replace_missing_values: Replace missing values in a CF-MS chromatogram matrix

In fosterlab/CFTK: The Co-Fractionation Toolkit

Description

Sanitize different representations of missing values (e.g., zeroes, NaNs, or infinite values) in a chromatogram matrix, then optionally replace them all with zeroes or with near-zero noise.

Usage

replace_missing_values(
  mat,
  missing = c("NA", "zero", "noise"),
  noise_floor = 0.001
)

Arguments

mat	a CF-MS chromatogram matrix, with proteins in rows and fractions in columns
missing	one of 'NA', 'zero', or 'noise'. If 'zero', missing values will be replaced with zeroes. If 'noise', missing values will be replaced with random noise, sampled from a uniform distribution with a maximum value of noise_floor. If 'NA', missing values will be left as NAs.
noise_floor	upper limit of the uniform distribution for missing = 'noise'

Value

a copy of the input matrix with missing values optionally replaced

Examples

mat = matrix(rnorm(100), ncol = 10, nrow = 10,
             dimnames = list(paste0('protein_', LETTERS[1:10]),
                             paste0('fraction_', seq_len(10))))
mat[sample(seq_along(mat), 5)] = 0
mat[sample(seq_along(mat), 8)] = NA
mat[sample(seq_along(mat), 2)] = Inf
sanitized = replace_missing_values(mat, missing = 'zero')
table(sanitized == 0)

fosterlab/CFTK documentation built on Jan. 19, 2021, 10:31 p.m.