The (Weighted) Interquartile Range

Compute the interquartile range of x, with optional weights.

Usage

iqrX(x, weights = NULL, na.rm = FALSE, type = NULL)

Arguments

x

numeric vector

weights

optional numeric vector of non-negative sample weights, of the same length as x

na.rm

logical; whether to remove missing values. If FALSE and x contains missing values, the result is NA.

type

integer selecting the quantile algorithm, see quantileX(). The default NULL picks the one that suits the branch taken: 7 without weights, as in stats::IQR(); 2 with weights, which reads them as relative weights and depends only on their ratios. Pass type = 7 explicitly to read the weights as replication counts.

Value

numeric scalar containing the interquartile range

Details

The IQR is the difference of the 0.75 and the 0.25 quantile as computed by quantileX(). Without weights the result is identical to stats::IQR(), except that a missing value yields NA instead of an error when na.rm = FALSE.

See also

medianX(), quantileX(), stats::IQR(), stats::quantile()

Other dispersion: coefVar(), madX(), meanAbsDev(), meanSE(), rangeX(), varX()

Examples

x <- c(3.7, 3.3, 3.5, 2.8)
w <- c(5, 5, 4, 1) / 15

iqrX(x)
#> [1] 0.375
iqrX(x, weights = w)
#> [1] 0.4
iqrX(x, weights = w * 15)                 # same: only the ratios count
#> [1] 0.4
iqrX(x, weights = w * 15, type = 7)       # replication counts
#> [1] 0.4

iqrX(c(x, NA))                            # NA, no error
#> [1] NA
iqrX(c(x, NA), na.rm = TRUE)
#> [1] 0.375