Settings for Which to Construct Learning Curves

Use this feature to select the settings for which to construct learning curves .

For detailed information about the files and data sets used or created by JMP Life Sciences software, see Files and Data Sets .

To Select the Settings:

Select the folder containing all of the predictive modeling settings.

Note : All of the settings to be cross validated must be located in the same folder.

All of the settings files in the selected folder are displayed in the Available Settings field, as shown below:

Highlight specific settings and click

to add the settings to the Settings for Which to Construct Learning Curves field.

A number of buttons enable you to add, remove, and manipulate the selected settings prior to cross validation :

•

To remove one or more settings from the Settings for Which to Construct Learning Curves field, highlight the selected settings in the Settings for Which to Construct Learning Curves field. The

button changes to a

button. Click the

button to remove the selected settings.

•

The

and

buttons enable you to change the order in which the selected processes are run during the analysis.

•

Select any process in the Settings for Which to Construct Learning Curves field and click

and

to move it within the list of processes.

•

The

button opens the dialog of a highlighted process, allowing you to make specific changes to the process settings.

•

The

button can be used to apply parameters specified in a selected setting to all other settings. Examples of parameters that can be set with this option include input data sets, the experimental design data set, and the output folder. Be aware that using this option can cause unintended changes in saved settings files. You should test any settings so modified before running full analyses.

Note : The

button, which is normally located with the other action buttons along the bottom of the dialog, has been moved to facilitate cross validation involving multiple processes, and because it functions somewhat differently in this process.

•

The

button enables you to run the model comparison using a select group of rows from the input data set to quickly validate all the input parameters. The test runs on all highlighted processes in the right panel or the entire set if none are highlighted. The input data set rows are selected using the inputs entered on the Options tab.

•

The

button enables you to update the list of available settings, as they are created in the individual models, without having to close and reopen the model comparison analytical procedure .

•

Click

to remove all of the settings from the Settings for Which to Construct Learning Curves field.

Note : The Input SAS Data Set and Dependent Variable must be identical for all processes specified in the Settings for Which to Construct Learning Curves field.