Transpose to long, then sort:
data long (rename=(dcdate=date));
set have;
array ic {*} icode:;
do i = 1 to dim(ic);
if not missing (ic{i}
then do;
icode = ic{i};
output;
enđ;
end;
keep id dcdate icode;
run;
proc sort data=long nodupkey;
by id date icode;
run;
... View more
Transpose to a long dataset, so data (the years) is not hidden in structure. Maxim 19!!
As soon as data is data, no dynamic code is needed, as the data drives the process.
... View more
Since no variables with names starting with nr_laks are present in your DATA step, and custid is automatically retained (because it is present in an incoming dataset), the RETAIN statement is completely irrelevant and does nothing.
... View more
At the moment you have two concurrent network streams competing for bandwidth: the download from your SAS server to EG, and the output from EG to the network share. If the share is also available on the SAS server, writing to it directly with PROC EXPORT (or a DATA step you write yourself) will perform better.
... View more
The 49-169 points to a problem with unbalanced quotes or similar which has left the SAS system in an unstable state. Start a new session and run only your import code, then copy/paste the whole log into a window opened with this button:
You can overwrite sensitive information with X's.
... View more
Writing ugly spaghetti code like this is a recipe for problems.
DATA Count;
SET Medications;
BY Patient;
IF First.Patient THEN DO;
IF Medication = 'Aspirin' THEN DO;
Count = 0;
Count + 1;
IF LAST.SubjectNumber;
END;
END;
RUN;
Now you can clearly see that the increment is part of the FIRST. block and therefore executed only once per patient.
You also used a variable not included in the BY for your LAST.
data count;
set medications;
by patient;
if first.patient then count = 0;
if medication = 'Aspirin' then count + 1;
if last.patient;
run;
... View more
If your SAS variable is not formatted as a datetime, the ACCESS module you use to connect to the database does not know that the target should be of type datetime. Apply an appropriate format in SAS.
... View more
We started with scheduling long before SAS had built-in options for this.
Scheduling through the IT-wide system (Control-M) was done only by the SAS administrators, but SAS users also had the option to schedule their programs by logging on to the SAS server and using cron.
... View more
If you‘re concerned about storage consumption on the network share, you need to set quotas there and force the users to clean their own house.
... View more
Really no need for hash objects, the standard tools of sorting and merging are sufficient. Particularly with extremely small datasets of just 10000 observations.
proc sort data=old;
by usubjid visit date;
run;
proc sort data=new;
by usubjid visit date;
run;
data compare;
merge
old (in=o)
new (in=n)
;
length flag $7;
if o and not n then flag = "DELETED";
else if not o and n then flag = "NEW";
else flag = "UPDATED";
run;
How should "observation numbers" be calculated?
... View more
