Keyboard shortcuts

Press ← or → to navigate between chapters

Press S or / to search in the book

Press ? to show this help

Press Esc to hide this help

Statscraft 0.7.0, release v0.7.0

CLUSTER

Hierarchical clustering of cases by one of 37 proximity measures and seven methods.

Syntax

CLUSTER varlist | /MATRIX=IN('file' | dataset | *)
        [/MEASURE=SEUCLID | EUCLID | COSINE | CORRELATION | BLOCK | CHEBYCHEV | MINKOWSKI(p) |
                  POWER(p,r) | CHISQ | PH2 | binary[(p[,np])]]
        [/METHOD=BAVERAGE | WAVERAGE | SINGLE | COMPLETE | CENTROID | MEDIAN | WARD [(rootname)] ...]
        [/PRINT=[SCHEDULE] [CLUSTER(min[,max])] [DISTANCE] [NONE]] [/SAVE=CLUSTER(min[,max])]
        [/ID=var] [/MISSING=EXCLUDE | INCLUDE] [/PLOT=[VICICLE | HICICLE[(min,max,inc)]] [DENDROGRAM] [NONE]].
binary: RR SM JACCARD DICE SS1 RT SS2 K1 SS3 K2 SS4 HAMANN OCHIAI SS5 PHI LAMBDA D Y Q BEUCLID SIZE
PATTERN BSEUCLID BSHAPE DISPER VARIANCE BLWMN, from the 2 x 2 table of presence (p, default 1) and
absence (np, default 0; with p alone every other value is absence). The items are the cases valid on
every variable; the weight variable is not used. Each step joins the most similar pair of clusters
(smallest distance or largest similarity) into the lower-numbered one and updates the proximities by
the formula of the method; BAVERAGE (default) and WAVERAGE choose by the mean proximity between or
within the clusters. Tables: Case Processing Summary, Proximity Matrix (DISTANCE), and for each
method the Agglomeration Schedule (default; for WARD the coefficient is the within-cluster sum of
squares) and Cluster Membership (CLUSTER), the clusters of a solution numbered in the order of their
first case. ID labels the cases. SAVE adds the membership of each solution, named rootname plus the
number of clusters, or CLUn_m. MATRIX=IN reads the rows of type PROX of a matrix file, a dissimilarity
matrix unless the value label of PROX says SIMILARITY. PLOT is accepted with a warning that charts
are not drawn yet. Not supported yet: MATRIX=OUT, SAVE with MATRIX=IN.

Example

CLUSTER q1 TO q5 /MEASURE=SEUCLID /METHOD=WARD /PRINT=SCHEDULE CLUSTER(2,4) /SAVE=CLUSTER(3).

statscraft help CLUSTER prints the same text in the terminal. All commands