Properties Assigned and Controlled by JMP
To control how information flows between platforms, JMP assigns some properties that you cannot control. These properties do not appear on the Column Properties menu.
Response Probability
In Profiler reports, the Response Probability property makes all levels of the categorical response variable appear in a single row. JMP automatically assigns the Response Probability property when certain probability formulas are saved to the data table.
Follow these steps to save the property:
1. Fit a logistic regression model using the Fit Model platform.
2. Click the report red triangle and select Save Probability Formula.
JMP automatically assigns the Response Probability property to the new probability columns.
For more information about the prediction profiler, see Profilers.
Expand Formula
In the Profiler and Nonlinear platforms, if an input column to the model has a formula, JMP substitutes the inner formula, as long as it refers to other columns. To prevent an input column from expanding, assign a value of 0 to the Expand Formula property.
See “Expand Intermediate Formulas” in Profilers.
Expand Intermediates
The Expand Intermediates property is assigned to saved formula columns that refer to intermediate formula columns. This is done in several JMP platforms to avoid repeating some calculations. When formula columns with the Expand Intermediates column property are specified in the Profiler, JMP automatically profiles them using the original factor columns.
Note: The Expand Intermediates column property overrides the Expand Intermediate Formula option in the profiler launch.
See “Expand Intermediate Formulas” in Profilers.
Intermediate
The Intermediate property is assigned to intermediate formula columns that are used to calculate final saved output columns.
Predicting
JMP automatically assigns the Predicting column property when you fit a model to a continuous response and save the prediction formula or prediction values. The Predicting column property identifies the platform used to create the prediction formula. That platform is listed as the Creator in the Model Comparison platform. In many cases, the Predicting column property also includes an estimate of the standard deviation from the platform used to create the prediction formula. This standard deviation estimate is used in the Prediction Profiler platform.
For platforms that also assign a Std Error column property to a saved standard error formula column, the Predicting column property contains an ID clause. The ID number in the Predicting column property matches the ID number in the Std Error column property of the corresponding standard error formula column. These ID numbers are used by the Prediction Profiler platform and are especially useful when there are multiple prediction formula columns and standard error formula columns.
Std Error
JMP automatically assigns the Std Error column property when you fit a model to a continuous response and save a formula column that calculates the standard error of the predicted values. The Std Error column property contains an ID clause and a DF clause. The ID number matches the ID number in the corresponding prediction formula column. These ID numbers are used by the Prediction Profiler platform and are especially useful when there are multiple prediction formula columns and standard error formula columns. The DF value of this column property is used by the Prediction Profiler to draw confidence intervals when both a prediction formula column and a standard error formula column are provided. If the model was fit on a transformed response, the Std Error column property also contains a Transform clause. The Transform clause ensures that the Prediction Profiler applies the correct confidence intervals to the transformed response.
Caution: Removing or changing this property is not recommended. Doing so could adversely affect the confidence intervals in the Prediction Profiler.
Informative Missing Terms
Missing levels of a categorical variable are treated as informative missing in certain JMP platforms, either by default or if you request an informative missing fit. In these cases, JMP automatically assigns the Informative Missing Terms column property to a prediction formula column that includes the categorical variable. This column property ensures that the missing value category is treated as a distinct level of the categorical variable in plots and analyses that involve the prediction formula. In particular, profiler plots show the missing values as a level.
Column ID
JMP automatically assigns the Column ID column property to the column that is used to split the initial Bootstrap results data table. This column property assists the split operation and ensures that the column names in the split Bootstrap results table are meaningful.
ConstraintState
When you select Save Constraints, the coefficients of each linear constraint appear in a column in a data table. Each constraint column is assigned the ConstraintState column property. This property specifies the direction of the inequality that defines the constraint. When you select Load Constraints from a design platform, the ConstraintState column property tells JMP the direction of the inequality.
RunsPerBlock
The RunsPerBlock property indicates the maximum allowable number of runs in each block. This property is used by the Evaluate Design and Augment Design platforms to indicate the blocking structure for the factor.