Statscraft 0.7.0, release v0.7.0

QUICK CLUSTER

K-means clustering of cases into a given number of clusters.

Syntax

QUICK CLUSTER varlist | ALL [/MISSING=[LISTWISE | PAIRWISE | DEFAULT] [INCLUDE]]
              [/FILE='file' | dataset] [/INITIAL=(values)]
              [/CRITERIA=[CLUSTER(2)] [NOINITIAL] [MXITER(10)] [CONVERGE(0)]]
              [/METHOD=KMEANS[(NOUPDATE | UPDATE)] | CLASSIFY]
              [/PRINT=[INITIAL] [CLUSTER] [ID(var)] [DISTANCE] [ANOVA] [NONE]]
              [/OUTFILE='file' | dataset] [/SAVE=[CLUSTER[(name)]] [DISTANCE[(name)]]].
Distances are Euclidean. Initial centers: by default the first k complete cases, then a case
replaces a center when it lies farther from the centers than the two closest centers lie from
each other (NOINITIAL keeps the first k cases); INITIAL gives them cluster by cluster, FILE reads
them from the clustering variables of a data file. KMEANS moves each center to the mean of its
cases after each pass (UPDATE after each case) until no center moves more than CONVERGE times the
smallest distance between the initial centers, or MXITER passes; CLASSIFY keeps the initial
centers. Then every case goes to its nearest center, and the final centers are the means of
their cases. Tables: Initial Cluster Centers, Iteration History, Cluster Membership (CLUSTER,
with ID), Final Cluster Centers, Distances between Final Cluster Centers (DISTANCE), ANOVA and
Number of Cases in each Cluster. LISTWISE (default) leaves out a case missing on any variable;
PAIRWISE uses the variables a case has. SAVE adds the cluster number and the distance to the
classification center, named QCL_1, QCL_2 and so on unless given. OUTFILE writes the final
centers with CLUSTER_. Weighted when a weight variable is set.

Example

QUICK CLUSTER q1 TO q5 /CRITERIA=CLUSTER(3) MXITER(20) /PRINT=INITIAL ANOVA /SAVE=CLUSTER(segment) DISTANCE.

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