We expect an kmeans_cluster
HDF5 group at the root of the file, containing the k-means clustering parameters and results.
The group itself should contain the parameters
and results
subgroups.
Definitions:
in_use
: whether k-means is the chosen clustering method inchoose_clustering
.num_cells
: number of cells remaining after QC filtering.
parameters
should contain:
k
: a scalar integer specifying the number of clusters to create.
If in_use = true
, results
should contain:
-
clusters
: an integer dataset of length equal tonum_cells
. This contains the cluster assignment for each cell, which should lie in[0, k)
.The total number of clusters may be less than
k
, e.g., when there are too few cells. ForN
clusters, there should be at least one occurrence of each integer in[0, N)
.
If in_use = false
, clusters
may be absent.
Nonetheless, if it is present, it should follow the constraints listed above.
Added in version 1.0.