Using either the DATDIF() or YRDIF() functions with datetime values as the arguments to the function might cause a "Floating Point Overflow" error. The arguments to these functions should be date values, not datetime values.

The "Floating Point Overflow" errors occur in DATA Step or the SQL procedure (PROC SQL), and only on AIX systems.

In PROC SQL, the following error is produced:

ERROR: Floating Point Overflow.
ERROR: Termination due to Floating Point Exception

If you are using a DATA Step, this error is produced:

ERROR: An unknown, abnormal error has occurred during execution at line xxx
       column y.
ERROR: Termination due to Floating Point Exception

The correct behavior is that a note is written to the SAS log:

NOTE: Invalid argument to function Function Name. Missing values may be generated.

Operating System and Release Information