How Do I Get Help with my SAS Problem?

If you are stuck on a problem with SAS, there are several ways to get assistance:

• Read your log file carefully. In many instances, the error messages will show exactly where the error occurs and give an explanatory message. The log file should always be examined before reading output. Common errors are
  • omitting the semicolon at the end of a command;
  • forgetting to close a title with quotation marks (or enclosing an apostrophe in single quotes);
  • forgetting to terminate a comment with a semicolon;
  • omitting or reversing the /* */ notation on comment blocks;
  • placing a statement meant for a data step after a PROC (without beginning a data step).

• Consult the manuals. Learning your way around the documentation will save you time in the long run and enable you to be self sufficient in debugging SAS problems. The complete set of manuals (and PDF files for printing) are available on the U cluster. The version 6 hard copy manuals are still very useful. SAS Institute has a link to the complete system of online documentation. You will need to obtain a username and password by pointing your browser to http://www.sas.com/v9doc http://www.sas.com/v9doc and completing the usage agreement. The documentation is available at http://v9doc.sas.com/sasdoc. For specific problems in running SAS locally, check the man pages. At the system prompt, type

  man sas

• Examples of SAS programs can be found in the directory /usr/local/sas/sas9.00/samples/x where x is the name of the product containing the SAS module you are using. For example, sample programs using PROC GLM can be found in /usr/local/sas/sas9.00/samples/stat under the names glm*.sas. For a description of these products, see SAS Products. Some older example programs can also be found in /usr/local/sas/sas612/samples/x.

• There are many resources available through the Internet to provide help. One possibility is to check the SAS Institute's web page for FAQs (Frequently Asked Questions). Another is to check a local page which provides links to helpful information.

• Help on simple SAS problems may be obtained at the CCIT Help Desk, Room 224 (621-4357). If consultants at the desk are not able to help you, they will refer you to an individual knowledgeable in the area.

• Help on problems related to a current job running SAS (e.g., a full temporary disk, unable to access SAS) can be obtained by sending email to the CONSULT list.

• Help on statistical problems or on setting up SAS analyses can be obtained by sending email to the STATHELP list. For lengthy problems, contact the CCIT Help Desk for a referral to a consultant who specializes in this area.

• Two excellent references for new users are
  • Delwiche and Slaughter, "The Little SAS Book: A Primer", and
  • Cody and Smith, "Applied Statistics and the SAS Programming Language".

  ---

  Return to Top Document