[下载]Introduction To SAS Programming

SN-009419 ***Alert Note***

Forecast output may be incorrect when exogenous variable is differenced

Product: SAS/ETS Component: VARMAX procedure Priority: ALERT

If an exogenous variable for a VARX model is differenced by a greater number of lags than the number by which each endogenous variable is differenced, then the results in the OUT= data set and the ODS Forecasts table are incorrect.

To circumvent the problem, perform the differencing of the exogenous variable in a DATA step prior to estimating the model with PROC VARMAX.

SN-009376 ***Alert Note***

Correlations of the parameter estimates may be incorrect in PROC VARMAX

Product: SAS/ETS Component: VARMAX procedure Priority: ALERT

The PRINT=(CORRB) option results in incorrect correlations of the parameter estimates for any VARMAX model except for the Error Correction Model.

To obtain correct estimates, you can compute the correct correlations manually from the covariance matrix of the parameter estimates. Use the PRINT=(COVB) option to request the covariances and then compute the correct correlations as follows:

CORR(i,j) = COV(i,j) / SQRT( COV(i,i) * COV(j,j) );

SN-009424 ***Alert Note***

Contradictory differencing specifications result in incorrect differencing

Product: SAS/ETS Component: VARMAX procedure Priority: ALERT

If differencing is specified with both the DIFX= option and the DIFY= option on the MODEL statement, and if the DIF= option on the MODEL statement specifies differencing which is contradictory with either DIFX= or DIFY= or both, then the variables are differenced incorrectly, the resulting output is incorrect, and one or both of the following warnings appear in the SAS log:

WARNING: Since the DIF= option is specified, the DIFX= option is ignored.

WARNING: Since the DIF= option is specified, the DIFY= option is ignored.

To circumvent the problem, specify the differencing for the exogenous variables using either the DIFX= option or the DIF= option and specify the differencing for the endogenous variables using either the DIFY= option or the DIF= option.

SN-009415 ***Alert Note***

Values in the OUTEST= data set are incorrect for VAR-GARCH models

Product: SAS/ETS Component: VARMAX procedure Priority: ALERT

The values in the OUTEST= data set are incorrect for any VAR-GARCH model, which is specified with the P= and the GARCH= options on the MODEL statement.

The GARCH= option is considered experimental in Release 8.2.

To circumvent the problem, output the parameter estimates and standard errors using the appropriate ODS tables instead of using the OUTEST= option. For example:

proc varmax data=work.a; ods output parameterestimates=work.b garchparameterestimates=work.c; model y1 y2 / p=1 garch=(q=1); nloptions tech=tr; run;

2003 SASware Ballot Results

2004 SASware Ballot Results

SAS Online GuidesNote that many of these guides have been developed to support local users at the host university; location-specific information may not be applicable to the Mississippi State University computing environment.

SAS Home Page

http://www.sas.com

Home page for SAS Inc., including links to user support and documentation.

An Introduction to the SAS System

http://stat-www.berkeley.edu/classes/s100/sas.pdf

Written by Phil Spector at the University of California, Berkley, this (pdf) guide provides an extensive introduction to SAS.

Regression with SAS

http://www.ats.ucla.edu/stat/sas/webbooks/reg/default.htm

Provides online access to the web book Regression with SAS by Xiao Chen, et. al.

SAS Guide

http://academic.csuohio.edu/holcombj/mth567/descrip.pdf

Short pdf guide to data entry and command structure.

SAS Online Samples

http://support.sas.com/documentation/onlinedoc/code.samples.html

Online examples of various types of SAS code, designed to accompany SAS user guides.

SAS/STAT User's Guide

http://jeff-lab.queensu.ca/stat/sas/sasman/sashtml/stat/index.htm

Provides online access to the SAS/STAT User's Manual.

[此贴子已经被作者于2005-4-9 11:24:59编辑过]

SAS Print Reference

Reference Resources

The following resources are available in the Reference Room.

Spector, Paul E. SAS Programming for Researchers and Social Scientists. Thousand Oaks, CA: Sage, 2001. Ref QA276.4 .S645 2001

Other Print Resources

The following print resources are available in the Mississippi State University Libraries' circulating collection. Additional materials may be located by searching the Galaxy Online Catalog.

Allison, Paul David. Logistic Regression Using the SAS System: Theory and Applications. Cary,NC: SAS Institute, 1999. QA278.2 .A4348 1999

Arthur, Winfred. Conducting Meta-analysis Using SAS. Mahwah, NJ: Lawrence Erlbaum, 2001. BF39 .A78 2001 (Also available online through netLibrary.)

Der, Geoff. Handbook of Statistical Analyses Using SAS. Boca Raton, FL: Chapman & Hall/CRC, 2002. QA276.4 .D47 2002

Freund, Rudolf . SAS System for Regression. Cary, NC: SAS Institute, 2000. QA276.4 .F74 2000

Gravely, Archer R. Your Guide to Survey Research Using the SAS System. Cary, NC: SAS Institute, 1998. QA276.4 .G73 1998

Hatcher, Larry. A Step-by-step Approach to Using the SAS System for Factor Analysis and Structural Equation Modeling. Cary, NC: SAS Institute, 1994. QA278.5 .H386 1994

Hatcher, Larry. A Step-by-step Approach to Using the SAS System for Univariate and Multivariate Statistics. Cary, NC: SAS Institute, 1994. QA276.4 .H37 1994

Khattree, Ravindra. Applied Multivariate Statistics with SAS Software. Cary, NC: SAS Institute, 1999. QA278 .K43 1999

Scerbo, Marge. Health Care Data and the SAS System. Cary, NC: SAS Institute, 2001. R864 .S24 2001

Stokes, Maura Ellen. Categorical Data Analysis Using the SAS System. Cary, NC: SAS Institute, 2000. QA276.4 .S74 2000

so great

fantasitc

wow，收集的好全！感谢！

非常非常感谢。