Launch Windows

The launch window is your point of entry into a platform. Launch Window Panels describes three panels that all launch windows have in common.

Launch Window Panels

Select Columns

Lists all of the variables in your current data table. Note the following:

•	Right-click on a column name to change the modeling type.

•	Filter the columns using the options in the red triangle menu. See Columns Filter Menu.

Cast Selected Columns into Roles

Moves selected columns into roles (such as Y, X, and so on.) You cast a column into the role of a variable (like an actor is cast into a role). See Cast Selected Columns into Roles Buttons.

This panel does not exist in the Graph Builder platform.

Action

performs the analysis.

Cancel

stops the analysis and quits the launch window.

Remove

deletes any selected variables from a role.

Recall

populates the launch window with the last analysis that you performed.

Help

takes you to the Help for the launch window.

Cast Selected Columns into Roles Buttons

Descriptions of Role Buttons describes buttons that appear frequently throughout launch windows. Buttons that are specific to certain platforms are described in the chapter for the specific platform.

Descriptions of Role Buttons

Identifies a column as a response or dependent variable whose distribution is to be studied.

Identifies a column as an independent, classification, or explanatory variable that predicts the distribution of the Y variable.

Weight

Identifies the data table column whose variables assign weight (such as importance or influence) to the data.

Freq

Identifies the data table column whose values assign a frequency to each row. This option is useful when a frequency is assigned to each row in summarized data. If the value is 0 or a positive integer, then the value represents the frequencies or counts of observations for each row when there are multiple units recorded.

Validation

Identifies the data table column whose values assign rows to training, validation, and test sets for crossvalidation in fitting models.

Note the following:

•	For some platforms, KFold validation is available if you specify more than three levels in the Validation column.

•	If you click the Validation button with no columns selected in the Select Columns list, you have the option to create a Validation column. For more information about the Make Validation Column utility, see Basic Analysis.

Identifies a column that creates a report consisting of separate analyses for each level of the variable.

Columns Filter Menu

A Column Filter menu appears in most of the launch windows. The Column Filter menu is a red triangle within the Select Columns panel. Use these options to sort columns, show or hide columns, or search columns.

The Columns Filter Menu


Reset	Resets the columns to its original list.
Sort by Name	Sorts the columns in alphabetical order by name.
Continuous	Shows or hides columns whose modeling type is continuous.
Ordinal	Shows or hides columns whose modeling type is ordinal.
Nominal	Shows or hides columns whose modeling type is nominal.
Numeric	Shows or hides columns whose data type is numeric.
Character	Shows or hides columns whose data type is character.
Match case	(Only applicable to the Name options below) Makes your search case-sensitive.
Name Contains	Searches for column names containing specified text. To remove the text box, select Reset.
Name Does Not Contain	Searches for column names that do not contain specified text. To remove the text box, select Reset.
Name Starts With	Searches for column names that begin with specified text. To remove the text box, select Reset.
Name Ends With	Searches for column names that end with specified text. To remove the text box, select Reset.
Exclude Formats	Excludes columns with specific formats from the column selection list. Select from the following formats: date, time, duration, geographic, or all numeric formats.
Column Groups	Shows or hides groups of columns. See Group Columns in Enter and Edit Data.
Ungrouped Columns	Shows or hides columns that have not been grouped.