If there was a SAS performance tool that could
drastically reduce your program’s I/O’s, lower its CPU
time, and decrease its run time, would you use it?
Of course you would! Such a performance tool
exists; it is called a SAS index. SAS Indexes can
dramatically improve the performance of programs
that access small subsets of observations from large
SAS data sets. They do this by only accessing and
returning the observations that you specify in a
WHERE expression, instead of reading the entire SAS
data set.
It is easy to understand how a SAS index can help
you to directly access the observations that you
need in a particular SAS data set. As an exercise,
do the following: Open SAS Online Documentation,
Click on the tab, enter the word “rtrace”
into the Search tab’s window, and click on
. The SAS Online Documentation search
function returns about a dozen links. When you click
on any of those links, you get a page in the
documentation that discusses the SAS RTRACE
facility. This saves you the tedious effort of going
through the entire SAS Online documentation, page-
by-page, looking for occurrences of the
word “rtrace”.
A SAS index is analogous to the search function,
above. A good index allows your programs to quickly
access the subset of SAS observations that you
need from a large SAS data set when you specify a
key variable value (or values) that must be
matched. This can dramatically improve the speed
and efficiency of your SAS programs.
Conversely, badly conceived SAS indexes return far
too many observations and are no better than
reading the entire data set sequentially. In the
analogy, above, consider how many pages would be
returned and how much longer it would take if you
searched the SAS Online Documentation for the
word “SAS”. That is why it is important to know
more about the selection criteria for index variables,
as well as the actual creation and use of SAS
indexes.
After deciding that an index is appropriate for your
subsetting purposes, you have three tools to choose
from to create one: the DATASETS procedure, the
SQL procedure, and the DATA Option in a Procedure
or in the DATA step. When you do so, SAS creates
an index file and associates it with your SAS data
set. SAS stores additional indexes in that file and
deletes the file when all indexes have been removed
from the data set.
You can create a Simple index from a single variable,
or a Composite index from two or more variables. A
single SAS data set can have as many indexes as
you think are necessary. However, each index you
create increases the size of the SAS index file.
You can exploit indexes with the WHERE statement,
the BY statement, or the KEY statement used in
either a SET or MODIFY statement. In doing so, you
will be increasing the efficiency of your SAS programs
that use the index. That is what SAS indexes are
good for!
There is enough information about SAS indexes to fill
an entire book. If you are interested in learning
more, either check the SAS online documentation, or
consider my new book: The Complete Guide to SAS
Indexes, at:
http://www.sas.com/apps/pubscat/bookdetails.jsp?
catid=1&pc=60409
Michael A. Raithel, a Senior Systems Anaylst for
Westat, has published two previous SAS books and
over twenty SAS technical papers. He has been a
section chair at SUGI, SESUG, and NESUG, and co-
chaired NESUG in 1995. A copy of his first SAS book,
Tuning SAS Applications in the MVS Environment,
resides in the Smithsonian Institution of American
History’s Permanent Research Collection of
Information Technology. You can reach him at:
michaelraithel@westat.com
