[R-sig-ME] glmer error

[André Luis Neves]

Dear all:

I`m trying to fit a model using glmer from qPCR counts. The model is
running fine to all the variables (archaea, protozoa and fungi counts), but
not for bacteria, which have by far a larger population than the other
communities.

--------------------------
Here is the model:

M1 <- glmer (Baci ~ Starch + Crossover + (1|fID:Crossover) + Period, family
= gaussian(log), data=qPCR)

Here is the error:

'Error in (function (fr, X, reTrms, family, nAGQ = 1L, verbose = 0L, maxit
= 100L,  :Downdated VtV is not positive definite'

---------------------------

I have tried centering the continuous predictor variables and optimizing
the glmerControl
(for instance,optimizer="bobyqa",check.conv.grad=.makeCC("warning",2e-3)),
nAGQ = 10), but haven`t got anything.

Could anyone please help me to find a solution?

Thank you so much for your attention.

-- 
Andre

	[[alternative HTML version deleted]]