cenNID {censNID} | R Documentation |
MLE estimation using AS 138 for censored and confined normal.
cenNID(y, L=rep(-Inf, length(y)), U=rep(Inf,length(y)))
y |
observations |
L |
lower bound for each observation that is equal to -Inf when none, default -Inf |
U |
upper bound for each observation that is equal to Inf when none, default Inf |
a list with the following components:
est |
matrix with mean and sd and their se |
covMat |
covariance matrix |
nobs |
number of complete observations |
itercount |
number of iterations |
ifault |
fault indicator. 0-converged. -1:convergence not obtained in default number of allowed iterations. -2:fewer than 2 complete obserations. -3:error with confined estimator initial values (should not occur in R version). -4:invalid specification for limits with confined observation. |
A. I. McLeod aimcleod@uwo.ca
M. S. Wolynetz (1979). Algorithm AS 138: Maximum Likelihood Estimation from Confined and Censored Normal Data. Journal of the Royal Statistical Society. Series C (Applied Statistics), Vol. 28, No. 2, pp. 185-195
set.seed(32173217) n<-100 z <- rnorm(n) U <- rep(Inf, n) L <- rep(-Inf, n) cy <- -1 ind <- z > cy m <- sum(as.integer(ind)) y <- c(z[ind], rep(cy, n-m)) L[(m+1):n] <- cy cenNID(y, L, U)