getVariancePartialRConditional:

In BERENZ/RISQ: What the package does (short line)

Usage

getVariancePartialRConditional(partialIndicator, indicator, sampleData, sampleDesign)

Arguments

partialIndicator
indicator
sampleData
sampleDesign

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 (partialIndicator, indicator, sampleData, sampleDesign) 
{
    byCategory <- partialIndicator$byCategory
    variable <- partialIndicator$variable
    sigma <- indicator$sigma
    weights <- sampleDesign$weights
    nPopulation <- sum(weights)
    propDelta <- indicator$prop - partialIndicator$propMeanByOthers
    zDelta <- indicator$z - partialIndicator$zMeanByOthers
    variance <- numeric()
    for (index in seq(nrow(byCategory))) {
        label <- byCategory[index, "category"]
        delta <- ifelse(sampleData[[variable]] == label, 1, 0)
        zDeltaWeight <- zDelta * delta * weights
        propDelta2 <- propDelta * propDelta * delta
        A <- matrix(propDelta, nrow = 1) %*% zDeltaWeight
        B <- t(zDelta) %*% zDeltaWeight
        V1 <- 4 * A %*% sigma %*% t(A)
        V2 <- 2 * getTrace(B %*% sigma %*% B %*% sigma)
        V3 <- sampleDesign$getVarTotal(sampleDesign, propDelta2)
        variance[index] <- 0.25 * (V1 + V2 + V3)/(nPopulation * 
            sum(propDelta2 * weights))
    }
    partialIndicator$byCategory <- within(byCategory, PcUnadjSE <- sqrt(variance))
    return(partialIndicator)
  }

BERENZ/RISQ documentation built on May 5, 2019, 10:25 a.m.