Get the multi-run NMF diagnostics

Getter function for the diagnostics of a multi-run NMF fit. For a stabilised_nmf_bulk() fit that is the per-run losses, convergence flags and the best-run index. For a consensus_nmf_bulk() fit it is the cluster stability, the relative errors and the per-component clustering table. Returns NULL for a single-run fit.

Usage

get_nmf_stability(object)

Arguments

object

The class, see BulkCoExp().

Value

For a stabilised fit, a list with losses, converged and best_idx. For a consensus fit, a list with stability, rel_error, rel_run_errors, clusters, cluster_sizes, n_dropped and n_empty_clusters. NULL if neither was run.

Examples

# clustering diagnostics of a consensus fit
syn <- generate_gene_module_data(n_samples = 24L, n_genes = 60L)
# NMF needs a non-negative matrix
mat <- syn$data - min(syn$data)
obj <- BulkCoExp(mat, syn$meta_data)
obj <- preprocess_bulk_coexp(
  obj, hvg = NULL, scaling = FALSE, .verbose = FALSE
)
obj <- consensus_nmf_bulk(obj, k = 4L, n_runs = 10L, .verbose = FALSE)
get_nmf_stability(obj)$cluster_sizes
#>    cluster     n
#>      <int> <int>
#> 1:       1    10
#> 2:       2    10
#> 3:       3    10
#> 4:       4    10