Name: PCA_illustrativevariable.jsl Author: Florence Kussener, SAS Institute Inc. History: Kussener - 03Jan13 Initial coding Support Contact: Florence.Kussener@jmp.com Copyright (c) 2010, SAS Institute Inc., Cary, NC, USA, All Rights Reserved Description: This script builds a dialog for performing a Principal Component Analysis and display illustrative variables on the Loadings 2D graphic. The correlation matrix between the principal components and illustrative variables is computed and allow to display the relation of new variables on the Loading graph. Instructions: GENERAL TAB: Active Variable, Columns: Select variables over which to compute PCA. Only numerical values are accepted. Illustrative Variable Column: Select numerical variable to add on loading graph. ANALYSIS PLATFORM: PCA with illustrative variables: See the 2D loading graph with illustrative variables displayed in RED. Summary plots: usual summary plot for 2D PCA. See the corresponding PCA platform for further details. KNOWN ISSUES: For the moment, it is only for 2D representation. A sample data set (PCA_Table_Example.jmp) from the SAPORTA's book, is included with this add-in and can be found in the home folder where the add-in was installed. This dataset collect information from different cars and look for correlated variables. CYL, POWER, LENGTH, WIDTH, WEIGHT and SPEED MAX are active variables. PRICE and WEIGHT/POWER are illustrative variables Go to View->Add-Ins... to view the Home Folder for the PCA with illustratives variables Add-in.