SEVSELECT Procedure
Functional Summary
Table 1 summarizes the statements and options that control the SEVSELECT procedure.
Table 1: Functional Summary
| Description | Statement | Option |
|---|---|---|
| Statements | ||
| Specifies BY-group processing | BY | |
| Specifies the response variable to model along with censoring and truncation effects | LOSS | |
| Specifies the weight variable | WEIGHT | |
| Specifies distributions to fit | DIST | |
| Specifies the classification variables | CLASS | |
| Specifies the constructed regression effects | EFFECT | |
| Specifies the regression effects to model | SCALEMODEL | |
| Specifies the options for selecting the best subset of regression effects | SELECTION | |
| Specifies the ODS tables to display (an alternative to the PRINT= option) | DISPLAY | |
| Specifies the ODS tables to save as output data tables | DISPLAYOUT | |
| Specifies linear restrictions on parameters | RESTRICT | |
| Specifies the scoring functions and quantiles to write | OUTPUT | |
| Specifies the library to write scoring functions to | OUTSCORELIB | |
| Specifies the optimization options | NLOPTIONS | |
| Specifies programming statements that define an objective function | Programming statements | |
| Input and Output Options | ||
| Specifies the input data table | PROC SEVSELECT | DATA= |
| Specifies the input data table for parameter estimates | PROC SEVSELECT | INEST= |
| Specifies the output data table for estimates of scoring functions and quantiles | OUTPUT | OUT= |
| Specifies the output data table for parameter estimates | PROC SEVSELECT | OUTEST= |
| Specifies the output data table for model information | PROC SEVSELECT | OUTMODELINFO= |
| Specifies the output data table for statistics of fit | PROC SEVSELECT | OUTSTAT= |
| Specifies the output item store for context and estimation results | PROC SEVSELECT | OUTSTORE= |
| Data Interpretation Options | ||
| Specifies left-censoring | LOSS | LEFTCENSORED= |
| Specifies left-truncation | LOSS | LEFTTRUNCATED= |
| Specifies the probability of observability | LOSS | PROBOBSERVED= |
| Specifies right-censoring | LOSS | RIGHTCENSORED= |
| Specifies right-truncation | LOSS | RIGHTTRUNCATED= |
| Distribution Processing Options | ||
| Specifies the input data table to read the distribution definitions from | DIST | INFUNCDEF= |
| Specifies that distributions be listed in the log without estimating any models that use them | DIST | LISTONLY |
| Specifies the input data table to write the distribution definitions to | DIST | OUTFUNCDEF= |
| Specifies that distributions be validated without estimating any models that use them | DIST | VALIDATEONLY |
| Model Estimation Options | ||
| Specifies the method for computing mixture distribution | SCALEMODEL | DFMIXTURE= |
| Specifies that informative missingness be used to model missing values of regressors | SCALEMODEL | INFORMATIVE |
| Specifies initial values for model parameters | DIST | INIT= |
| Specifies the sample to be used for computing initial parameter estimates | PROC SEVSELECT | INITSAMPLE |
| Specifies that observations do not need to be balanced among threads of computation | PROC SEVSELECT | NOBALANCETHREADS |
| Specifies role of constant distribution parameters in fit statistic calculations | PROC SEVSELECT | NOCONSTFITSTATS |
| Specifies the objective function symbol | PROC SEVSELECT | OBJECTIVE= |
| Specifies the offset variable in the scale regression model | SCALEMODEL | OFFSET= |
| Specifies the denominator for computing covariance estimates | PROC SEVSELECT | VARDEF= |
| Regression Effect Selection Options | ||
| Specifies whether and how the model hierarchy requirement is applied | SELECTION | HIERARCHY= |
| Specifies the set of regression effects to be forced in all models | SCALEMODEL | INCLUDE= |
| Specifies the selection method | SELECTION | METHOD= |
| Specifies the selection order for displaying effects in the selected model | SELECTION | ORDERSELECT |
| Specifies the optimization technique to use for the intermediate selection steps | PROC SEVSELECT | SELECTNLOTECH= |
| Specifies the set of regression effects to begin the selection process | SCALEMODEL | START= |
| Specifies how to apply the STOP= criterion | SELECTION | STOPHORIZON= |
| Regression Effect Selection Method Options | ||
| Specifies a criterion for choosing the best model at each step | SELECTION | CHOOSE= |
| Specifies that the competitive form of the stepwise selection method be used | SELECTION | COMPETITIVE |
| Specifies the maximum number of effects in the model | SELECTION | MAXEFFECTS= |
| Specifies the maximum number of selection steps | SELECTION | MAXSTEPS= |
| Specifies the minimum number of effects in the model | SELECTION | MINEFFECTS= |
| Specifies a criterion to decide which effects enter or leave the model at each step | SELECTION | SELECT= |
| Specifies a criterion for stopping the selection process | SELECTION | STOP= |
| Empirical Distribution Function (EDF) Estimation Options | ||
| Specifies the nonparametric method of CDF estimation | PROC SEVSELECT | EMPIRICALCDF= |
| EMPIRICALCDF=MODIFIEDKM Options | ||
Specifies the  value for the lower bound on risk set size | PROC SEVSELECT | ALPHA= |
| Specifies the c value for the lower bound on risk set size | PROC SEVSELECT | C= |
| Specifies the absolute lower bound on risk set size | PROC SEVSELECT | RSLB= |
| EMPIRICALCDF=TURNBULL Options | ||
| Specifies that the final EDF estimates be maximum likelihood estimates | PROC SEVSELECT | ENSUREMLE |
| Specifies the relative convergence criterion | PROC SEVSELECT | EPS= |
| Specifies the maximum number of iterations | PROC SEVSELECT | MAXITER= |
| Specifies the threshold below which an EDF estimate is deemed to be 0 | PROC SEVSELECT | ZEROPROB= |
| OUT= Data Table Generation Options | ||
| Specifies the variables to copy from the DATA= data table to the OUT= data table | OUTPUT | COPYVARS= |
| Specifies the scoring functions to estimate | OUTPUT | FUNCTIONS= |
| Specifies the quantiles to estimate | OUTPUT | QUANTILES= |
| OUTEST= Data Table Generation Options | ||
| Specifies that the OUTEST= data table contain covariance estimates | PROC SEVSELECT | COVOUT |
| Specifies that only the selected regression parameters be written | PROC SEVSELECT | SELECTOUT |
| Specifies that estimates of parameters not in the final model be 0 | PROC SEVSELECT | ZEROEST |
| Scoring Function Generation Options | ||
| Writes scoring functions of all models to one package | OUTSCORELIB | COMMONPACKAGE |
| Specifies the output data table for BY-group identifiers | OUTSCORELIB | OUTBYID= |
| Specifies the output library for scoring functions | OUTSCORELIB | OUTLIB= |
| Displayed Output Options | ||
| Specifies the criterion to report in the model selection table | PROC SEVSELECT | CRITERION= |
| Specifies the level of detail to be produced about the selection process | SELECTION | DETAILS= |
| Specifies the confidence level for reporting the confidence interval for EDF estimates | PROC SEVSELECT | EDFALPHA= |
| Limits or suppresses the display of class levels | PROC SEVSELECT | NOCLPRINT |
| Suppresses all displayed and graphical output | PROC SEVSELECT | NOPRINT |
| Specifies which graphical output to prepare | PROC SEVSELECT | PLOTS= |
| Specifies which graphical output to prepare about the selection process | SELECTION | PLOTS= |
| Specifies which output to display | PROC SEVSELECT | PRINT= |
| Prints internal names assigned to parameters | PROC SEVSELECT | PRINTINTERNALNAMES |
Last updated: July 09, 2026