Cluster Profiles offers an easy way of quickly identifying the key variables which discriminate between clusters.  Having produced a hierarchical cluster analysis, you can select any cluster level simply by pointing to a section through the tree and clicking the mouse.

Clicking Profiles reveals our unique ClustanGraphics Profiles chart.  This example lists the mean percentages of Protein in the milk of the mammals in each cluster.  The chart shows that cluster 4 (Deer) has the highest average protein, with cluster 5 (Seal) second highest and cluster 1 (Orangutan).  Click the chart and it advances to the next variable.  Click Pivot, and the chart changes to show a summary for all the variables by each cluster in turn.

Because we used Reorder Tree to arrange the tree in an optimal order, it's clear from the profile that protein is closely correlated with this optimum case order. This is because the average cluster protein generally increases with the cluster number (1-5), and the clusters are numbered sequentially from the top of the tree.

Click Table to view the full table of cluster means or to copy it into a report, presentation or spreadsheet.  Perhaps you'd like to try clustering by another method - ClustanGraphics has 11 methods of hierarchical cluster analysis.  Click Cluster Prox to see how you can compute a proximity matrix and cluster it for up to 10,000 cases.

Cluster Profiles now allows for missing values in your data matrix.  Cluster means are estimated from the observed values within a cluster, and can themselves be missing in clusters that contain no valid observations on one or more variables.

To order ClustanGraphics on-line click ORDER now.