SAVE Subcommand (PRINCALS command)

SAVE lets you add variables containing the object scores computed by PRINCALS to the active dataset.

  • If SAVE is not specified, object scores are not added to the active dataset.
  • A variable rootname can be specified on the SAVE subcommand to which PRINCALS adds the number of the dimension. Only one rootname can be specified, and it can contain up to six characters.
  • If a rootname is not specified, unique variable names are automatically generated. The variable names are PRI n_m, where n is a dimension number and m is a set number. If three dimensions are saved, the first set of names is PRI1_1, PRI2_1, and PRI3_1. If another PRINCALS is then run, the variable names for the second set are PRI1_2, PRI2_2, PRI3_2, and so on.
  • Following the name, the number of dimensions for which you want to save object scores can be listed in parentheses. The number cannot exceed the value of the DIMENSION subcommand.
  • If the number of dimensions is not specified, the SAVE subcommand saves object scores for all dimensions.
  • If you replace the active dataset by specifying an asterisk (*) on a MATRIX subcommand, the SAVE subcommand is not executed.
  • The prefix should be unique for each PRINCALS command in the same session. If it is not, PRINCALS replaces the prefix with DIM, OBJ, or OBSAVE. If all of these already exist, SAVE is not executed.

Example

PRINCALS CAR1 CAR2 CAR3(5) PRICE (10)
/ANALYSIS=CAR1 TO CAR3(SNOM) PRICE(NUM)
/DIMENSIONS=3
/SAVE=DIM(2).
  • Three nominal variables, CAR1, CAR2, and CAR3, each with five categories, and one numerical (interval level) variable, with ten categories, are analyzed in this PRINCALS example.
  • The DIMENSIONS subcommand requests results for three dimensions.
  • SAVE adds the object scores from the first two dimensions to the active dataset. The names of these new variables will be DIM00001 and DIM00002, respectively.