SET commands can be issued to Hive using explicit pass-through code in the SQL procedure. Here is an example:

proc sql noprint; connect to hadoop (server=server-name subprotocol=hive2 user=user-name password=password schema=hive-schema config=config-file); execute(set hive.exec.compress.output=true) by hadoop; execute(set mapred.output.compression.type=BLOCK) by hadoop; disconnect from hadoop; quit;

This code sends the contents of both EXECUTE statements directly to Hive for execution.