Evaluates whether simulated or integrated multi-batch single-cell data preserve
cell-type identity across distinct experimental batches or donors.
Trains a k-NN classifier on each individual batch and tests prediction accuracy
and macro F1 score on all other remaining batches.
Usage
evaluate_cross_batch_prediction(coords, cell_types, batch_info, k = 5)
Arguments
- coords
Embedding coordinates (cells x dimensions).
- cell_types
Factor or character vector of ground-truth cell type labels.
- batch_info
Factor or character vector of batch assignments.
- k
Integer number of nearest neighbors for the k-NN classifier (default 5).
Value
A list containing:
- mean_cross_batch_accuracy
Overall mean classification accuracy across all directed batch pairs
- mean_cross_batch_F1
Overall mean macro-averaged F1 score across all directed batch pairs
Examples
coords <- matrix(stats::rnorm(100), 50, 2)
batch <- factor(rep(c("B1", "B2"), length.out = 50))
ct <- factor(rep(c("TypeA", "TypeB"), length.out = 50))
evaluate_cross_batch_prediction(coords, ct, batch)
#> $mean_cross_batch_accuracy
#> [1] NaN
#>
#> $mean_cross_batch_F1
#> [1] NaN
#>
