scppNJW

Clustering of projected data using Ng, Jordan and Weiss (2002) normalised spectral clustering algorithm
IDX = SCPPNJW(K,V,X,WEIGHTS,SIGMA,PARAMS)

 Inputs:
    (K): number of clusters
    (V): Matrix defining projection subspace
    (X): Dataset (potentially micro-cluster centers)
    (WEIGHTS): Observations per microcluster (empty for no micro-clustering)
    (SIGMA): scaling parameter for Gaussian kernel
    (BETA,DELTA): parameters of similarity transformation function:
        if empty similarity between projections is based on Euclidean distance

 Output:
    (IDX): Cluster assignment vector \in {1,...,K}