Nothing
THRESH <- 1e-10
# #' Refit a row problem with given support
# #'
# #' @param r row index
# #' @param ind the index set of support
# #' @param S p-by-p sample covariance matrix
# #' @param delta a nonnegative tuning parameter for the ridge penalty for numerical stability, default is 0.01
refit_row <- function(r, ind, S, delta = 0.01){
# Refitting by plain MLE with
# bandwidth k = r-1-J given our estimator
p <- ncol(S)
res <- rep(0, r)
if(is.null(ind)){
res[r] <- 1/sqrt(S[r, r])
}
else{
# if J < r-1
# ind <- (J+1):(r-1)
# If S[ind,ind] is not invertible
# add a little ridge penalty to that
tmpVec <- solve((S[ind, ind] +
delta * diag(rep(1, length(ind)))),
S[ind, r])
res[r] <- 1/sqrt(S[r, r] + delta -
crossprod(S[r, ind], tmpVec))
res[ind] <- -tmpVec * res[r]
}
return(res)
}
# #' Refit the estimate of lower triangular matrix L with given support
# #'
# #' @param S p-by-p sample covariance matrix
# #' @param mat p-by-p estimate of lower triangular matrix L
# #' @param delta a nonnegative tuning parameter for the ridge penalty for numerical stability, default is 0.01
refit_matrix <- function(S, mat, delta = 0.01){
p <- ncol(S)
refit <- matrix(0, p, p)
refit[1, 1] <- 1/sqrt(S[1, 1])
for(r in seq(2, p)){
ind <- c()
for(j in seq(1,r-1)){
if(abs(mat[r, j]) >= THRESH){
ind <- c(ind, j)
}
}
refit[r, 1:r] <- refit_row(r = r, ind = ind,
S = S, delta = delta)
}
return(refit)
}
# #' Refit a path of estimates of lower triangular matrix L with given support
# #'
# #' @param S: p-by-p sample covariance matrix
# #' @param path: a list of p-by-p estimate of lower triangular matrix L along a path of tuning parameters
# #' @param delta: a nonnegative tuning parameter for the ridge penalty for numerical stability, default is 0.01
refit_path <- function(S, path, delta = 0.01){
p <- dim(path)[1]
nlam <- dim(path)[3]
refit <- array(NA, c(p, p, nlam))
for(i in seq(nlam)){
refit[, , i] <- refit_matrix(S, path[, , i], delta)
}
return(refit)
}
Any scripts or data that you put into this service are public.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.