Partition the matrix
Usage
partition_by_kmeans(mat, n_repeats = 10)
partition_by_pam(mat)
partition_by_hclust(mat)
partition_by_kmeanspp(mat)Arguments
- mat
The submatrix in the binary cut clustering process.
- n_repeats
Number of repeated runs of k-means clustering.
Value
All partitioning functions split the matrix into two groups and return a categorical vector of labels of 1 and 2.
Details
These functions can be set to the partition_fun argument in binary_cut().
partition_by_kmeans(): Since k-means clustering brings randomness, this function performs
k-means clustering several times (controlled by n_repeats) and uses the final consensus partitioning results.
partition_by_pam(): The clustering is performed by cluster::pam() with the pamonce argument set to 5.
partition_by_hclust(): The "ward.D2" clusering method was used.
partition_by_kmeanspp(): It uses the kmeanspp method from the flexclust package.