Launch the Discriminant Analysis Platform

Launch the Discriminant platform by selecting Analyze > Multivariate Methods > Discriminant.

Figure 5.3 Discriminant Launch Window for Iris.jmp 

For more information about the options in the Select Columns red triangle menu, see Column Filter Menu in Using JMP.

Note: The Validation button appears in JMP Pro only. In JMP, you can define a validation set using excluded rows. See Validation in JMP and JMP Pro.

Y, Covariates

Columns that contain the continuous variables used to classify observations into categories.

X, Categories

A column that contains the categories or groups into which observations are to be classified.

Weight

A column whose values assign a weight to each row for the analysis.

Freq

A column whose values assign a frequency to each row for the analysis. In general terms, the effect of a frequency column is to expand the data table, so that any row with integer frequency k is expanded to k rows. You can specify fractional frequencies.

Validation

A numeric column that defines the validation sets. This column should contain at most three distinct values:

– If there are two values, the smaller value defines the training set and the larger value defines the validation set.

– If there are three values, these values define the training, validation, and test sets in order of increasing size.

– If the validation column has more than three levels, the rows that contain the smallest three values define the validation sets. All other rows are excluded from the analysis.

The Discriminant platform uses the validation column to train and evaluation the model, unless Stepwise Variable Selection is used. If the Stepwise Variable Selection option is selected in the launch, the Discriminant platform uses the validation column to train and tune the model or to train, tune, and evaluate the model. For more information about validation, see Validation in JMP Modeling.

Tip: If Stepwise Variable Selection is not used, the validation column should contain only two distinct values.

If you click the Validation button with no columns selected in the Select Columns list, you can add a validation column to your data table. For more information about the Make Validation Column utility, see Make Validation Column in Predictive and Specialized Modeling.

By

Performs a separate analysis for each level of the specified column.

Stepwise Variable Selection

Performs stepwise variable selection using covariance analysis and p-values. See Stepwise Variable Selection.

If you have specified a validation set, statistics for the validation set also appear. The validation set statistics are used to determine how many steps to take if you use the Go button.

Note: This option is not provided for the Wide Linear discriminant method.

Discriminant Method

Provides four methods for conducting discriminant analysis. See Discriminant Methods.

Shrink Covariances

Shrinks the off-diagonal elements of the pooled within-group covariance matrix and the within-group covariance matrices. This can improve stability and reduce the variance of prediction. See Shrink Covariances.

Advanced Options

Contains the following options:

Uncentered Canonical

Suppresses centering of canonical scores for compatibility with older versions of JMP.

Use Pseudoinverses

Uses Moore-Penrose pseudoinverses in the analysis when the covariance matrix is singular. The resulting scores involve all covariates. If left unchecked, the analysis drops covariates that are linear combinations of covariates that precede them in the list of Y, Covariates.

Cross Validate by Excluded Rows

Specifies that the excluded rows form a validation set for which statistics of fit are calculated.