For the latest version of JMP Help, visit JMP.com/help.


Multivariate Methods > Cluster Variables > Example of the Cluster Variables Platform
Publication date: 04/21/2023

Example of the Cluster Variables Platform

In this example, you cluster continuous baseline variables used in modeling diabetes disease progression.

1. Select Help > Sample Data Folder and open Diabetes.jmp.

2. Select Analyze > Clustering > Cluster Variables.

3. Select the columns Age through Glucose except for Gender (Age, BMI, BP, Total Cholesterol, LDL, HDL, TCH, LTG, and Glucose) and click Y, Columns.

The Gender column cannot be included because Cluster Variables requires numeric continuous variables.

4. Click OK.

Figure 17.2 Cluster Variables Report for Diabetes Data 

Cluster Variables Report for Diabetes Data

The Cluster Summary report shows that the variables were grouped into three clusters:

Cluster 1 consists of TCH, HDL, LTG, and BMI, as shown in the Cluster Members report. The Cluster Summary report shows that TCH is the most representative variable for Cluster 1 and that for the variables in Cluster 1, 62.8% of the variation is explained by the first principal component.

Cluster 2 consists of Total Cholesterol and LDL. The Cluster summary report shows that Total Cholesterol is the most representative variable for Cluster 2 and that for the variables in Cluster 2, 94.8% of the variation is explained by the first principal component.

Cluster 3 consists of BP, Age, and Glucose. The Cluster Summary report shows that the most representative variable is BP and that for the variables in Cluster 3, 56.2% of the variation is explained by the first principal component.

Want more information? Have questions? Get answers in the JMP User Community (community.jmp.com).