doKmeans#
Last Updated: Oct 02, 2023
Description#
cluster cells using kmeans algorithm
Usage#
doKmeans(
gobject,
feat_type = NULL,
spat_unit = NULL,
expression_values = c("normalized", "scaled", "custom"),
feats_to_use = NULL,
genes_to_use = NULL,
dim_reduction_to_use = c("cells", "pca", "umap", "tsne"),
dim_reduction_name = "pca",
dimensions_to_use = 1:10,
distance_method = c("original", "pearson", "spearman", "euclidean", "maximum",
"manhattan", "canberra", "binary", "minkowski"),
centers = 10,
iter_max = 100,
nstart = 1000,
algorithm = "Hartigan-Wong",
name = "kmeans",
return_gobject = TRUE,
set_seed = TRUE,
seed_number = 1234
)
Arguments#
|
giotto object |
|
feature type (e.g. “cell”) |
|
spatial unit (e.g. “rna”, “dna”, “protein”) |
|
expression values to use (e.g. “normalized”, “scaled”, “custom”) |
|
subset of features to use |
|
deprecated use feats_to_use |
|
dimension reduction to use (e.g. “cells”, “pca”, “umap”, “tsne”) |
|
dimensions reduction name, default to “pca” |
|
dimensions to use, default = 1:10 |
|
distance method (e.g. “original”, “pearson”, “spearman”, “euclidean”, “maximum”, “manhattan”, “canberra”, “binary”, “minkowski”) |
|
number of final clusters, default = 10 |
|
kmeans maximum iterations, default = 100 |
|
kmeans nstart, default = 1000 |
|
kmeans algorithm, default to “Hartigan-Wong” |
|
name for kmeans clustering, default to “kmeans” |
|
boolean: return giotto object (default = TRUE) |
|
set seed (default = TRUE) |
|
number for seed |
Details#
Description on how to use Kmeans clustering method.
Value#
giotto object with new clusters appended to cell metadata
See Also#
kmeans