fcd.cluster: Clustering the estimators along the path.

In fcd: Fused Community Detection

Description

To use k-means to cluster the estimators along the path and get the community labels.

Usage

fcd.cluster(obj, K = 2)

Arguments

obj	it is an object generated by fcd.start,
K	input integer – the pre-specified number of communities, with the default value 2.

Value

an array of community labels list, of dimension number of non-isolated nodes x number of effective estimators. Each entry has value from 1 to K, as an index of the community label. Notice, the community labels are usually permutation-invariant.

Author(s)

Yang Feng, Richard J. Samworth and Yi Yu

References

Yang Feng, Richard J. Samworth and Yi Yu, Fused Community Detection, manuscript.

fcd documentation built on May 2, 2019, 3:28 p.m.