apply_per_group: Apply functions
In Rundmus/BAf-R_package: Binder Array data hanlding package in facility

Description Usage Arguments Value Author(s) Examples

apply_per_group : Apply a function to each group which is a subset of samples and binders stratified by the variables in @sinfo and @binder. The variable names are given by the argument by_s and by_b.

apply_per_group(baf, FUN, ..., by_s = NULL, by_b = NULL, passBAf = FALSE)

sapply_per_group(
  baf,
  FUN,
  ...,
  by_s = NULL,
  by_b = NULL,
  b_simplify = T,
  s_simplify = b_simplify
)

`baf`	an object of `BAf-class`
`FUN`	the function
`...`	the arguments that will be passed to FUN
`by_s, by_b`	by which the samples (`by_s`) or binders (`by_b`) are grouped. The `FUN` is applied to each group and the results are combined. This can be given by 1) a `character` string of a column name in `@sinfo` or `@binder`, and 2) a `factor` vector that can be used in the stratification. If it is NULL as the default, there will be no stratification. So, the `FUN` will be applied to entire data.
`passBAf`	if BAf object instead of matrix in @.Data is passed when applying the `FUN`.
`b_simplify`	if the output should be simplified like `simplify` for `sapply`
`s_simplify`	if the output within stratified group by the `by_s` should be simplified.

an object of BAf-class

Mun-Gwan Hong <mun-gwan.hong@scilifelab.se>

data(sba)
summary(sba)

# mean per plate and assay
apply_per_group(sba, mean, by_s= "plate", by_b= "assay")

# calculate median after stratifying samples by "plate" and "sex"
apply_per_group(sba, median, na.rm= TRUE, by_s= c("plate", "sex"))

# verify the first median computed by above line
median(sba@.Data[with(sba@sinfo, plate == "1" & !is.na(sex) & sex == "female"), ])

# compute coefficients of variation (CV) of \code{MIX_1} wells
apply_per_group(sba[sba@sinfo$cohort == "MIX_1", 1:5], FUN= function(j) {
    apply(j, 2, function(k) sd(k, na.rm= TRUE)/mean(k, na.rm= TRUE))
})