R help - May 2013 - how to get samples from rtmvnorm with large dimensions

Hi, dear all, 

I wish to get one sample (2500-d vector)  from the truncated multivariate
normal distribution, so I choose use the R function rtmvnorm() to do this. 
But the error information shows that for this function, the dimension should
be lower than 1000,
So could you help me to find out if there's any solution could do such
sampling from a truncated multivariate normal distribution with the sigma
dimension is 2500*2500??

Here is parts of my code for sampling: 

rg[k,]<- rtmvnorm(n=1, as.vector(SG1u), SG1, lower=rep(0,2500),
upper=rep(Inf,2500), algorithm="rejection")
and the error information is:
 "Error: mvt(lower = lower, upper = upper, df = 0, corr = corr, delta mean,
:
  only dimensions 1 <= n <= 1000 allowed" 

Thanks a lot. 
	[[alternative HTML version deleted]]