reorder.default {stats}

Reorder Levels of a Factor

Description

reorder is a generic function. The "default" method treats its first argument as a categorical variable, and reorders its levels based on the values of a second variable, usually numeric.

Usage

reorder(x, ...)

## Default S3 method:
reorder(x, X, FUN = mean, ...,
        order = is.ordered(x), decreasing = FALSE)

Arguments

x

an atomic vector, usually a factor (possibly ordered). The vector is treated as a categorical variable whose levels will be reordered. If x is not a factor, its unique values will be used as the implicit levels.

X

a vector of the same length as x, whose subset of values for each unique level of x determines the eventual order of that level.

FUN

a function whose first argument is a vector and returns a scalar, to be applied to each subset of X determined by the levels of x.

...

optional: extra arguments supplied to FUN

order

logical, whether return value will be an ordered factor rather than a factor.

decreasing

logical, whether the levels will be ordered in increasing or decreasing order.

Details

This, as relevel(), is a special case of simply calling factor(x, levels = levels(x)[....]).

Value

A factor or an ordered factor (depending on the value of order), with the order of the levels determined by FUN applied to X grouped by x. By default, the levels are ordered such that the values returned by FUN are in increasing order. Empty levels will be dropped.

Additionally, the values of FUN applied to the subsets of X (in the original order of the levels of x) is returned as the "scores" attribute.

Author(s)

Deepayan Sarkar [e-mail]

See Also

reorder.dendrogram, levels, relevel.

Examples

require(graphics)

bymedian <- with(InsectSprays, reorder(spray, count, median))
boxplot(count ~ bymedian, data = InsectSprays,
        xlab = "Type of spray", ylab = "Insect count",
        main = "InsectSprays data", varwidth = TRUE,
        col = "lightgray")

bymedianR <- with(InsectSprays, reorder(spray, count, median, decreasing=TRUE))
stopifnot(exprs = {
    identical(attr(bymedian, "scores") -> sc,
              attr(bymedianR,"scores"))
    identical(nms <- names(sc), LETTERS[1:6])
    identical(levels(bymedian ), nms[isc <- order(sc)])
    identical(levels(bymedianR), nms[rev(isc)])
})