2.6 Database Processing with PROC SQL

As an alternative to DATA step manipulation, the SAS implementation of a structured query language in PROC SQL might be a better alternative. (This is above and beyond the obvious applications of querying a data table and constructing summary tables.) In fact, programmers who come to SAS with a broad exposure to database programming (say, via formal computer science or MIS training) might find SQL to be more natural for data set manipulation than old-time SAS programmers. (Yeah, this is a confession. One of my former students, Matt Wheeler, did a fair amount of project work using PROC SQL that convinced me that I should learn more about this procedure.) This section provides an overview of this procedure with an emphasis on using PROC SQL to concatenate and merge data sets. Background references that might be interesting include documentation about SAS Advanced Certification.

The big idea is that PROC SQL looks at data sets as tables that can be manipulated, processed, and displayed using command queries. Here is the basic syntax, excerpted from SAS Help: