Data Analysis with IBM SPSS Statistics

K-means cluster analysis of the Old Faithful data

Since charting suggests a two-cluster spherical structure, you could try k-means clustering with k tentatively identified as 2.

Here is SPSS code for k-means analysis of the standardized input variables:

QUICK CLUSTER Zeruption Zwaiting /MISSING=LISTWISE /CRITERIA=CLUSTER(2) MXITER(10) CONVERGE(0) /METHOD=KMEANS(NOUPDATE) /SAVE CLUSTER /PRINT INITIAL ANOVA CLUSTER DISTAN.

Key elements of the SPSS code are:

• The input variables are the standardized forms of eruption and waiting
• /CRITERIA specifies two clusters, as well as default values for iteration and convergence criteria
• /SAVE specifies that cluster membership should be saved in a new variable that will be added to the active file
• /PRINT ...