rowGrpNA: Count Number Of NAs Per Row And Group Of Columns

Description

This functions allows easy counting of the number of NAs per row and group in data organized multiple sub-groups (in columns).

Usage

rowGrpNA(
  mat,
  grp = NULL,
  initColOrder = TRUE,
  mode = "simple",
  digits = 2,
  silent = FALSE,
  debug = FALSE,
  callFrom = NULL
)

Value

This function returns a matrix (or numeric vector if all grouping falls in a single class or charcter-vector if all groups get collapsed in single chain of characters) with number of NAs per group

Arguments

mat: (matrix, data.frame, list or 'MArrayLM') data to count the number of NAs; in case of furnishing a list or MArrayLM-class a list-element called 'isNA' will be used as mat or all NA positions of list-element called 'raw'
grp: (character or factor) defining which columns should be grouped (considered as replicates). If grp is not provided and mat is list or 'MArrayLM'-class containg a list-element 'setup' (as produced with testing from package wrProteo the group-association will be taken from there
initColOrder: (logical) ifFALSE the columns may be in (alphabetically) sorted order (and thus not in the order of argument grp)
mode: (character) allows to switch between different types of output : mode="simple" will return simply the number of NAs per line and group as integer; mode="simpleCollapse" will collapse all columns into single chain of characters with the number of NAs per line and group (as character); mode="complete" will return number of NAs per line and group accompanied by the number of columns associated with the very group (as character); mode="completeCollapse" will collapse all columns into single chain of characters with the number of NAs per line and group accompanied by the number of columns associated with the very group (as character); mode="ratio" will return the ratio of NAs per line and group (as numeric); mode="ratioCollapse" will collapse all columns into single chain of characters with the ratio of NAs per line and group(as character);
digits: (integer, length=1) number of decimal digits from rounding (when ratios are calculated);
silent: (logical) suppress messages
debug: (logical) additional messages for debugging
callFrom: (character) allows easier tracking of messages produced

Details

First of all, this function allows counting the number of NAs per line and sub-group of columns as defined by argument grp. Furthermore, the counting can be expressed relative to the size of the sub-groups, either as ratio or by giving the NA-count using mode and all results may also get collasped in as single (character) vector, however, the names of the groups do not get conserved.

This function has several options via tehe argument 'mode': - mode="simple" will return simply the number of NAs per line and group as integer; - mode="simpleCollapse" will collapse all columns into single chain of characters with the number of NAs per line and group (as character); - mode="complete" will return number of NAs per line and group accompanied by the number of columns associated with the very group (as character); - mode="completeCollapse" will collapse all columns into single chain of characters with the number of NAs per line and group accompanied by the number of columns associated with the very group (as character); - mode="ratio" will return the ratio of NAs per line and group (as numeric); - mode="ratioCollapse" will collapse all columns into single chain of characters with the ratio of NAs per line and group(as character);

With mode="complete" or mode="ratio" the resulting column-names will state 'nNA.' plus 'ratio.' (if a ratio-mode was chosen).

Examples

Run this code

mat2 <- matrix(1:72, nrow=9, ncol=8, dimnames=list(letters[1:9],LETTERS[1:8]))
set.seed(2025); mat2[sample.int(prod(dim(mat2)), 2*nrow(mat2), replace=FALSE)] <- NA
set.seed(2026); grp2 <- as.factor(sample.int(n=3, size=ncol(mat2), replace=TRUE))
  table(grp2)

## overal number of NAs per row
rowSums(is.na(mat2)) 

## number of NAs per row and group
head(rowGrpNA(mat2, grp2))
head(rowGrpNA(mat2, grp2, mode="complete"))
head(rowGrpNA(mat2, grp2, mode="ratio"))   

## mimick output from testing from package wrProteo
dat1 <- list(isNA =is.na(mat2), setup=list(grp=grp2))
head(rowGrpNA(dat1, mode="simple"))
head(rowGrpNA(dat1, mode="complete"))
head(rowGrpNA(dat1, mode="ratioCollapse"))