There is no such thing as a macro array. You can simulate arrays in macros in two ways: As you did, put the stuff into a single string, with a delimiter (blank in your case). Generate a number of macro variables with the same prefix. In case one, you can use the %SCAN macro function, (or, if you have quoted delimiters, %sysfunc(SCAN()) with the Q option) to access individual elements, in your case   %put %scan(&numVarlist,1,str( )); You could also use the second solution:     proc SQL; select distinct variable into: numVar1-numVar999 from Map_num:   In which case you can get the first element like this:   %put &numVar1; Or, if you are using macro variables to access the "array":     %let i=1; %put &&numVar&i; The double ampersand works like this: the macro interpreter scans the codes several times. Each time, double ampersands become single ampersands, and "unmatched" ampersands trigger interpretation of variables. So, after the first scan, &&numVar&i becomes &numVar1, in the second scan that becomes the value (var1 in your example).   When using the first method, you can use the COUNTW function to find the number of values, with the second option, it is not so obvious (you may look at which macro variables exist, but then, they may be variables generated in an earlier, similar SQL call). So in any case I advice you to store the value of the SQLobs macro variable after the SQL call: proc SQL; select distinct variable into: numVar1-numVar999 from Map_num: %let nVars=&SQLobs; ... View more

Why not convert the two variables to numeric in the SQL call? proc sql; create table combo as select dataset2.*, dataset1.var1, dataset1.var2 from dataset2 left join dataset1 on input(dataset2.id,best6.) = input(left(dataset1.id),best12.) ; quit; run; And be careful with those SQL aliases. Everybody uses them, but they very often make the code harder to read - in this case dataset1 was mapped to "b" and dataset2 to "a",  very confusing, IMO. It is not that hard to use a change command to get rid of the aliases in the final code, and it normally makes everything much easier to read. ... View more

You can do it with a hash table, e.g.: data have; input id1 @6 id2 $ @9 time yymmdd8.; format time date9.; cards; 1 A 20131128 1 B 20140214 1 C 20140530 1 C 20140622 1 D 20140831 1 D 20141220 1 A 20150217 1 A 20150302 1 C 20150410 1 D 20150425 1 E 20150609 1 F 20151025 1 C 20160612 1 D 20180101 2 A 19900515 2 B 19900813 2 E 19910522 2 A 19910524 2 F 19910919 2 G 19920101 2 A 19930321 ;run; data want; set have; by id1; if first.id1 then do; if _N_=1 then do; declare hash h(); rc=h.definedata('lasttime'); rc=h.definekey('id2'); h.definedone(); declare hiter iter('h'); end; else h.clear(); end; rc=h.find(); lasttime=time; if rc then h.add(); else h.replace(); runningcount1=h.num_items; rc=iter.first(); starttime=intnx('year',time,-1,'same'); runningcount2=0; do until(iter.next()); if lasttime>=starttime then runningcount2=runningcount2+1; end; keep id1 id2 time runningcount1 runningcount2; run; ... View more

If your data is character, you may as well treat is as such, instead of messing round with numeric formats: data have; length ln_no $20; input ln_no $ ; datalines; 21555555 5544000 6654444 1145555555555 ;run; data want; set have; if lengthn(ln_no)<10 then ln_no=repeat('0',9-lengthn(ln_no))!!ln_no; run; I used the LENGTHN function because the LENGTH function will report a length of 1 for an empty string (LENGTHN returns 0). ... View more

What's the mistake in the above which is causing the "forever" run? Could be that one of your data files contains a record which does not begin with "11" or "22" - in which case your datastep will not execute any INPUT statements, only the final OUTPUT statement, and will be stuck in the DO loop. Or it may be a problem with the END=DONE condition, which may not be set correctly when you read from multiple infiles like that. In the last case, you may want to take a look at the EOF= option for infiles.   All in all, I would try something like this: DATA HISTORY_&BORD_NAME; Next_file: SET FILELIST; FILEPATH = "&DIRNAME"||FILE_NAME; INFILE TEMP FILEVAR = FILEPATH EOF=Next_file TRUNCOVER; DO WHILE(1); INPUT @1 RECORD_ID $2. @; IF RECORD_ID = "11" THEN INPUT @3 DEALER_NUMBER $4. ; ELSE IF RECORD_ID = "22" THEN INPUT @3 INVOICE_NUMBER $7. ; else do; /* In case we have an invalid(?) record ID */ input; /* skip to next line */ error; /* write the stuff to log */ continue; /* do not output */ end; OUTPUT; END; FORMAT RECORD_ID $CHAR2. DEALER_NUMBER $CHAR4. INVOICE_NUMBER $CHAR4. ; RUN; This EOF= option makes the program jump to the label indicated when a read after end of file occurs. And the SET statement terminates the datastep, when there are no more records in FILELIST.   I moved the FORMAT statement to the end, as it is not executable (it makes no sense to put it inside a conditionally executed block). ... View more