Skip to contents

Scale data using a subset of the data

Source: R/LFQDataTransformer.R
scale_with_subset.Rd

this should reduce the overall variance.

Usage

scale_with_subset(lfqdata, lfqsubset, preserve_mean = FALSE, get_scales = TRUE)

Arguments

lfqdata

LFQData object with full dataset

lfqsubset

LFQData object with subset for computing scales

preserve_mean

default FALSE - sets mean to zero

get_scales

return a list of transformed data and the scaling parameters

Value

list with data, scales, and colname

See also

Other preprocessing: INTERNAL_FUNCTIONS_BY_FAMILY, apply_to_response_matrix(), filter_proteins_by_peptide_count(), get_robscales(), robust_scale()

Examples


bb <- sim_lfq_data_peptide_config(Nprot = 100)
#> creating sampleName from file_name column
#> completing cases
#> completing cases done
#> setup done
lfqdata <- LFQData$new(bb$data, bb$config)
lfqdata <- lfqdata$get_Transformer()$log2()$lfq
#> Column added : log2_abundance
s1 <- get_robscales(lfqdata)
res <- scale_with_subset(lfqdata, lfqdata)
#> Joining with `by = join_by(sampleName, isotopeLabel, protein_Id, peptide_Id)`
cfg <- lfqdata$get_config()$clone(deep = TRUE)
cfg$set_response(res$colname)
lfqres <- LFQData$new(res$data, cfg)
s2 <- get_robscales(lfqres)
stopifnot(abs(mean(s1$mads) - mean(s2$mads)) < 1e-6)