Common Features among Nodes |
Most Enterprise Miner nodes have a Variables tab. Typically, the Variables tab displays variable attribute information, such as
the names of the variables in the input data set. The values that are displayed in the Name column of most Variable tabs are protected.
the status of the variables, which can be use or don't use. You can change the status of a variable in one of the sampling or modeling nodes. For most explore nodes, you cannot change the variable's status. To change the status of a variable, follow these steps:
Right-click on the appropriate cell of the Status column.
Select Set Status.
Select the appropriate status (either don't use or use).
Note: The Variables tab of the Input Data Source node does not contain a Status column. To assign the rejected model role to a variable, right-click on the appropriate cell of the Model Role column to open a pop-up menu. Select Set Model Role and then select rejected. You use this same protocol to assign the rejected role to variables in the Manual Selection tab of the Variable Selection node.
This example shows the Variables tab of the Data Partition node:
In this node, the Variables tab contains the following columns:
Name -- the variable name.
Model Role -- the variable model role. Examples of model roles include cost, freq, ID, input, predict, rejected, sequence, target, and trial. The role of a variable is automatically assigned in the Input Data Source node based on the information in the metadata sample. You can change the role that is automatically assigned to a variable by right-clicking on the appropriate role cell.
Measurement -- the measurement type of the variable. Examples of measurement types include binary, interval, ordinal, and nominal. The measurement type of variables is automatically assigned in the Input Data Source node based on the information in the metadata sample.
Type -- either character (char) or numeric (num). The variable type is displayed in the Input Data Source node, but it cannot be changed (the values in this column are protected).
Format -- the format of the variable. Examples of formats include $18. (18 characters) and BEST12. (12-digit numeric). The variable formats are assigned in the Input Data Source node.
Label -- the descriptive variable label, which is assigned in the Input Data Source node. To change the label for a variable, type a new value in the appropriate cell of the Label column.
You can sort by clicking the column header. Successive clicks on the header toggle the sort between ascending and descending order.
Subsetting (displaying a subset of the variables) can be performed for the variable attributes of status, model role, measurement, and type:
Right-click in any row of the attribute that you want to subset.
Select Subset by <attribute> from the menu. A selection list box appears that contains the set of attribute values that are available for subsetting.
Select the appropriate attribute values for subsetting.
To accept the selections and display the subsetting, click
. To cancel the subsetting dialog box, click .To find a value within a column:
Right-click on the column heading or a cell within the column and select Find column name. The Find column name window opens, which is modeled after the Find dialog box in Internet Explorer.
Type the text that you want to find.
Specify the search direction (up or down) and whether you want the text to match the whole field.
Click
.
Copyright © 2006 by SAS Institute Inc., Cary, NC, USA. All rights reserved.