Variable Lists (PLS command)
The command line variable lists specify the dependent variables,
any categorical predictors (factors), and any scale predictors (covariates). ALL
and TO
keywords are supported in all variable lists.
Dependent Variables
A list of one or more dependent variables must be the first specification
on the PLS
command.
- Dependent variables can be numeric or string.
- If a dependent variable is specified more than once, only the last specification is honored for the variable.
MLEVEL Keyword
By default, the measurement level recorded in the data dictionary is honored for dependent variables.
- The
MLEVEL
keyword can be specified after the name of any dependent variable to override its dictionary level of measurement. SpecifyN
(nominal),O
(ordinal), orS
(scale). Note that thePLS
procedure does not distinguish between ordinal and nominal dependent variables; it treats both as categorical. - If
MLEVEL
is not specified and no setting is recorded in the data dictionary, then a numeric variable is treated as scale and a string variable is treated as categorical. - A string dependent variable may be defined as ordinal or nominal only.
-
MLEVEL
applies only to the variable that immediately precedes it in the dependent variable list. For example,Age Income [S]
andAge TO Income [S]
assign the scale level of measurement to Income only. The keywordALL
followed by a measurement level specification applies that measurement level to all dependent variables.
REFERENCE Keyword
The REFERENCE
keyword specifies
the value of the dependent variable to use as the reference category
for parameter estimation. REFERENCE
can be specified after the name of any dependent variables but is
honored only for categorical dependent variables. It is silently ignored
for scale variables.
- Specify
FIRST
,LAST
, or a value. If you specifyFIRST
, the lowest value is treated as the reference category.LAST
(the default) treats highest value of the dependent variable as the reference category. - If you specify a value it must exist in the data and be consistent with the type and format of the dependent variable. String and formatted (for example, date or time) values must be quoted.
-
REFERENCE
applies only to the variable that immediately precedes it in the dependent variable list. The keywordALL
followed by aREFERENCE
specification applies that reference category to all categorical dependent variables.
Predictor Variables
- At least one predictor must be specified.
- The names of the categorical predictors, if any, must
be preceded by the keyword
BY
. If keywordBY
is specified with no factors, then a warning is issued andBY
is ignored. - The names of the
covariates, if any, must be preceded by the keyword
WITH
. If keywordWITH
is specified with no covariates, then a warning is issued andWITH
is ignored. - Categorical predictor variables can be numeric or string.
- Covariates must be numeric.
- If the dependent variable is specified within a categorical predictor list or a covariate list, then it is ignored in the list.
- All variables specified within a categorical predictor or covariate list must be unique. If duplicate variables are specified within a list, then the duplicates are ignored.
- If duplicate variables are specified across the predictor lists, then an error is issued.