yuendna:

In musto101/wilcox_R: What the package does (short line)

Usage

yuendna(x, y = NULL, tr = 0.2, alpha = 0.05)

Arguments

x
y
tr
alpha

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 (x, y = NULL, tr = 0.2, alpha = 0.05) 
{
    if (is.null(y)) {
        if (!is.matrix(x)) 
            stop("y is null and x is not a matrix")
        y = x[, 2]
        x = x[, 1]
    }
    if (length(x) != length(y)) 
        stop("The number of observations must be equal")
    m <- cbind(x, y)
    flag = (apply(is.na(m), 1, sum) == 2)
    m = m[!flag, ]
    x <- m[, 1]
    y <- m[, 2]
    flagx = is.na(y)
    flagy = is.na(x)
    m <- elimna(m)
    n = nrow(m)
    n1 = sum(flagx)
    n2 = sum(flagy)
    h = n - 2 * floor(tr * n)
    h1 = n1 - 2 * floor(tr * n1)
    h2 = n2 - 2 * floor(tr * n2)
    xbarn = mean(x, tr = tr, na.rm = TRUE)
    xbarn1 = 0
    if (h1 > 0) 
        xbarn1 = mean(x[flagx], tr = tr)
    ybarn = mean(y[!flagy], tr = tr, na.rm = TRUE)
    ybarn1 = 0
    if (h2 > 0) 
        ybarn1 = mean(y[flagy], tr = tr)
    lam1 = h/(h + h1)
    lam2 = h/(h + h2)
    est = lam1 * xbarn - lam2 * ybarn + (1 - lam1) * xbarn1 - 
        (1 - lam2) * ybarn1
    sex = trimse(elimna(x), tr = tr)
    sey = trimse(elimna(y), tr = tr)
    q1 <- (n - 1) * winvar(m[, 1], tr)
    q2 <- (n - 1) * winvar(m[, 2], tr)
    q3 <- (n - 1) * wincor(m[, 1], m[, 2], tr)$cov
    sen = sqrt((lam1^2 * q1 + lam2^2 * q2 - 2 * lam1 * lam2 * 
        q3)/(h * (h - 1)))
    SE = sqrt(sen^2 + (1 - lam1)^2 * sex^2 + (1 - lam2)^2 * sey^2)
    test = est/SE
    list(estimate = est, test = test, se = SE)
  }

musto101/wilcox_R documentation built on May 23, 2019, 10:52 a.m.