Getting around convergence issues by restarting it from the fail point

I'm running some glmms and when I run the code: Model5 <- glmer.nb(Count ~ Light + (1|Site), data = dframe5)

I get the error messages:

1: In checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
unable to evaluate scaled gradient
2: In checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, :
Model failed to converge: degenerate Hessian with 2 negative eigenvalues

It was suggested that I could get around these messages by restarting the model from a different point using the code:

ss <- getME(Model5,c("theta","fixef"))
Model5b <- update(Model5,start=ss)

This has stopped me from getting convergence errors and given me normal-seeming results but I wanted to double check that this is an OK thing to do.

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.