Reports high-level diagnostics from [super_cluster2()] output to help sanity-check `distance` and `merge_ratio` choices before running downstream time-series analysis. A very high singleton fraction usually means `distance` was too tight; a dominant single cluster with a small `top1_frac` << expected can indicate over-aggressive merging.
Usage
cluster_stats(
clusters,
top_n = c(1, 10, 100),
singleton_max = 1L,
verbose = TRUE
)Arguments
- clusters
Tibble returned by [super_cluster2()], or a path to a CSV of that shape. Must contain a `sum_counts` column.
- top_n
Integer vector. Rank thresholds for cumulative-abundance share. Default `c(1, 10, 100)`.
- singleton_max
Integer. Maximum `sum_counts` treated as a singleton. Default `1L`.
- verbose
Logical. Print the summary to console. Default `TRUE`.
Value
A one-row tibble with columns:
`n_clusters`
`total_reads`
`n_singletons`, `singleton_frac`
`largest_cluster_size`, `largest_cluster_frac`
`topK_frac` for each `K` in `top_n`
Examples
if (FALSE) { # \dontrun{
result <- super_cluster2("sample1_barcodes.csv", distance = 3)
cluster_stats(result)
} # }