Skip to contents

Calculate the Hedge's G effect between two matrices

calculate_effect_size.Rd

This function takes two matrices in and calculate on a per column basis the Hedge's G effect size and the standard error. These results can be subsequently used for meta-analyses or other approaches.

Usage

calculate_effect_size(
  mat_a,
  mat_b,
  small_sample_correction = NULL,
  .verbose = TRUE
)

Arguments

mat_a

Numerical matrix. Contains the values for group a. Assumes that rows = samples, and columns = features.

mat_b

Numerical matrix. Contains the values for group b.

small_sample_correction

Can be NULL (automatic determination if a small sample size correction should be applied) or Boolean.

.verbose

Boolean that controls verbosity of the function.

Value

x, robustly scaled.

Examples

# Hedge's G between two groups of ten samples
set.seed(42)
mat_a <- matrix(rnorm(100), nrow = 10, ncol = 10)
mat_b <- matrix(rnorm(100, mean = 1), nrow = 10, ncol = 10)
colnames(mat_a) <- colnames(mat_b) <- sprintf("gene_%i", 1:10)
res <- calculate_effect_size(mat_a, mat_b, .verbose = FALSE)
head(res$effect_sizes)
#> [1] -0.7178191 -0.6166706 -0.3550481 -1.4490642 -0.9636552 -0.6221444