Calculates the weighted medians for each row (column) in a matrix

Calculates the weighted medians for each row (column) in a matrix.

rowWeightedMedians(x, w = NULL, rows = NULL, cols = NULL,
  na.rm = FALSE, ..., useNames = TRUE)

colWeightedMedians(x, w = NULL, rows = NULL, cols = NULL,
  na.rm = FALSE, ..., useNames = TRUE)

Arguments

x

An NxK matrix or, if dim. is specified, an N * K vector.

w

A numeric vector of length K (N).

rows

A vector indicating subset of rows to operate over. If NULL, no subsetting is done.

cols

A vector indicating subset of columns to operate over. If NULL, no subsetting is done.

na.rm

If TRUE, missing values are excluded.

...

Additional arguments passed to weightedMedian().

useNames

If TRUE (default), names attributes of the result are set, otherwise not.

Value

Returns a numeric vector of length N (K).

Details

The implementations of these methods are optimized for both speed and memory. If no weights are given, the corresponding rowMedians()/colMedians() is used.

See also

Internally, weightedMedian() is used. See rowMedians() and colMedians() for non-weighted medians.

Author

Henrik Bengtsson

Examples

x <- matrix(rnorm(20), nrow = 5, ncol = 4)
print(x)
#>            [,1]       [,2]       [,3]       [,4]
#> [1,] -0.6487490  1.7879817 -0.6099912 -1.4171540
#> [2,] -1.8689043  0.8673005 -0.3132811  0.6107810
#> [3,] -1.2069966  0.7025809 -0.9221298  0.2415208
#> [4,]  1.8972581  1.6235322 -0.2564806  1.3039648
#> [5,] -0.2615856 -1.5901587 -2.0828546 -0.3345286

# Non-weighted row averages
mu_0 <- rowMedians(x)
mu <- rowWeightedMedians(x)
stopifnot(all.equal(mu, mu_0))

# Weighted row averages (uniform weights)
w <- rep(2.5, times = ncol(x))
mu <- rowWeightedMedians(x, w = w)
stopifnot(all.equal(mu, mu_0))

# Weighted row averages (excluding some columns)
w <- c(1, 1, 0, 1)
mu_0 <- rowMedians(x[, (w == 1), drop = FALSE])
mu <- rowWeightedMedians(x, w = w)
stopifnot(all.equal(mu, mu_0))

# Weighted row averages (excluding some columns)
w <- c(0, 1, 0, 0)
mu_0 <- rowMedians(x[, (w == 1), drop = FALSE])
mu <- rowWeightedMedians(x, w = w)
stopifnot(all.equal(mu, mu_0))

# Weighted averages by rows and columns
w <- 1:4
mu_1 <- rowWeightedMedians(x, w = w)
mu_2 <- colWeightedMedians(t(x), w = w)
stopifnot(all.equal(mu_2, mu_1))