Checks sets of factors to remove any that define duplicate classifications (R.W. Payne).
No options
Parameters
SET1 = pointers |
First set of factors |
|---|---|
SET2 = pointers |
Second set of factors |
DISTINCTSET = pointers |
Saves the distinct factors |
Description
FDISTINCTFACTORS checks sets of factors to remove any that divide the data units into identical groups. The levels of the factors need not be in the same order – it is the composition of the groups that they define that is important. Also, any null groups (containing no units) are ignored.
The SET1 and SET2 parameters supply pointers containing sets of factors. The DISTINCTSET parameter saves the set of distinct factors (i.e. those that all define different groupings). If only SET1 is set, DISTINCTSET saves the set of factors from SET1 that are distinct from each other. Alternatively, if both SET1 and SET2 are set, DISTINCTSET saves the factors in SET1, plus the factors in SET2 that are distinct from each other and from the factors in SET1. Thus, if SET2 is specified, it is assumed the factors in SET1 are already distinct from each other (so this provides a way of augmenting an already distinct set).
Options: none.
Parameters: SET1, SET2, DISTINCTSET.
Action with RESTRICT
Any restrictions are ignored.
See also
Directives: FACTOR, SETCALCULATE.
Commands for: Calculations and manipulation.
Example
CAPTION 'FDISTINCTFACTORS example'; STYLE=meta FACTOR [NVALUES=36; LEVELS=2] A,B,DA FACTOR [NVALUES=36; LEVELS=3] C,D,DB,DC,DD GENERATE A,B,C,D CALCULATE DA,DB,DC,DD = A,B,C,D FDISTINCTFACTORS !p(B,D); SET2=!p(DA,DB,DC,DD); DISTINCTSET=dfac PRINT dfac FDISTINCTFACTORS !p(A,B,C,D,DA,DB,DC,DD); DISTINCTSET=dfac PRINT dfac