Hi List, Is there anyway to estimate the parameters with integral forms? Here is the function. The idea is to calculate the real mean from the censored mean given normal distribution. I have the function as trun.mean<- function(mu,sigma,t) # t is the threshold { f <- function(x) (1/(sigma*sqrt(2*pi)))*exp(-(x-mu)^2/(2*sigma^2)); pdf.fun <- function(x) x*f(x); mut <- integrate(f,thre,upper=Inf)$value/integrate(pdf.fun,thre,upper=Inf)$value ; # true mean cat( mut,"\n"); } The pain is the integral forms in the function so I don't know how to put it in a parameter optimizator. Is there any way to do simulation in R to get true mu and sigma? if I know both censored mean and standard deviation? Thanks! Jen. [[alternative HTML version deleted]]