R/bg.checks.r

In bgeva: Binary Generalized Extreme Value Additive Models

bg.checks <- function(x){

e.v <- eigen(x$He)$values

cat("\nLargest absolute gradient value: ",max(abs(x$fit$gradient)))

if (min(e.v) > 0) cat("\nInformation matrix is positive definite\n")
else cat("\nEigenvalue range of the information matrix: [",min(e.v),",",max(e.v),"]\n", sep = "")

if(x$l.sp!=0 && x$fp==FALSE){
cat("\nTrust region iterations before smoothing parameter selection: ",x$iter.if)
cat("\nLoops for smoothing parameter selection: ",x$iter.sp)
cat("\nTrust region iterations within smoothing parameter selection loops: ",x$iter.inner,"\n\n")}
                        else{cat("\nTrust region iterations: ",x$iter.if,"\n\n")}

}