Chapter Contents
Chapter Contents
The SYSLIN Procedure

Functional Summary

The SYSLIN procedure statements and options are summarized in the following table.

Description Statement Option
Data Set Options   
specify the input data setPROC SYSLINDATA=
specify the output data setPROC SYSLINOUT=
write parameter estimates to an output data setPROC SYSLINOUTEST=
write covariances to the OUTEST= data setPROC SYSLINOUTCOV
write the SSCP matrix to an output data setPROC SYSLINOUTSSCP=
Estimation Method Options   
specify full information maximum likelihood estimationPROC SYSLINFIML
specify iterative SUR estimationPROC SYSLINITSUR
specify iterative 3SLS estimationPROC SYSLINIT3SLS
specify K-class estimationPROC SYSLINK=
specify limited information maximum likelihood estimationPROC SYSLINLIML
specify minimum expected loss estimationPROC SYSLINMELO
specify ordinary least squares estimationPROC SYSLINOLS
specify seemingly unrelated estimationPROC SYSLINSUR
specify two-stage least-squares estimationPROC SYSLIN2SLS
specify three-stage least-squares estimationPROC SYSLIN3SLS
specify Fuller's modification to LIMLPROC SYSLINALPHA=
specify convergence criterionPROC SYSLINCONVERGE=
specify maximum number of iterationsPROC SYSLINMAXIT=
use diagonal of S instead of SPROC SYSLINSDIAG
exclude RESTRICT statements in final stagePROC SYSLINNOINCLUDE
specify criterion for testing for singularityPROC SYSLINSINGULAR=
specify denominator for variance estimatesPROC SYSLINVARDEF=
Printing Control Options   
print first-stage regression statisticsPROC SYSLINFIRST
print estimates and SSE at each iterationPROC SYSLINITPRINT
print the restricted reduced form estimatesPROC SYSLINREDUCED
print descriptive statisticsPROC SYSLINSIMPLE
print uncorrected SSCP matrixPROC SYSLINUSSCP
print correlations of the parameter estimatesMODELCORRB
print covariances of the parameter estimatesMODELCOVB
print Durbin-Watson statisticsMODELDW
print Basmann's testMODELOVERID
plot residual values against regressorsMODELPLOT
print standardized parameter estimatesMODELSTB
print unrestricted parameter estimatesMODELUNREST
print the model crossproducts matrixMODELXPX
print the inverse of the crossproducts matrixMODELI
suppress printed outputMODELNOPRINT
suppress all printed outputPROC SYSLINNOPRINT
Model Specification   
specify structural equationsMODEL 
suppress the intercept parameterMODELNOINT
specify linear relationship among variablesIDENTITY 
perform weighted regressionWEIGHT 
Tests and Restrictions on Parameters   
place restrictions on parameter estimatesRESTRICT 
place restrictions on parameter estimatesSRESTRICT 
test linear hypothesisSTEST 
test linear hypothesisTEST 
Other Statements   
specify BY-group processingBY 
specify the endogenous variablesENDOGENOUS 
specify instrumental variablesINSTRUMENTS 
write predicted and residual values to a data setOUTPUT 
name variable for predicted valuesOUTPUTPREDICTED=
name variable for residual valuesOUTPUTRESIDUAL=
include additional variables in X'X matrixVAR 

Chapter Contents
Chapter Contents

Copyright © 1999 by SAS Institute Inc., Cary, NC, USA. All rights reserved.