batch_separation: Compute the batch separation metric
In YutongWangUMich/corgi: CORGI: Curve-fit Objective Ranking of Gene Importance.

Description Usage Arguments Value Examples

Compute the batch separation metric

1	batch_separation(emb, batch)

`emb`	the embedding of the cells where each row is a cell, and each column is a feature (e.g. principal component)
`batch`	the batch labels of the cells

Cohen's kappa of the the predicted batch labels with the true batch label on the entire dataset (i.e., training accuracy)

# suppose that iris[["Species"]] are the batch labels
batch_separation(emb = iris[,1:2],batch = iris$Species) # lower batch separation present in dim = 1:2
plot(iris[,1],iris[,2],col = iris[["Species"]])

batch_separation(emb = iris[,3:4],batch = iris$Species) # higher batch separation in dim = 3:4
plot(iris[,3],iris[,4],col = iris[["Species"]])