Example of a Distance Matrix

The proper data table structure for a distance matrix consists of the following:

•

An identifier column (usually the first column) that has a Character data type.

•

A set of n columns, where n is also the number of rows. These n columns define a symmetric matrix with zero or missing values on the diagonal.

Notice that the distance matrix in Flight Distances.jmp follows the preceding format.

Select Help > Sample Data Library and open Flight Distances.jmp.

Select Analyze > Clustering > Hierarchical Cluster.

In the list at the bottom left corner of the launch window, change Data as usual to Data is distance matrix.

Select Cities and click Label.

Select all remaining columns and click Y, Columns.

Figure 10.6 Completed Distance Matrix Launch Window

Click OK.

Click the red triangle next to Hierarchical Clustering and select Color Clusters.

Figure 10.7 Dendrogram Report for Flight Distances

Figure 10.7 shows the Dendrogram report for the flight distances. The placement of the diamonds indicates that the model has grouped the cities into three clusters, which are color-coded on the dendrogram. For more details about how to interpret the report, see Dendrogram Report.