Dear list,
I am facing a problem with the rtmvtnorm function to sample from a
truncated multivariate normal when some of the truncation boundaries are
equal, e.g.:
>rtmvnorm(1, mean = rep(0,3), sigma = diag(3), lower=c(-1,0,-1),
upper=c(1,0,1),
algorithm="gibbs")
Error in checkTmvArgs(mean, sigma, lower, upper) :
lower must be smaller than or equal to upper (lower<=upper)
Of course, since it is all about numerics I could do
>rtmvnorm(1, mean = rep(0,3), sigma = diag(3), lower=c(-1,0,-1),
upper=c(1,1e-16,1),
algorithm="gibbs")
[,1] [,2] [,3]
[1,] -0.6221186 0 -0.6435531
but it is not entirely satisfying. Is it a normal behaviour of rtmvtnorm?
Thanks in advance,
jurgis
[[alternative HTML version deleted]]
______________________________________________
[email protected] mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
