Centered Log-Ratio Transformation — clr

Applies the centered log-ratio transform row-wise. For log-scale inputs this subtracts the within-sample mean from each feature. For raw positive inputs the function applies `log(x + pseudocount)` before centering.

Usage

clr_transform(x, input_scale = c("log", "raw"), pseudocount = 1e-08)

Arguments

x: Numeric vector or matrix with samples in rows and features in columns.
input_scale: Either `"log"` (default) or `"raw"`.
pseudocount: Small positive constant added to raw inputs before logging. Ignored when `input_scale = "log"`.

Value

A numeric vector or matrix of the same dimensions as `x`.

References

Aitchison J. (1986). The Statistical Analysis of Compositional Data. Chapman and Hall.

Egozcue JJ, Pawlowsky-Glahn V, Mateu-Figueras G, Barcelo-Vidal C. (2003). Isometric logratio transformations for compositional data analysis. Mathematical Geology, 35(3), 279-300.

Gloor GB, Macklaim JM, Pawlowsky-Glahn V, Egozcue JJ. (2017). Microbiome datasets are compositional: and this is not optional. Frontiers in Microbiology, 8, 2224.

Examples

clr_transform(c(5, 3, 1))
#> [1]  2  0 -2
clr_transform(matrix(c(5, 3, 1, 4, 2, 0), nrow = 2, byrow = TRUE))
#>      [,1] [,2] [,3]
#> [1,]    2    0   -2
#> [2,]    2    0   -2