Lets asume you have a dataset A. We then create another dataset B and a view C that are exact copies of A.  If we then make a change to dataset A by adding or removing rows, etc. there will be no change to dataset B until we programatically update.  The view C, however, will be updated without submitting any statements. data A; val=1; output; val=2;output; val=3;output; run; data B; set A; run; proc sql; create view C as select * from A; quit; data A; set A; where val ne 2; run; proc print data=B; run; proc print data=C; run; ... View more

For finding valid adresses I think perl regular expressions functions would be a good way (prxparse, prxmatch). You can output data to different datasets within the same data step. data A B; set C; if condition then output A; else output B; run; Good luck! ... View more

It seems like you want to group by admit and svrty_lvl_cd unless I am missing something else. ... View more

To se the discussions you created you can click on your profile and then click "content." While there you can also go to your bio page and click "edit profile and privacy settings".  Under privacy settings you can change who can see your name/user name.  Your name is always visible to yourself ( I currently only see ncsthbell ). ... View more

Also %let group1=%sysfunc( translate( &group, ',' , ' ' )); works. ... View more

Hi Jason, You can try using the following code. %let vars=country state city county; %do i=1 %to 3; %let group=%substr(&vars,1,%eval(  %index(&vars,  %scan(&vars,%eval(5-&i))  )-1)  ); %put &group; %end; The nested functions can be a little hard to follow. Using another %do loop and concatenating the variables(%let group=&group %scan(&vars,&j) ) might be a better option if you want to improve readability of the program. Good luck! ... View more

Do you have country and dadbirthplacecountry in the wrong sides of the equals sign? If all you are doing is renaming then proc datasets might be a better choice. ... View more

SGPLOT may be a good place to look. Is this what you are trying to do? http://support.sas.com/documentation/cdl/en/grstatproc/62603/HTML/default/a003199396.htm ... View more

Year cut off doesn't need to be 100's of years. Why not use yearcutoff=1980 or whatever the smallest reasonable year you need? ... View more

It is possible to use UTF-16 encoding and the hexadecimal code ( var='03b1'x; ) but there is no way to ensure that it will appear correctly for someone else. I would suggest contacting tech support to see if they have a solution. ... View more

First off, || doesn't remove blanks and the new variables are not long enough to store the entire new string. The catx function might be a better way to go. data test; input (var1-var8) ($); cards; a a b b c c d d ; run; Data test2; set test; length newvar1-newvar4 $20; array intermediate[8] $ var1-var8; array newvar(4) $; array newvarA(4) $; do i=1 to 4; newvar(i)= intermediate[2*i-1]||"/"||intermediate[2*i]; newvarA(i)=catx("/",intermediate[2*i-1],intermediate[2*i]); end; keep newvar1-newvar4 newvarA1-newvarA4; run; ... View more

If your date is a character variable then use the compress funtion. If it is a numeric SAS date then use the format. data _NULL_; a='05/2011'; b= compress(a,'/'); c=today(); d=c; format d mmyyn6.; put a= b= c= d=; run; results in : a=05/2011 b=052011 c=18765 d=052011 ... View more