Center a delayed matrix along rows or columns

Usage

d_center(x, dim = c("rows", "cols"), na.rm = FALSE)

Arguments

x: A delarr.
dim: Dimension along which to subtract the mean.
na.rm: Logical; remove missing values when computing the centre.

Value

A delarr with a deferred centering operation.

Examples

mat <- matrix(c(1, 2, 3, 10, 20, 30), nrow = 2, ncol = 3)
darr <- delarr(mat)

# Center rows (subtract row means)
centered_rows <- darr |> d_center(dim = "rows") |> collect()
centered_rows
#>      [,1] [,2] [,3]
#> [1,]   -7   -5   12
#> [2,]  -12   -4   16
rowMeans(centered_rows)  # Should be ~0
#> [1] 0 0

# Center columns (subtract column means)
centered_cols <- darr |> d_center(dim = "cols") |> collect()
colMeans(centered_cols)  # Should be ~0
#> [1] 0 0 0