checking per species », you mean you studied each row of this
> table separatly and for each of them, constructed the
> species:treatment contingency table and saw that in each cell there
> were several observations (at least 3 or 4)?
>
> something like by( Bud_type, table( Species, Treatment ) )...
>
> After that... I must say I have no further idea.
>
> Best regards,
>
> On Tue, Mar 05, 2013 at 12:05:32PM +0100, PALACIO BLASCO, SARA wrote:
> «
> « Below you can see the contingency table for Treatment*Bud_type with
> « the observations in each combination. As you can see the design is
> « not balanced but there are no combinations with particularly few
> « observations. I have checked this table per species and (taking into
> « account that each species only has one Bud_type), all treatment
> « levels have a representative number of observations...
> «
> « 	Treatment
> « Bud_type-80	-34.7	-30.2	-26.3	-22.1	-17.2	-13.1	-6.6	4	Total
> « hy	42	45	47	47	54	71	79	64	39	488
> « na	24	31	30	34	30	30	30	30	24	263
> « sc	62	71	72	68	71	82	69	78	62	635
> « Total  128	147	149	149	155	183	178	172	125	1386
> «
> «
> « The only issue I can see is that of each species only having one  
> Bud_type...
> «
> « Thanks for your help!
> «
> « sara palacio
>
> --
>                                 Emmanuel CURIS
>                                 emmanuel.curis at parisdescartes.fr
>
> Page WWW: http://emmanuel.curis.online.fr/index.html