multi-trait model without repeated individual observations

[Victor Papin]

Dear Facundo,

I would like to fit a model with 2 traits, corresponding to the circumference of the trees in two different sites. However, as this is not a clonal trial, each individual/genotype is only present at one site. Thus, for trait 1, I have observations for the first 50% of individuals, and NA for the remaining 50%. The reverse is true for trait 2. When I perform the adjustment with breedR I get this error message:

Error in eigen(x, symmetric = TRUE, only.values = TRUE) :

infinite or missing values in 'x'  

Is it not possible to fit a multi-trait model when no individual is observed for the 2 traits despite connections between individuals in the pedigree?

Thanks in advance,

Victor