varcov1: varcov1

In kloke/rbgee: Rank-based GEE fits

Description

Computes asymptotic form for standard errors as descibed in Section 8.6.4 of Kloke and McKean (2014).

Usage

varcov1(eitbg, center, scores, v12inv, Dmat, wts)

Arguments

eitbg
center
scores
v12inv
Dmat
wts

Author(s)

Joseph W. McKean

References

Kloke and McKean (2014), Nonparametrics Using R, Boca Raton: Chapman-Hall.

Examples

##---- Should be DIRECTLY executable !! ----
##-- ==>  Define data, use random,
##--	or do  help(data=index)  for the standard data sets.

## The function is currently defined as
function (eitbg, center, scores, v12inv, Dmat, wts) 
{
    numc <- max(center)
    n <- length(eitbg)
    u <- (1:n)/(n + 1)
    sc <- getScores(scores, u)
    rs <- sc[rank(eitbg, ties.method = "first")]
    brrp <- matrix(rep(0, n * n), ncol = n)
    veeinv <- v12inv %*% diag(wts) %*% v12inv
    part1 <- t(Dmat) %*% veeinv %*% Dmat
    i1 <- 1
    i2 <- 1
    for (i in 1:numc) {
        rsc <- rs[center == i]
        ni <- length(rsc)
        rscbar <- mean(rsc)
        rsc <- rsc - rscbar
        rrp <- rsc %*% t(rsc)
        i2 <- i1 + ni - 1
        brrp[i1:i2, i1:i2] <- rrp
        i1 <- i1 + ni
    }
    mid <- t(Dmat) %*% v12inv %*% brrp %*% v12inv %*% Dmat
    varcov1 <- solve(part1) %*% mid %*% solve(part1)
    return(varcov1)
  }

kloke/rbgee documentation built on May 20, 2019, 12:34 p.m.