try this, your sign is backwards and you only want to pull the date, not the date and time: where datepart(mydate)<'01Aug2013'd; ... View more

I'm guessing there will be more elegant solutions than this, but here's one way to do it: data have; infile cards dsd; input bep_fisc_yr    bep_app_code   bep_org_code   bep_pgm_ele_code  tran_pgm_ele_code   end_date$ ; cards; 2009,10,45,34,234,null 2009,10,45,34,234,current 2009,10,45,23,234,null 2009,10,45,34,234,current 2009,10,45,23,234,current 2009,10,45,13,234,null ; run; data want; set have; lbep_f = lag(bep_fisc_yr); lbep_a = lag(bep_app_code); lbep_o = lag(bep_org_code); lbep_p = lag(bep_pgm_ele_code); ltran = lag(tran_pgm_ele_code); if _N_ > 1 then do; if lbep_f ne bep_fisc_yr then lbep_f_c = 1; if lbep_a ne bep_app_code then lbep_a_c = 1; if lbep_o ne bep_org_code then lbep_o_c = 1; if lbep_p ne bep_pgm_ele_code then lbep_p_c = 1; if ltran ne tran_pgm_ele_code then ltran_c = 1; end; if sum(lbep_f_c,lbep_a_c,lbep_o_c,lbep_p_c,ltran_c) < 3 then change = 'No Change';     else change = 'Change'; run; ... View more

I think this is what you are looking for but I"m not sure: data want; set have; by type notsorted; format l_edate mmddyy10.; l_edate = lag(edate); if first.type and type = 'Sdate' then do; diff_date = edate - l_edate; end; run; ... View more

Is it only this instance, or is it every rdate/sdate combination? ... View more

Here's a solution, if want you mean by age is 15 days then you do not need the function, if you want month or years just replace day with what you want: data have; infile cards dsd; length defect $10.; informat col1 yymmdd8.; format col1 mmddyy10.; input BODYNO$ DEFECT$ ADJKM former_variable$ former_label $ COL1; cards; 0NP07185,99600668MP,15332,DREP,DREP,20130116 0NP07185,99600668MP,29407,DREP,DREP,20131029 0NP07185,99600668MP,41802,DREP,DREP,20140602 0NP07185,99600668MP,51477,DREP,DREP,20141222 0VU43599,99600668MP,15572,DREP,DREP,20130227 0VU43599,99600668MP,29010,DREP,DREP,20130619 0VU43599,99600668MP,44477,DREP,DREP,20131114 0VU43599,99600668MP,59886,DREP,DREP,20140614 0VU43599,99600668MP,72938,DREP,DREP,20150325 8DN12628,99600668MP,74186,DREP,DREP,20100727 8DN12628,99600668MP,101807,DREP,DREP,20110919 8DN12628,99600668MP,126877,DREP,DREP,20120229 8DT43009,99600668MP,18829,DREP,DREP,20091215 8DT43009,99600668MP,18999,DREP,DREP,20100514 8DT43009,99600668MP,40864,DREP,DREP,20110912 8DT43009,99600668MP,50756,DREP,DREP,20111129 8DT43009,99600668MP,66477,DREP,DREP,20120822 8DT43009,99600668MP,77629,DREP,DREP,20130823 8DT43009,99600668MP,87800,DREP,DREP,20140805 8DT43009,99600668MP,90984,DREP,DREP,20150407 ; run; proc sort data=have;by bodyno defect; data all want; set have; format l_col1 mmddyy10.; by bodyno defect; l_adjkm = lag(adjkm); l_col1 = lag(col1); if not first.bodyno then do; diff_adjkm = adjkm - l_adjkm; diff_date = intck('day',l_col1,col1); end; output all; if diff_adjkm > 10000 and diff_date > 15 then output want; run; ... View more

I think you were looking for a way to do all of this in one step, not so much getting away from sql, is that right?  If so this will get you what you want: data have1;   input ID cat_var1 $ cat_var2 $ cat_var3 $; cards; 1 a d i 2 b f h 3 c e j 4 a e h 5 c d h ; data have2;   input ID     count_var     group_var; cards; 1 1 1 1 1 1 1 1 2 1 1 2 1 1 2 2 1 1 2 1 1 2 1 1 ; proc sql; create table want as select distinct a.id, a.cat_var1, a.cat_var2, a.cat_var3, count(b.group_var) as count_group from have1 a left join      have2 b on a.id = b.id group by b.id,b.group_var; ... View more

Here you go, I changed a little. data have ;   input geo_key $ var1 var2 var3 var4 var5 var6; cards; 001 1 1 1 1 1 0 002 1 1 1 1 0 0 003 1 1 1 0 0 0 004 1 1 0 0 0 0 005 1 0 0 0 0 0 ; proc transpose data=have out=prep;by geo_key; proc sql; create table want as select distinct geo_key,sum(col1) as result from prep group by geo_key order by geo_key; ... View more

Slight variation, this should do it: data have ;   input geo_key var1 var2 var3 var4 var5 var6; cards; 1 1 0 0 0 0 0 2 0 0 0 0 0 0 3 1 1 1 1 1 1 4 0 0 1 0 1 0 5 1 0 1 0 1 0 ; proc transpose data=have out=prep;by geo_key; proc sql; create table want as select distinct _name_,sum(col1) as result from prep group by _name_ order by _name_; ... View more