Workspace Node: Lasso Regression - Specifications - Quick Tab

The Quick tab within the Specifications group of the Lasso Regression workspace node dialog box is displayed by default.

Option Description
Variables Click the Variables button to display a standard variable selection dialog box. Select one dependent variable and two or more independent variables. The independent variables can be categorical or continuous, or a combination of both.
Algorithm The following are the types of algorithms. Choose either the Linear Regression or Logistic Regression algorithm.
Alpha Specify the value of the mixing parameter in the penalty term. The valid range of values are 1 for Lasso penalty, 0 for ridge penalty and (0, 1) for elastic‐net penalty.
Number of lambda Specify the number of lambda values for the coordinate descent.
Lambda ratio Specify the minimum value for lambda as a fraction of maximum. If the number of variables in the analysis exceeds the number of observations, a value of 0.01 is prescribed.
Response codes Click the Response codes button to display the Specify Codes dialog box, which contains options to select the codes identifying the levels or categories for the categorical (binary) dependent (response) variables. Codes must be integer values or text labels, they can be dates, times, and two codes must be specified. This button is only available if the user chooses Logistic Regression as the Algorithm.
Options / C / W For more information on Options/C/W, see to "Common Options" in Statistica Electronic Manual.
OK Click the OK button to accept all the specifications made in the dialog box and to close it. The analysis results are placed in the Reporting Documents node after running or updating the project.
Cancel Click the Cancel button to close the Lasso Regression dialog box without making any changes to the current specifications.
Note: Statistica ignores all cases that have missing data for any of the variables selected in the list.
Note: All cases with weight less than or equal to zero will be treated as missing data.