A PLS-DA (Partial Least Squares Discriminant Analysis) model is a special type of PLS where the Y variables consist of a set of discrete dummy variables. PLS-DA is used for classification of observations into different classes. In the normal case, a dummy variable consists of the two discrete levels "0" and "1".

In Evince, the dummy variables can be created from the classes of a category. When the category is selected in the DataSet, the settings panel (shown below) to the right will show how the dummy variables are set up. Please see the section about categories for more information about dummy variables.

The dummy variables are set to Y in the variable view of the DataSet.

When this is done, it is possible to create a PLS-DA model from the Data Tree. For a PLS-DA model, it is possible to create a prediction table when a test set is used.