CLUSTROC

Relative operating characterics of multilevel clustering

     [E1,E2,N] = CLUSTROC(LABC,LABT)
     [E1,E2,N] = CLUSTROC(LABC,A)
      ROC = CLUSTROC(LABC,LABT)
      ROC = CLUSTROC(LABC,A)

Input
LABC	Index array, size [M,K], indices of cluster prototypes for M objects in K clusterings.
LABT	Double arrys, size [M,1] with true object labels.
A	Labeled PRTools dataset with M objects. It used for obtaining  the clusterings stored in LABC. LABT = GETNLAB(A)

Output
E1	Double array, size [1,NUMEL(N)], fraction of object pairs  erroneously assigned to the same cluster (Error I in ROC plot).
E2	Double array, size [1,NUMEL(N)], fraction of object pairs  erroneously assigned to different clusters (Error II in ROC plot).
N	Row vector with K or less elements with the corresponding number  of clusters in the clustering used for the vectors E1 and E2.
ROC	Relative operating characteristic, ready to be plotted by PLOTE.

Description

This routine is used for some performance measures used in CLUSTEVAL.

See also

datasets, cluste, clusth, clustk, clustkh, clustm, clustf, clustr, dcluste, dclustf, dclusth, dclustk, dclustm, dclustr, clusteval, plote,