Error with covariate.predictions

[rcscott]

Hi all,

I am working with a robust design multi-scale model. I am attempting to use model averaging, and would like to retain all models (drop = FALSE), but when I do I get the error "Error in if (any(diag(x$vcv[[i]]) < 0)) { :
missing value where TRUE/FALSE needed"." My guess is that this is because of the the vcv in my model list of -INF or NaN values. However my understanding is that the drop = F argument allows for inclusion of models even with vcv that include NaN or inf.

So my question is whether this can be fixed, and whether it should be fixed for my models? I have included a brief snippet of my code below incase it's just a syntax error. Apologies for not providing a short example: I tried running this on sample data provided with RDMultScalOCc, but there it seems to work just fine.

temp_data <- data.frame(
temp1 = seq(from = 12.0, to = 31.0, by = 1)
)
test <- covariate.predictions(psi1eps1gamma1, #marklist
data = temp_data,
indices = c(14,66), #2 indices for 2 different spp
drop = FALSE
)