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


Fitting Linear Models > Stepwise Regression Models > The All Possible Models Option
Publication date: 05/05/2023

The All Possible Models Option

In the Stepwise personality of the Fit Model platform, use the All Possible Models option to investigate all models that can be constructed using your predictors. This option is accessed from the Stepwise red triangle menu.

Note the following:

This option is not practical for large problems, when the number of models is greater than 5 million.

Categorical predictors are represented by indicator variables. See Models with Nominal and Ordinal Effects.

The following options restrict the number of models that appear:

Maximum number of terms in a model

Enter a value for the maximum number of terms in a model.

Number of best models to see

Enter the maximum number of models of each size to display. The best models according to RSquare value appear.

Restrict to models where interactions imply lower order effects (Heredity Restriction)

Shows only models that contain all lower-order effects when a higher-order effect is included. These models satisfy strong effect heredity. This option is useful when your predictors include interaction or polynomial terms.

See Example of the All Possible Models Option.

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