10-09-2015 03:18 PM
I have one dataset "A" which has one variable "test" which came out with zero observation
i wants to mention it in auto email from SAS to team that "we have &no_of_obs observation for test" but problem is my below macro variable,no_of_obs, is resolving to &no_of_obs only since i have zero observation,
is there any work around to resolve it to zero (0)?
10-09-2015 03:40 PM
Change the order of the statements. SAS stops the data step as soon as it reads past the end of the input stream. So the CALL SYMPUT() never runs.
data _null_; call symputX("no_of_obs",_n_-1); set a; run;
10-13-2015 08:30 AM
@ Miller - in my case, its not a case where i will always get 0 obs, could be possible obs>0
Yes, the code as I wrote it will produce a result of 0 if there is no data, and a positive number equal to the number of observations when there is a positive number of data points.
10-09-2015 08:57 PM
For efficiency purposes, this sort of problem is usually handled a little differently. Notice that your original code executes CALL SYMPUT for each observation in the data set. It that's a million times, you might be slowing the program down. An alternative:
call symputx("no_of_obs", _nobs_);
set a nobs=_nobs_;
Note that switching to CALL SYMPUTX will remove from your log the message about numeric to character conversion, and will remove leading blanks from the value of &NO_OF_OBS.