PRINT Subcommand (CLUSTER command)

PRINT controls the display of cluster output (except plots, which are controlled by the PLOT subcommand).

  • If the PRINT subcommand is omitted or included without specifications, an agglomeration schedule is displayed. If any keywords are specified on PRINT, the agglomeration schedule is displayed only if explicitly requested.
  • CLUSTER automatically displays summary information (the method and measure used, the number of cases) for each method named on the METHOD subcommand. This summary is displayed regardless of specifications on PRINT.

You can specify any or all of the following on the PRINT subcommand:

SCHEDULE. Agglomeration schedule. The agglomeration schedule shows the order and distances at which items and clusters combine to form new clusters. It also shows the cluster level at which an item joins a cluster. SCHEDULE is the default and can also be requested with the keyword DEFAULT.

CLUSTER(min,max). Cluster membership. For each item, the display includes the value of the case identifier (or the variable name if matrix input is used), the case sequence number, and a value (1, 2, 3, and so on) identifying the cluster to which that case belongs in a given cluster solution. Specify either a single integer value in parentheses indicating the level of a single solution or a minimum value and a maximum value indicating a range of solutions for which display is desired. If the number of clusters specified exceeds the number produced, the largest number of clusters is used (the number of items minus 1). If CLUSTER is specified more than once, the last specification is used.

DISTANCE. Proximities matrix. The proximities matrix table displays the distances or similarities between items computed by CLUSTER or obtained from an input matrix. DISTANCE produces a large volume of output and uses significant CPU time when the number of cases is large.

NONE. None of the above. NONE overrides any other keywords specified on PRINT.

Example

CLUSTER V1 V2 V3 /PRINT=CLUSTER(3,5).
  • This example displays cluster membership for each case for the three-, four-, and five-cluster solutions.