How do I get VARMAX to Complete in a Macro with a Floating Point Error

[markgggg]

I’m running Proc Varmax on sparse data and I’m receiving an “Invalid Operation, Floating Point Error” message and Proc Varmax.stops midway and my whole bigger program stops.  I’m assuming that since the input Time Series data is so sparse it’s dividing by zero somewhere. 

Here’s the issue.  I notice that when this happens in a Macro, Varmax stops (and stops my bigger program) and when I run it in open code (no macro call) Varmax gives the error but completes and gives no output (in which case my bigger program would continue running).  For my purpose it’s OK for Varmax to produce no output as long as it keeps running and doesn’t stop my whole bigger program.

How do get Varmax to keep running in this situation when it’s in a Macro?

I’m using SAS enterprise guide Version 7.1

[WarrenKuhfeld]

Perhaps this will help?  There are system options that control what SAS does after encountering an error.

https://support.sas.com/resources/papers/proceedings15/1565-2015.pdf