The GAMSELECT Procedure

PROC GAMSELECT Statement

PROC GAMSELECT <options>;

The PROC GAMSELECT statement invokes the procedure. Table 2 summarizes the available options in the PROC GAMSELECT statement by function. The options are then described fully in alphabetical order.

Table 2: PROC GAMSELECT Statement Options

Option	Description
Basic Options
DATA=	Specifies the CAS input data table
SEED=	Sets the seed for pseudorandom number generation
Display Options
ITDETAILS	Displays the "Iteration Details" table
NOCLPRINT	Limits or suppresses the display of classification variable levels
PLOTS=	Controls plots that are produced through ODS Graphics
Tolerance Options
SINGCHOL=	Tunes the singularity criterion for Cholesky decompositions
SINGULAR=	Tunes the general singularity criterion

You can specify the following options in the PROC GAMSELECT statement.

DATA=CAS-libref.data-table

names the input data table for PROC GAMSELECT to use. The default is the most recently created data table. CAS-libref.data-table is a two-level name, where

CAS-libref: refers to a collection of information that is defined in the LIBNAME statement and includes the caslib, which includes a path to the data, and a session identifier, which defaults to the active session but which can be explicitly defined in the LIBNAME statement. For more information about CAS-libref, see the section Using CAS Sessions and CAS Engine Librefs.
data-table: specifies the name of the input data table.

ITDETAILS

generates the "Iteration Details" table.

NOCLPRINT<=number>

suppresses the display of the "Class Level Information" table if you do not specify number. If you specify number, the values of the classification variables are displayed for only those variables whose number of levels is less than number. Specifying a number helps reduce the size of the "Class Level Information" table if some classification variables have a large number of levels.

PLOTS <(global-plot-option)> <= plot-requests <(option)>>

controls the plots that are produced through ODS Graphics. When ODS Graphics is enabled, PROC GAMSELECT produces by default a panel of plots of partial prediction curves or surfaces of smoothing components. For example:

ods graphics on;

proc gamselect plots;
   model y=spline(x1) spline(x2);
   selection;
run;

ods graphics off;

You can specify the following global-plot-option, which applies to the smoothing component plots that the GAMSELECT procedure generates:

UNPACK UNPACKPANEL: suppresses paneling. By default, multiple smoothing component plots can appear in some output panels. Specify this option to display each plot individually.

You can specify the following plot-requests:

ALL

produces all default plots.

COMPONENTS <(component-option)>

plots a panel of smoothing components of the fitted model. You can specify the following component-option:

COMMONAXES: requests that the smoothing component plots, except for bivariate contour plots, use a common vertical axis. This option enables you to visually judge the relative effect size.

NONE

suppresses all plots.

SEED=number

specifies an integer to use to start the pseudorandom number generator for subset sampling from observations to form knots if necessary, for random cross validation if the boosting method is used for model selection, and for random data partitioning if a separate random number seed is not supplied. If you do not specify this option or if number 0, the seed is generated by reading the time of day from the computer’s clock. If you specify this option with a nonzero value and you also specify the PARTITION statement without specifying the random number seed there, then the same random number seed is used to partition the input data. Note that the random sampling of observations for cross validation and data partitioning depends on how the data are distributed across machines and threads. Therefore, even if you specify the same seed, results might differ between runs of PROC GAMSELECT.

SINGCHOL=number

tunes the singularity criterion in Cholesky decomposition and matrix inversion operations. The default number is 1E4 times the machine epsilon; this product is approximately 1E–12 on most computers.

SINGULAR=number

tunes the singularity criterion that is used to solve ridge regressions in searching upper bounds of the sparsity parameter and the smoothness parameter, if you specify the shrinkage method of model selection. The default number is 1E4 times the machine epsilon; this product is approximately 1E–12 on most computers.

Last updated: April 08, 2021