That user has 4 posts counted, but none appear in the list, which means they all were marked as spam or deleted.
I assume that one of the moderators (@ChrisHemedinger !) will soon take care of the account.
... View more
In a DATA step, use
by usubjid;
At first.usubjid, set join to 1.
If not first.usubjid, and either ae is not equal to lag(ae) or aestdtc is greater than lag(aendt), increment join by 1.
... View more
Try this:
proc sort
data=mydataset (
keep=id var1 var2 var3 var4 time_flag
)
out=sorted
;
by id var1 var2 time_flag;
run;
data spine;
set sorted;
by id var1 var2;
if last.var2;
run;
SQL will also sort, and it is usually less efficient than PROC SORT.
... View more
In the SET clause of the UPDATE statement, only variables of the dataset named in UPDATE can be set, so no identification of a specific table is needed and therefore not recognized by the SQL compiler.
... View more
Your problem is not the same one. The original question was about SASUSER.PROFILE, your issue is with WORK.SASMACR. Different catalog file, different location. You probably exceeded your quota in the WORK location, maybe because of remnants of crashed or improperly terminated SAS sessions.
... View more
Since we needed to secure our "data at rest" per company policy, we switched to metadata-bound and encrypted libraries, which worked quite well. We did uncover a bug in the SAS metadata software that led to a race condition, it has been fixed in the meantime.
... View more
This will not really work. For a view to successfully execute, the user running it has to have read access to the underlying tables, which means they also can directly read those.
I don't know if it is possible to restrict access to specific variables via metadata when libraries are metadata-bound.
... View more
