Called by link{saws}
PsiTildeCalc(u, H, test, omega, vm, vminv, p, K)
PsiHatCalc(u, H)
dfCalc(Psi, u, omega, H, test, vm)
sawsCalc(beta, u, omega, test = diag(p),
beta0 = matrix(0, p, 1), conf.level = 0.95,
method = c("d3", "d5", "d1", "d2", "d4", "dm"),
bound=.75)
length of coefficient vector
number of terms in estimating equation
coefficient vector of length p
K by p score matrix, u[i,] is the ith term in the estimating equation
K by p adjustment matrix, diag(H[i,])=Hi in Fay and Graubard, 2001
r by p test matrix
K by p by p array, omega[i,,] estimates -du/dbeta.
p by p matrix,model based variance
inverse of vm
confidence level for confidence intervals
see saws
bound for bias correction, denoted b in Fay and Graubard, 2001