topic Re: Does sas stored process ignore some errors? in SAS Programming

Does sas stored process ignore some errors?

yassine_ben — Mon, 04 Dec 2023 00:30:04 GMT

I have a piece of SAS code that make some calculation, and to finish, add a line to a table "T", when I run the code on sas enterprise guide, gives the error below, after a "proc transpose" step, and no line is added to the table "T"

but when I run a stored process, having the same code, it doesn't return the error, and the line is added to the table "T"

anyone ever dealt with the same situation?

Re: Does sas stored process ignore some errors?

Sajid01 — Mon, 04 Dec 2023 01:29:39 GMT

Can you please share the log?

Re: Does sas stored process ignore some errors?

yassine_ben — Mon, 04 Dec 2023 09:42:40 GMT

hello

you mean the stored process execution log?

Re: Does sas stored process ignore some errors?

andreas_lds — Mon, 04 Dec 2023 10:25:31 GMT

Are you 100% sure that SP and EG (?) are processing the same input files?

Re: Does sas stored process ignore some errors?

Quentin — Mon, 04 Dec 2023 12:11:46 GMT

No, that should not happen. Most likely you don't have the same data in scoring_client_p3 for some reason.

That error message means that in your data, there is at least one case where there are multiple records that have the same value for NumDossier and Score_Name.

You could add a step to identify these duplicate values:

proc sort data=scoring_client_p3;
by NumDossier  Score_Name;
run;

data dups;
  set scoring_client_p3;
  by NumDossier Score_Name;
  if not (first.score_name and last.score_name);
run;

When you run that in EG, you should see at least two records in dups. Sounds like when you run in it in the stored process, dups will have 0 records. That will confirm there is a difference in your data, which you would need to investigate.

Re: Does sas stored process ignore some errors?

Sajid01 — Mon, 04 Dec 2023 14:15:36 GMT

Yes.
After all you are executing a SAS code in STP.
Proc transpose is outputing an error and it is quite possible that the source datasets are different.