bff: Find features most associated with cluster membership

In RoheLab/vsp: Vintage Sparse PCA for Semi-Parametric Factor Analysis

Find features most associated with cluster membership

Description

Find features most associated with cluster membership

Usage

bff(loadings, features, num_best)

Arguments

loadings	An n by k matrix of weights that indicates how important that ith user is to the jth cluster, i.e., the Z or Y matrix calculated by vsp().
features	An n by d matrix of features measured for each node in the network.
num_best	An integer indicating how many of the top features for differentiating between loadings you want.

Details

See vignette("bff").

Value

An n by k matrix whose ⁠[i, j]⁠ entry is the ith "most important" feature for cluster j.

RoheLab/vsp documentation built on Nov. 6, 2024, 10:08 p.m.