DCrr {quantiseqr} | R Documentation |
Perform robust regression
DCrr(b, A, method = c("hampel", "huber", "bisquare"), scaling = NULL)
b |
Numeric vector containing the right-hand side of the quadratic function to be minimized. |
A |
Numeric matrix containing the coefficients of the quadratic function to be minimized. |
method |
Character specifying the robust regression method to be used among deconvolution methods: "hampel", "huber", or "bisquare". Default: "hampel". |
scaling |
A vector of scaling factors to by applied to the estimates. Its length should equal the number of columns of A. |
The MASS::rlm()
function is used as underlying framework. Please
refer to that function for more details.
A vector containing robust least-square estimates.
data(dataset_racle) mixture <- dataset_racle$expr_mat signature.file <- system.file( "extdata", "TIL10_signature.txt", package = "quantiseqr", mustWork = TRUE) signature <- read.table(signature.file, header = TRUE, sep = "\t", row.names = 1) scaling.file <- system.file( "extdata", "TIL10_mRNA_scaling.txt", package = "quantiseqr", mustWork = TRUE) scaling <- as.vector( as.matrix(read.table(scaling.file, header = FALSE, sep = "\t", row.names = 1))) cgenes <- intersect(rownames(signature), rownames(mixture)) b <- as.vector(as.matrix(mixture[cgenes,1, drop=FALSE])) A <- as.matrix(signature[cgenes,]) # cellfrac <- quantiseqr:::DCrr(b = b, A = A, scaling = scaling)