The TRANSREG Procedure

PROC TRANSREG Statement

PROC TRANSREG <DATA= SAS-data-set><PLOTS= (plot-requests)><OUTTEST= SAS-data-set> <a-options > <o-options >;

The PROC TRANSREG statement invokes the TRANSREG procedure. Optionally, this statement identifies an input and an OUTTEST= data set, specifies the algorithm and other computational details, requests displayed output, and controls the contents of the OUT= data set (which is created with the OUTPUT statement). The DATA=, OUTTEST=, and PLOTS= options can appear only in the PROC TRANSREG statement. Table 117.1 summarizes the options available in the PROC TRANSREG statement. The a-options are also available in the MODEL statement, and the o-options are also available in the OUTPUT statement.

Table 117.1: Options Available in the PROC TRANSREG Statement

Option	Description
Data Set Options (PROC Statement)
DATA=	Specifies input SAS data set
OUTTEST=	Specifies output test statistics data set
ODS Graphics (PROC Statement)
PLOTS=	Specifies ODS Graphics selection
Input Control (PROC or MODEL)
REITERATE	Restarts the iterations
TYPE=	Specifies input observation type
Method and Iterations (PROC or MODEL)
CCONVERGE=	Specifies minimum criterion change
CONVERGE=	Specifies minimum data change
MAXITER=	Specifies maximum number of iterations
METHOD=	Specifies iterative algorithm
NCAN=	Specifies number of canonical variables
NSR	Specifies no restrictions on smoothing models
SINGULAR=	Specifies singularity criterion
SOLVE	Attempts direct solution instead of iteration
Missing Data Handling (PROC or MODEL)
INDIVIDUAL	Fits each model individually (METHOD=MORALS)
MONOTONE=	Includes monotone special missing values
NOMISS	Excludes observations with missing values
UNTIE=	Unties special missing values
Intercept and CLASS Variables (PROC or MODEL)
CPREFIX=	Specifies CLASS coded variable name prefix
LPREFIX=	Specifies CLASS coded variable label prefix
NOINT	Specifies no intercept or centering
ORDER=	Specifies order of CLASS variable levels
REFERENCE=	Controls output of reference levels
SEPARATORS=	Controls CLASS coded variable label separators
Control Displayed Output (PROC or MODEL)
ALPHA=	Specifies confidence limits alpha
CL	Displays parameter estimate confidence limits
DETAIL	Displays model specification details
HISTORY	Displays iteration histories
NOPRINT	Suppresses displayed output
PBOXCOXTABLE	Prints the Box-Cox log likelihood table
RSQUARE	Displays the R square
SHORT	Suppresses the iteration histories
SS2	Displays regression results
TEST	Displays ANOVA table
TSUFFIX=	Shortens transformed variable labels
UTILITIES	Displays conjoint part-worth utilities
Standardization (PROC or MODEL)
ADDITIVE	Fits additive model
NOZEROCONSTANT	Does not zero constant variables
TSTANDARD=	Specifies transformation standardization
Predicted Values, Residuals, Scores (PROC or OUTPUT)
CANONICAL	Outputs canonical scores
CLI	Outputs individual confidence limits
CLM	Outputs mean confidence limits
DESIGN=	Specifies design matrix coding
DREPLACE	Replaces dependent variables
IREPLACE	Replaces independent variables
LEVERAGE	Outputs leverage
NORESTOREMISSING	Does not restore missing values
NOSCORES	Suppresses output of scores
PREDICTED	Outputs predicted values
REDUNDANCY=	Outputs redundancy variables
REPLACE	Replaces all variables
RESIDUALS	Outputs residuals
Output Data Set Coefficients (PROC or OUTPUT)
COEFFICIENTS	Outputs coefficients
COORDINATES=	Outputs ideal point coordinates
MEANS	Outputs marginal means
MREDUNDANCY	Outputs redundancy analysis coefficients
Output Data Set Variable Name Prefixes (PROC or OUTPUT)
ADPREFIX=	Specifies dependent variable approximations
AIPREFIX=	Specifies independent variable approximations
CDPREFIX=	Specifies canonical dependent variables
CILPREFIX=	Specifies conservative individual lower CL
CIPREFIX=	Specifies canonical independent variables
CIUPREFIX=	Specifies conservative-individual-upper CL
CMLPREFIX=	Specifies conservative-mean-lower CL
CMUPREFIX=	Specifies conservative-mean-upper CL
DEPENDENT=	Specifies METHOD=MORALS untransformed dependent
LILPREFIX=	Specifies liberal-individual-lower CL
LIUPREFIX=	Specifies liberal-individual-upper CL
LMLPREFIX=	Specifies liberal-mean-lower CL
LMUPREFIX=	Specifies liberal-mean-upper CL
PPREFIX=	Specifies predicted values
RDPREFIX=	Specifies residuals
RPREFIX=	Specifies redundancy variables
TDPREFIX=	Specifies transformed dependents
TIPREFIX=	Specifies transformed independents
Macros Variables (PROC or OUTPUT)
MACRO	Creates macro variables
Other Options (PROC or OUTPUT)
APPROXIMATIONS	Outputs dependent and independent approximations
CCC	Outputs canonical correlation coefficients
CEC	Outputs canonical elliptical point coordinates
CPC	Outputs canonical point coordinates
CQC	Outputs canonical quadratic point coordinates
DAPPROXIMATIONS	Outputs approximations to transformed dependents
IAPPROXIMATIONS	Outputs approximations to transformed independents
MEC	Outputs elliptical point coordinates
MPC	Outputs point coordinates
MQC	Outputs quadratic point coordinates
MRC	Outputs multiple regression coefficients

DATA=SAS-data-set

specifies the SAS data set to be analyzed. If you do not specify the DATA= option, PROC TRANSREG uses the most recently created SAS data set. The data set must be an ordinary SAS data set; it cannot be a special TYPE= data set.

OUTTEST=SAS-data-set

specifies an output data set to contain hypothesis tests results. When you specify the OUTTEST= option, the data set contains ANOVA results. When you specify the SS2 a-option, regression tables are also output. When you specify the UTILITIES o-option, conjoint analysis part-worth utilities are also output. For more information about the OUTTEST= data set, see the section OUTTEST= Output Data Set.

PLOTS <(global-plot-options)> <= plot-request <(options)>> PLOTS <(global-plot-options)> <= (plot-request <(options)> <... plot-request <(options)>>)>

controls the plots produced through ODS Graphics. When you specify only one plot request, you can omit the parentheses around the plot request. Here are some examples:

plots=none
plots=(residuals transformation)
plots(unpack)=boxcox
plots(unpack)=(transformation boxcox(p=0))
plots=(residuals(unpack) transformation(dep unp) boxcox(t rmse))

ODS Graphics must be enabled before plots can be requested. For example:

ods graphics on;

proc transreg plots=all;
   model identity(y) = pbspline(x);
run;

ods graphics off;

For more information about enabling and disabling ODS Graphics, see the section Enabling and Disabling ODS Graphics in Chapter 21: Statistical Graphics Using ODS.

If ODS Graphics is enabled, but you do not specify the PLOTS= option, then PROC TRANSREG produces a default set of plots. The fit, scatter, residual, and observed-by-predicted plots are available with METHOD= MORALS and also with METHOD=UNIVARIATE when there is only one dependent variable. When no method is specified and there is more than one dependent variable, and when regression plots are requested, the default method is set to METHOD=MORALS. When there is more than one dependent variable, when METHOD= is not specified, or when METHOD=MORALS is specified and PLOTS=ALL is specified, the plots that are produced might be different from those you would see with METHOD=UNIVARIATE and PLOTS=ALL. Certain plots appear by default when ODS Graphics is enabled and certain combinations of options are specified. The Box-Cox $F = t^2$ and log-likelihood plots appear when a BOXCOX dependent variable transformation is specified. The regression fit plot appears for models with a single dependent variable that is not transformed (for example, IDENTITY (y)), a single quantitative independent variable that might or might not be transformed, and at most one CLASS independent variable. Preference mapping plots appear when the COORDINATES o-option is used.

The global plot options include the following:

INTERPOLATE INT: uses observations that are excluded from the analysis for interpolation in the fit and transformation plots. By default, observations with zero weight are excluded from all plots. These include observations with a zero, negative, or missing weight or frequency and observations excluded due to missing and invalid values. You can specify PLOTS(INTERPOLATE)=(plot-requests) to include some of these observations in the plots. You might want to use this option, for example, with sparse data sets to show smoother functions over the range of the data (see the section The PLOTS(INTERPOLATE) Option). Observations with missing values in CLASS variables are excluded from the plots even when PLOTS(INTERPOLATE) is specified.
ONLY ONL: suppresses the default plots. Only plots specifically requested are displayed.
UNPACKPANEL UNPACK UNP: suppresses paneling. By default, multiple plots can appear in some output panels. Specify UNPACKPANEL to get each plot in a separate panel. You can specify PLOTS(UNPACKPANEL) to unpack the default plots. You can also specify UNPACKPANEL as a suboption with TRANSFORMATION, RESIDUALS, PBSPLINE, and BOXCOX.

The plot requests include the following:

ALL

produces all appropriate plots. You can specify other options with ALL; for example, to request all plots and unpack only the residuals, specify PLOTS=(ALL RES(UNP)).

BOXCOX <(options )> BOX <(options )>

requests a display of the results of the Box-Cox transformation. These results are displayed by default when there is a Box-Cox transformation. The BOXCOX plot request has the following options:

P=n: adds t or $F = t^2$ curves to the legend for the functions where $p(t) < n$ , where t is the t statistic corresponding to the optimal lambda. You can specify P=0 to suppress the legend and P=1 to see all curves in the legend. The default value comes from the BOXCOX(variable / ALPHA=p) specification, which by default is 0.05.
RMSE RMS: plots the root mean square error as a function of lambda.
T: plots t statistics rather than $F = t^2$ statistics.
UNPACKPANEL UNPACK UNP: plots the t or $F = t^2$ and log-likelihood plots in separate panels.

FIT <(options )>

requests a regression fit plot. This plot is produced by default whenever it is appropriate. It is produced when the dependent variable is specified with the IDENTITY transform, and when there is one quantitative independent variable (for example, IDENTITY for linear fit, SPLINE or one of the other transformations for a nonlinear fit, or PSPLINE) and at most one CLASS variable. When there is a CLASS variable, separate fits are produced within levels based on your model. You would specify the FIT plot request only to specify a FIT option or with the ONLY global plot option. The FIT plot request has the following options:

FORMULA FOR: displays the fit function as an equation in regression fit plots. This option is valid when a fit plot is produced and either an IDENTITY transform or a PSPLINE expansion with degree less than ten and no knots is specified for a single independent variable. When this option is specified, you can output the formula to a data set by using the ods output formula=SAS-data-set statement. This is the formula, complete with Unicode specifications for polynomials, that is used in the fit plot template to make the formula.
NOCLM: suppresses the confidence limits in regression fit plots.
NOCLI: suppresses the individual prediction limits in regression fit plots.
NOOBS: suppresses the observations showing only the fit function and optionally the confidence and prediction limits.

NONE

suppresses all plots.

OBSERVEDBYPREDICTED OBP OBS

plots the transformed dependent variable as a function of the regression predicted values.

PBSPLINE <(UNPACKPANEL)> PBS <(UNPACK)>

requests the penalized B-spline criterion plots. You would specify the PBSPLINE plot request only to specify a PBSPLINE option or with the ONLY global plot option. The PBSPLINE plot request has the following option:

UNPACKPANEL UNPACK UNP: plots each criterion plot in a separate panel.

PREFMAP PRE

plots ideal point or vector preference mapping results when either two IDENTITY or two POINT independent variables are specified along with the COORDINATES option.

RESIDUALS <(options)> RES <(options)>

plots the residuals as a function of each of the transformed independent variables, except coded CLASS variables. The RESIDUALS plot request has the following options:

CLASS CLA: plots the residuals as a function of each of the transformed independent variables, including coded CLASS variables. Note that the ALL plot request, which you use to request all plots, specifies the RESIDUALS plot request without the CLASS option.
UNPACKPANEL UNPACK UNP: plots the residuals in separate plots, not several per panel.
SMOOTH SMO: adds a LOESS smooth function to the residuals plots.

SCATTER SCA

plots the scatter plot of observed data, before the transformations, for models with a single quantitative dependent variable, a single quantitative independent variable, and at most one CLASS independent variable.

TRANSFORMATION <(options)> TRA <(options)>

plots the variable transformations. The TRANSFORMATION plot request has the following options:

DEPENDENTS DEP: plots only the dependent variable transformations.
INDEPENDENTS IND: plots only the independent variable transformations.
UNPACKPANEL UNPACK UNP: plots the transformations in separate plots, not several per panel.