Select Dependent Variables and Predictors - Advanced Tab

Select the Advanced tab of the Select Dependent Variables and Predictors dialog box to access options to specify special purpose variables, example, variables that are applicable to particular types of analyses only, and that are ignored by many (or most) types of analyses available in Statistica. It is recommended that you carefully review the documentation (overviews, examples) for the types of advanced analyses that might use these variables to make sure that your selections are applied as you intended.
Option Description
Counts Variable Displays a variable selection dialog box that contains options to select the variable with frequency counts. A counts variable is used in some advanced methods for nonlinear estimation, to handle aggregated data. For example, in the Generalized Linear Models module, you can specify a counts variable for certain models.
Censoring indicator Displays a standard variable selection dialog box that contains options to select the variable with the censoring codes, that is, codes that identify whether a respective observation is complete or censored. After selecting the censoring variable, you usually also must select the specific codes that were used in that variable to identify the cases that are to be treated as complete or censored in subsequent analyses; default (random) selection of codes in this instance could lead to highly misleading results. The Censoring variable is used in survival and failure time analyses, as well as reliability (Weibull) analyses.
Learning/testing indicator Displays a standard variable selection dialog box that contains options to select a variable with codes that identify the observations that are to be assigned to a learning or training sample, or to a testing or validation sample. Learning and testing samples are typically used in complex analytic procedures for predictive data mining, to avoid over-fitting or over-learning. Many of those procedures randomly select observations into a learning and testing sample if none are specified for the data source.

After selecting a learning/testing variable, select the codes in the data file that are to be used to identify the learning and testing sample cases for subsequent analyses.