wtedrb: Weighted Rank-Based Fit

In kloke/npsmReg2: R functions for Chapter 7 of Nonparametric Statistical Methods Using R

Description

This function computes the weighted rank-based fit. Input are the responses, design matrix, weights (diagonal or matrix, assumes postive definite), and scores. See Section 7.4 of Kloke and McKean (2014).

Usage

wtedrb(x, y, wts = diag(rep(1, length(y))), scores = wscores)

Arguments

x	design matrix
y	vector of responses
wts	either a matrix of weights or a vector of diagonal weights (assume weights are positive definite).
scores	rank-based scores to use, (default is Wilcoxon)

Value

yhatst	vector of weighted fitted values
ehatst	residuals of weighted fit
bstar	vector of weighted regression coefficients
se	vector of standard errors of bstar
vc	variance-covariance of bstar

Author(s)

Joe McKean ([e-mail]) and John Kloke ([e-mail])

References

Kloke, J. and McKean, J.W. (2014), Nonparametric statistical methods using R, Boca Raton, FL: Chapman-Hall

Examples

## Example 7.4.1 in Kloke and McKean (2014)
x <- seq(2,28,2)
y <- c(.08, 1.12,4.43,4.98,4.92,7.18,5.57,8.40,8.881,10.81,11.16,10.12,13.12,15.04)
wtedfit <- wtedrb(x,y)
wtedfit$bstar

kloke/npsmReg2 documentation built on May 20, 2019, 12:34 p.m.