VARREDUCE Procedure

REDUCE Statement

REDUCE UNSUPERVISED effects </ reduce-options>;

REDUCE SUPERVISED response-variable <response-variable> = effects </ reduce-options>;

PROC VARREDUCE can be used for either of the following types of variable selection:

For unsupervised variable selection, you specify the effects to be considered in the variable selection process. An effect can be an original variable in the input data table or a variable that is constructed from the original variables.
For supervised variable selection, you specify both the effects and one or more response-variables. A response-variable can be an original variable in the input data table or a variable that is constructed from the original variables. PROC VARREDUCE supports variable selection both in a classification (categorization) context, in which you specify one response-variable, and in a regression context, in which you can specify more than one response-variable.

Table 3 summarizes the reduce-options, which control the number of variables to be selected.

Table 3: reduce-options

Option	Description
AIC	Uses Akaike’s information criterion to perform model selection
AICC	Uses the corrected Akaike’s information criterion to perform model selection
BIC	Uses Schwarz Bayesian information criterion to perform model selection
MAXITER= \| MAXSTEPS=	Specifies the maximum number of steps to take; the number must be greater than or equal to 1
MAXEFFECTS=	Specifies the number of effects to select; the number must be greater than or equal to 1.
VARIANCEEXPLAINED= \| VAREXP=	Specifies the fraction of the total variance to be explained; the value must be between 0 and 1.
MINVARIANCEINCREMENT= \| VARINC=	Specifies the minimum increment of explained variance (in a fraction of the total variance); the value must be between 0 and 1.

The reduce-options determine the number of variables to be selected. You can specify the following reduce-options as stopping criteria for the VARREDUCE procedure. When you specify more than one option, PROC VARREDUCE stops whenever one of the specified options is satisfied, or when the explained variance equals the total variance. In the latter case, the procedure prints the following message in the log: "Early stop: the proportion of the explained variance to the total variance equals 1."

You can specify the following reduce-options only for supervised variable selection:

AIC: stops PROC VARREDUCE if the Akaike’s information criterion (AIC) value fails to decrease in three contiguous steps.
AICC: stops PROC VARREDUCE if the corrected Akaike’s information (AICC) value fails to decrease in three contiguous steps.
BIC: stops PROC VARREDUCE if the Schwarz Bayesian information criterion (BIC) value fails to decrease in three contiguous steps.

You can specify the following reduce-options for both supervised and unsupervised variable selection:

MAXITER=n MAXSTEPS=n

stops PROC VARREDUCE after it runs n steps.

By default, n is the number of input variables.

MAXEFFECTS=n

stops PROC VARREDUCE after n effects have been selected. Because individual levels of one classification variable can be selected in different steps of the variable selection process, selecting n effects might require more steps than are specified by the value of the MAXITER= option.

By default, MAXEFFECTS=200.

VARIANCEEXPLAINED=fraction VAREXP=fraction

stops PROC VARREDUCE when the fraction of the total variance can be explained by the selected variables.

By default, VARIANCEEXPLAINED=0.9.

MINVARIANCEINCREMENT=fraction VARINC=fraction

stops PROC VARREDUCE when the minimum increment of the explained variance is less than fraction of the total variance.

By default, MINVARIANCEINCREMENT=0.

Last updated: September 22, 2026