R/separable1v.R

In sensitivitymv: Sensitivity Analysis in Observational Studies

separable1v <-
function (ymat, gamma = 1)
    {
        n <- dim(ymat)[1]
        m <- dim(ymat)[2]
        mumax <- 0
        varmax <- 0
        tstat <- 0
        for (i in 1:n) {
            if (!is.na(ymat[i, 1])) {
                rk <- sort(as.vector(unlist(ymat[i, ])))
                rk <- sort(rk[!is.na(rk)])
                mi <- length(rk)
                mumaxi <- (-Inf)
                varmaxi <- (-Inf)
                for (ai in 1:(mi - 1)) {
                    denom <- ai + gamma * (mi - ai)
                    mua <- (sum(rk[1:ai]) + gamma * sum(rk[(ai +
                                                                1):mi]))/denom
                    vara <- ((sum(rk[1:ai]^2) + gamma * sum(rk[(ai +
                                                                    1):mi]^2))/denom) - (mua^2)
                    if (mua > mumaxi) {
                        mumaxi <- mua
                        varmaxi <- vara
                    }
                    else if (mua == mumaxi)
                        varmaxi <- max(varmaxi, vara)
                }
                mumax <- mumax + mumaxi
                varmax <- varmax + varmaxi
                tstat <- tstat + ymat[i, 1]
            }
        }
        tstat <- as.vector(tstat)
        expect <- mumax
        vartotal <- varmax
        dev <- (tstat - expect)/sqrt(vartotal)
        pval <- 1 - stats::pnorm(dev)
        list(pval = pval, deviate = dev, statistic = tstat, expectation = expect,
             variance = vartotal)
}