The SYSLIN Procedure
- Overview
-
Getting Started
An Example ModelVariables in a System of EquationsUsing PROC SYSLINOLS EstimationTwo-Stage Least Squares EstimationLIML, K-Class, and MELO EstimationSUR, 3SLS, and FIML EstimationComputing Reduced Form EstimatesRestricting Parameter EstimatesTesting ParametersSaving Residuals and Predicted ValuesPlotting Residuals -
Syntax
-
Details
-
Examples
- References
The SYSLIN procedure statements and options are summarized in the following table.
|
Description |
Statement |
Option |
|---|---|---|
|
Data Set Options |
||
|
specify the input data set |
PROC SYSLIN |
DATA= |
|
specify the output data set |
PROC SYSLIN |
OUT= |
|
write parameter estimates to an output data set |
PROC SYSLIN |
OUTEST= |
|
write covariances to the OUTEST= data set |
PROC SYSLIN |
OUTCOV |
|
OUTCOV3 |
||
|
write the SSCP matrix to an output data set |
PROC SYSLIN |
OUTSSCP= |
|
Estimation Method Options |
||
|
specify full information maximum likelihood estimation |
PROC SYSLIN |
FIML |
|
specify iterative SUR estimation |
PROC SYSLIN |
ITSUR |
|
specify iterative 3SLS estimation |
PROC SYSLIN |
IT3SLS |
|
specify K-class estimation |
PROC SYSLIN |
K= |
|
specify limited information maximum likelihood estimation |
PROC SYSLIN |
LIML |
|
specify minimum expected loss estimation |
PROC SYSLIN |
MELO |
|
specify ordinary least squares estimation |
PROC SYSLIN |
OLS |
|
specify seemingly unrelated estimation |
PROC SYSLIN |
SUR |
|
specify two-stage least squares estimation |
PROC SYSLIN |
2SLS |
|
specify three-stage least squares estimation |
PROC SYSLIN |
3SLS |
|
specify Fuller’s modification to LIML |
PROC SYSLIN |
ALPHA= |
|
specify convergence criterion |
PROC SYSLIN |
CONVERGE= |
|
specify maximum number of iterations |
PROC SYSLIN |
MAXIT= |
|
use diagonal of S instead of S |
PROC SYSLIN |
SDIAG |
|
exclude RESTRICT statements in final stage |
PROC SYSLIN |
NOINCLUDE |
|
specify criterion for testing for singularity |
PROC SYSLIN |
SINGULAR= |
|
specify denominator for variance estimates |
PROC SYSLIN |
VARDEF= |
|
Printing Control Options |
||
|
print all results |
PROC SYSLIN |
ALL |
|
print first-stage regression statistics |
PROC SYSLIN |
FIRST |
|
print estimates and SSE at each iteration |
PROC SYSLIN |
ITPRINT |
|
print the reduced form estimates |
PROC SYSLIN |
REDUCED |
|
print descriptive statistics |
PROC SYSLIN |
SIMPLE |
|
print uncorrected SSCP matrix |
PROC SYSLIN |
USSCP |
|
print correlations of the parameter estimates |
MODEL |
CORRB |
|
print covariances of the parameter estimates |
MODEL |
COVB |
|
print Durbin-Watson statistics |
MODEL |
DW |
|
print Basmann’s test |
MODEL |
OVERID |
|
plot residual values against regressors |
MODEL |
PLOT |
|
print standardized parameter estimates |
MODEL |
STB |
|
print unrestricted parameter estimates |
MODEL |
UNREST |
|
print the model crossproducts matrix |
MODEL |
XPX |
|
print the inverse of the crossproducts matrix |
MODEL |
I |
|
suppress printed output |
MODEL |
NOPRINT |
|
suppress all printed output |
PROC SYSLIN |
NOPRINT |
|
Model Specification |
||
|
specify structural equations |
MODEL |
|
|
suppress the intercept parameter |
MODEL |
NOINT |
|
specify linear relationship among variables |
IDENTITY |
|
|
perform weighted regression |
WEIGHT |
|
|
Tests and Restrictions on Parameters |
||
|
place restrictions on parameter estimates |
RESTRICT |
|
|
place restrictions on parameter estimates |
SRESTRICT |
|
|
test linear hypothesis |
STEST |
|
|
test linear hypothesis |
TEST |
|
|
Other Statements |
||
|
specify BY-group processing |
BY |
|
|
specify the endogenous variables |
ENDOGENOUS |
|
|
specify instrumental variables |
INSTRUMENTS |
|
|
write predicted and residual values to a data set |
OUTPUT |
|
|
name variable for predicted values |
OUTPUT |
PREDICTED= |
|
name variable for residual values |
OUTPUT |
RESIDUAL= |
|
include additional variables in |
VAR |
|
