downsample_count_vec: Downsample a vector of counts.
In mirvie/mirmisc: Non-Modeling Helper Functions For Mirvie R Coders

Description Usage Arguments Details Value Examples

Uniformly downsample a vector of non-negative integers to have a specified sum. That is, keep randomly subtracting 1 from nonzero elements of the vector until it has the desired sum. Each count is equally likely to be taken. That is, an element with value 8 is 4 times more likely to be decremented than an element with value 2.

downsample_count_vec(vec, end_sum)

downsample_count_mat_rows(mat, end_sum)

downsample_count_mat_cols(mat, end_sum)

downsample_gene_counts(df, end_sum)

`vec`	A vector of non-negative integers.
`end_sum`	The number that you would like vector to sum to after downsampling. This must be less than the initial sum.
`mat`	A matrix of non-negative integers.
`df`	A data frame with gene names as columns.

downsample_count_mat_rows() and downsample_count_mat_cols() just do downsample_vec() to all rows and columns of a matrix using apply().

downsample_gene_counts() downsamples on the subset of columns in the data frame that have names in get_gene_names().

If end_sum > sum(vec), vec is returned unchanged.

A vector of non-negative integers.

downsample_count_vec(1:24, 24)
mat <- matrix(sample.int(100, size = 6^2, replace = TRUE), nrow = 6)
downsample_count_mat_rows(mat, end_sum = 6)
downsample_count_mat_cols(mat, end_sum = 6)
if (rlang::is_installed("mirmodels")) {
  ms_data <- mirmodels::get_ms_data(gene_predicate = ~ median(.) > 0)
  downsampled_ms <- downsample_gene_counts(ms_data, end_sum = 1e6)
}