BUILD_OPTIONS Subcommand (GENLINMIXED command)

The BUILD_OPTIONS subcommand specifies the criteria used to build the model.

TARGET_CATEGORY_ORDER. This determines the order of the target's categories for purposes of determining the "last" category, which is used as the reference category if the REFERENCE keyword is not used. This keyword is ignored if the target is not categorical.

  • ASCENDING. Target categories are sorted in ascending order, from the lowest value to the highest value. The highest value is treated as the "last" category. This is the default.
  • DESCENDING. Target categories are sorted in descending order, from the highest value to the lowest value. The lowest value is treated as the "last" category.
  • DATA. Target values are not sorted. The first value encountered in the data defines the first category, the last value encountered defines the last category. This option is technically not valid if splits are defined on the SPLIT FILE command, but the procedure will run while it does not honor splits.

INPUTS_CATEGORY_ORDER. This determines the order of the categories for factors (categorical inputs) for purposes of determining the "last" category, which is used as the reference category. This keyword is ignored if there are no factors.

  • ASCENDING. Factor values are sorted in ascending order, from the lowest value to the highest value. The highest value is treated as the "last" category. This is the default.
  • DESCENDING. Factor values are sorted in descending order, from the highest value to the lowest value. The lowest value is treated as the "last" category.
  • DATA. Factor values are not sorted. The first value encountered in the data defines the first category, the last value encountered defines the last category. This option is technically not valid if splits are defined on the SPLIT FILE command, but the procedure will run while it does not honor splits.

MAX_ITERATIONS = number. The model building algorithm stops after a certain number of iterations. By default, this is 100. Alternatively, specify a non-negative integer maximum number of iterations. The algorithm uses a doubly iterative process that consists of an inner loop and an outer loop. The value that is specified for the maximum number of iterations applies to both loops.

CONFIDENCE_LEVEL. This is the level of confidence used to compute interval estimates of the model coefficients. Specify a value greater than 0 and less than 100. The default is 95.

DF_METHOD. This specifies how degrees of freedom are computed for significance tests.

  • RESIDUAL. The residual method has a fixed degrees of freedom for all tests. It is useful if your sample size is sufficiently large, or the data are balanced, or the model uses a simpler covariance type; for example, scaled identity or diagonal. This is the default.
  • SATTERTHWAITE. The Satterthwaite method has a field degrees of freedom across tests. It is useful if your sample size is small, or the data are unbalanced, or the model uses a complicated covariance type; for example, unstructured.

COVB. This is the method for computing the parameter estimates covariance matrix.

  • MODEL. The model-based estimate assumes that the model assumptions are correct. This is the default.
  • ROBUST. The robust estimate is useful if you are concerned that the model assumptions are violated.

The model building algorithm uses a doubly iterative process that consists of an inner loop and an outer loop. The following settings apply to the inner loop.

PCONVERGE = number (ABSOLUTE | RELATIVE). Parameter estimates convergence criterion. Convergence is assumed if the maximum ABSOLUTE or maximum RELATIVE change in the parameter estimates is less than the specified value. The criterion is not used if the specified value is 0. Specify a non-negative value and the ABSOLUTE or RELATIVE keyword in parentheses to define the type of convergence. The default value is 1E-6 (ABSOLUTE).
Note: The default setting of PCONVERGE=0.000001(ABSOLUTE) might produce results that differ from the results that are obtained in versions before version 22. To reproduce results from pre-22 versions, use PCONVERGE=0.000001(RELATIVE).

LCONVERGE = number (ABSOLUTE | RELATIVE). Log-likelihood function convergence criterion. Convergence is assumed if the ABSOLUTE or RELATIVE change in the log-likelihood function is less than the specified value. The criterion is not used if the specified value is 0. This criterion is not used by default. Specify a non-negative value and the ABSOLUTE or RELATIVE keyword in parentheses to define the type of convergence.

HCONVERGE = number (ABSOLUTE | RELATIVE). Hessian convergence criterion. Convergence is assumed if g' k H k -1 g k is less than a multiplier of the specified value. The multiplier is 1 for ABSOLUTE type and is the absolute value of the current log-likelihood function for RELATIVE type. The criterion is not used if the specified value is 0. This criterion is not used by default. Specify a non-negative value and the ABSOLUTE or RELATIVE keyword in parentheses to define the type of convergence.

SCORING = integer. Apply scoring algorithm. Specify a non-negative integer. The default is 0, which specifies the Newton-Raphson method. Values greater than 0 specify to use the Fisher scoring algorithm up to iteration number n, where n is the specified integer, and Newton-Raphson thereafter.

SINGULAR = number. Value used as tolerance in checking singularity. Specify a positive value. The default value is 1E-12.