Not sure what you have done.  The option cannot be assigned a missing value and the option only affects new 2-digit year data being read in, not existing SAS dates. The documentation can be found at: http://support.sas.com/documentation/cdl/en/lrdict/64316/HTML/default/viewer.htm#a000197881.htm ... View more

Use options obs=max. The actual value will vary depending on the version of SAS and your hardware. xx  options obs=100; xx  %put %sysfunc(getoption(obs)); 100 xx  options obs=max; xx  %put %sysfunc(getoption(obs)); 9223372036854775807 ... View more

I probably would have coded it differently, but just modifying Dhana's code (if I correctly understand what you want), you might be able to use something like: data count(keep=cus_id first last flag);   input cus_id jan_08 feb_08 mar_08 apr_08         may_08 jun_08 jul_08 aug_08 sep_08         oct_08 nov_08 dec_08         jan_09 feb_09 mar_09 apr_09         may_09 jun_09 jul_09 aug_09 sep_09         oct_09 nov_09 dec_09;   array my{*} jan_08--dec_09;   format first last date9.;     do pay_1=1 to dim(my);     if my(pay_1) ne . then do;     first=input(catt("01",       substr(vname(my(pay_1)),1,3),       "20",       substr(vname(my(pay_1)),5,2)),       date9.);     goto done;    end;   end;   done:   do pay_n=dim(my) to 1 by -1;     if my(pay_n) ^=. then do;       last=input(catt("01",         substr(vname(my(pay_n)),1,3),         "20",         substr(vname(my(pay_n)),5,2)),         date9.);       goto finish;     end;   end;   finish:   count=0;   do i=pay_1 to pay_n;     if my(i) =. then do;       count+1;     end;   end;   if count > 3 then flag=0;   cards; 1324  56 . . . 45 67 19 . . . . . 1 2 3 4 . . 7 8 9 0 1 2 1256   . . . . 45 . 78 . . . . 24 1 2 3 4 5 . . . . . . . 1257   1 2 3 4 45 . 78 . 5 5 5 24 1 2 3 4 . . . . . . . . ; run; HTH, Art ... View more

Give you an example. The code search all the sub-directory under c:\temp\ and convert xls file into csv file at respective sub-directory. [pre] filename _xls pipe 'dir c:\temp\*.xls /s'; data xls_file; infile _xls length=len; input whole $varying200. len; retain directory; length file_name $ 200; if strip(whole) eq: 'c:\temp' then directory=scan(whole,1,' '); else if scan(whole,-1,'.') eq 'xls' then do; _name=scan(trim(whole),-1,' '); file_name=catx('\',directory,_name); output; end; keep file_name; run; data _null_; set xls_file nobs=num_obs; call symputx(cats('path',_n_),file_name); call symputx(cats('name',_n_),scan(file_name,-2,'.\')); call symputx('nobs',num_obs); run; %put _user_; options mprint mlogic symbolgen; %macro xls_to_csv; %do i=1 %to &nobs; proc import datafile="&&path&i" out=&&name&i dbms=excel replace; getnames=yes; mixed=yes; run; proc export data=&&name&i outfile="%scan(&&path&i,1,'.').csv " dbms=csv replace;run; %end; %mend ; %xls_to_csv [/pre] Ksharp ... View more

Maybe you can code like ,Not tested. [pre] data temp; set sashelp.vtable(where=(libname='WORK') keep=memname nobs); if memname eq: 'DAILY'; run; options obs=1; data want; set daily1 - daily200 indsname=dsn; memname=scan(dsn,2,'.'); run; data want; merge want temp; by memname; run; [/pre] Ksharp ... View more

Presuming hi fun represents hyphen, the following might serve the purpose. There is probably a more direct way with regular expressions, but I am still just learning those: [pre] data a; informat s $200.; input id s &; datalines; 1 prg-code,main,IT-Hardware,Ins,Main,Non_IT- ,WOrk-Hardwrk,giveup,main,doit, -Nowork,ceo-cash,main,user,repair,nonit, ; run; data a2 (keep=id want); set a; length want $35; array x (99) $35.; i=1; do while (scan(s,i,',') ne ""); x(i)=scan(s,i,','); i+1; end; want=""; do j=1 to i; want=catx(',',want,x(j)); if j eq i or index(x(j+1),"-") gt 0 then do; output; want=""; end; end; run; [/pre] HTH, Art --------- > data a; > input id s $ 3-118; > datalines; > 1 prg-code,main,IT-Hardware,Ins,Main,Non_IT- > ,WOrk-Hardwrk,giveup,main,doit, > -Nowork,ceo-cash,main,user,repair,nonit, > run; > > LOGIC :I WANT TO MOVE IT IN TO ANOHTER ROW BASED ON > "- "(HI FUN) BEFORE HIFUN FIRST COMMA . > > i wnat it in data set a2 as > > output: > > 1 prg-code,main > 1 IT-Hardware,Ins,Main > 1 Non_IT- > 1 WOrk-Hardwrk,giveup,main,doit > 1 -Nowork > 1 ceo-cash,main,user,repair,nonit ... View more

Another solution. [pre] data a; input id s $ 3-118; datalines; 1 prg-code,main,IT-Hardware,Ins,Main,Non_IT- ,WOrk-Hardwrk,giveup,main,doit, -Nowork,ceo-cash,main,user,repair,nonit, 2 prgcode,main,IT-Hardware,Ins,Main,Non_IT- ,WOrk-Hardwrk,giveup,main,doit, -Nowork,ceo-cash,main,user,repair,non-it, 3 prg-code,main,IT-Hardware,Ins,Main,Non_IT- ,WOrk-Hardwrk,giveup,main,doit, -Nowork,ceo-cash,main,user,repair,non-it, ; run; data temp(drop=s i); set a; if id ne lag(id) then count+1; i=1;ss=scan(s,i,','); do while(not missing(ss)); if findc(ss,'-') then count+1; output; i+1; ss=scan(s,i,','); end; run; data want; set temp; by count notsorted; length s $ 100; retain s; if first.count then call missing(s); s=catx(',',s,ss); if last.count then output; drop ss count; run; [/pre] Ksharp ... View more

Thanks it worked ... View more

OK.It is easy. [pre] data want(where=( age is not missing)); length id no $ 8 name age $ 200; infile datalines length=len; input row $varying100. len; id=scan(row,1,' ,');no=scan(row,2,' ,'); i=1; do until(missing(name) or missing(age)) ; name=scan(scan(row,-2,' '),i,','); age=scan(scan(row,-1,' '),i,','); output; i+1; end; drop row i; datalines; 3 10 a,b,c 10,11,12 2 11 e,f,g 11,15,16 12 12 w,r,t 1,2,3 14 13 p,m 4,5 15 14 a,b,c,d,e 1,2,3,4,5 ; run; [/pre] Ksharp ... View more

Similar to Art.T's suggestion, but all in one datastep: [pre] data want(keep=id main_obs where=(main_obs is not missing)) ; infile datalines delimiter=', ' truncover; input id @; informat a b $20.; do until(missing(a) or missing(b)); input a b @; main_obs=catx(',',a,b); output; end; input; datalines; 1 typist-type,printing,computer-coding,writing,driver-driving,caring,ceo-mainhead,cordinator,asstgm-lesswrk,maintanace 2 typist-type,printing,computer-coding,writing,driver-driving,caring,ceo-mainhead,cordinator,asstgm-lesswrk,maintanace ; run; [/pre] Ksharp Message was edited by: Ksharp ... View more

Not sure about how the SUBJECT correlates to the post/message body. In a DATA step, if you need to "retain" an observation's value across DATA step passes, look at using a RETAIN statement to capture (with a SAS assignment statement), a temporary SAS variable, as required by your program. Scott Barry SBBWorks, Inc. Suggested Google advanced search argument, this topic / post: programming intro data step retain site:sas.com ... View more

Exactly how are you "importing data" - using SAS wizard, PROC IMPORT, or a DATA step that you coded? Also, what determines the DATA step INFILE statement (or SAS FILENAME) to setup the "fileref" to your input data? Suggest you share more about your actual current SAS program execution code, preferably as a forum post/reply and paste your SAS generated log output, in addition to explaining the relevant points above. Scott Barry SBBWorks, Inc. ... View more

The possible solution is to check date and time of creation of the text files and then read these files to SAS dataset. The code below is reading file attributes (date and time) to sas dataset. [pre] %let input=d:\dep\input\2010; %let dir=%str(%'dir %")&input.%str(\%" /A-D/-C/ON/TW/4 %'); filename myfiles pipe %unquote(&dir); data list1 (compress=no); infile myfiles truncover ; informat date ddmmyy10. time1 $5. bytes 8. name $100. ; input date ?? @1 @; if not missing(date) then do; input time1 13-17 bytes 19-35 name @37; end; else delete; time=input(time1,time5.); format date ddmmyy10. time hhmm.; drop time1; run; filename myfiles clear; [/pre] ... View more

Hi.Peter. Your 'base 26' would be some right.And something you thought would be further profound. I also am confused that why this author want to use this mapping. And your HyperLink,i readed,It is pretty useful. Learned it from you. 🙂 😉 ... View more

Just a variation: data _null_; a='01jan1990'd; b='31jul1992'd; c=yrdif(a,b,'ACT/ACT'); length d $ 22; d=cat(int(c),' year(s) ',ceil(12*mod(c,1)),' month(s)'); put d=; run; You might want to replace the ceil() function with round() or floor(). Your example does not really say what you intend to do with fractions of months, i.e. from 01Jan to 15Feb. Would this now be 1 month or 2 months. HTH Patrick Message was edited by: Patrick ... View more