Problem printing first value using if first.variable

[Iona]

Hi everyone!

I'm using the SAS sample data set called stress99.

data sasuser.stress99;
   input ID $ 1-4 Name $ 7-18 RestHR 25-26 MaxHR 32-34 RecHR 39-41
         TimeMin 48-49 TimeSec 57-58 Tolerance $ 65 Year 70-73;
datalines;
2501  Bonaventure, T    78     177    139      11       13      I    1999
2544  Jones, M          79     187    136      12       26      N    1999
2552  Reberson, P       69     158    139      15       41      D    1999
2568  Eberhardt, S      72     182    122      16       49      N    1999
2571  Nunnelly, A       65     181    141      15        2      I    1999
2578  Cameron, L        75     158    108      14       27      I    1999
2579  Underwood, K      72     165    127      13       19      S    1999
2588  Ivan, H           70     182    126      15       41      N    1999
2595  Warren, C         77     170    136      12       10      S    1999
;

Now what I would like to do is to create a data set that holds only the smallest value for TimeMin, which would be TimeMin=11.

I tried:

proc sort data = sasuser.stress99; by timemin; run; 

data bla (drop = resthr maxhr rechr tolerance timesec year);
	set sasuser.stress99;
	by  timemin;
	first_time= first.timemin;

        if first.timemin;
run; 

But what it will give me is the first UNIQUE value for TimeMin. Where am I going wrong?  I found this solution and I don't see why it works in their example but not in mine. 

Thank you!

[jarg]

the way that the unprinted variable first.[var] works is that it will resolve to 1 for the first observation with a distinct value - so if you are conditionally setting a table by saying if first.timein it will keep the first observation with a distinct value (this is also what's happening in the other solution you linked.)

if you're after the minimum value in the whole dataset, using  i would recommend turning it into a macro;

proc sql noprint; select min(timemin) into: mintimemin from stress99;
%put &mintimemin.;

this will allow you to run calculations off this value

[Iona · Accepted solution]

Thank you for your answer! I was able to use it. I wasn't too clear in my question but I wanted to keep the whole observation with the minimum timemin.

I used your solution to write

proc sql;
create table subset as
select ID, Name, min(timemin) as mintime from stress;

I never used proc SQL so while it's straightforward it's still new to me. 

Thanks!

[Astounding]

The link you posted is solving a different problem: the minimum for each category.

Try (after sorting) :

data bla (drop = ....) ;
set sasuser.stress99 (obs=1) ;
run;

[Iona]

Thank you, it's a very straightforward answer!