[Environment]
-
Version: JMP(r) Student Edition 19.1.0 (914688)
-
License: For Academic Use Only (Expires: 2027/04/13)
-
OS: Microsoft Windows 11 Home (10.0.26200.0)
[Inquiry] Regarding the procedure for performing PLS Discriminant Analysis (PLS-DA)
I am attempting to perform a PLS Discriminant Analysis (PLS-DA) using JMP Student Edition, as I understand this feature is supported in this edition. I have been following the procedure described in the documentation below:
(Link: https://www.jmp.com/support/help/ja/18.2/index.shtml#page/jmp/statistical-details-for-pls-discrimina...)
[Issue Details]
As shown in JMPpic1.png, I selected the PLS analysis and set the Target Variable (Y) as "BP Occurrence (0/1)". Although I set the modeling type to Nominal (JMPpic2.png) or Ordinal (JMPpic3.png) as required for PLS-DA, the following JMP alert appears:
"The following columns are not continuous. There may be a mistake: BP(1)/non-BP(0). If you are not sure you want this operation, we recommend you cancel."
Even after clicking "Continue," the analysis does not proceed, and I am returned to the data table screen (JMPpic4.png).
[Troubleshooting performed] I have already tried the following steps without success:
-
Re-assigning variables: Cleared all variables in the "Partial Least Squares" launch window and re-assigned the Nominal variable to [Y, Response].
-
Data cleaning: Removed rows containing non-numeric characters (e.g., full-width spaces or "Unknown") to ensure the column only contains "0" and "1".
Note: Standard PLS Regression (with Y and X set to Continuous) runs successfully.
Could you please advise if there are additional steps or specific settings required to perform PLS-DA correctly in my environment?
Thank you for your assistance.
