Dear All, Please, can someone guide how can I use proc rank to select the best 10% of the records by each treatment and then apply to proc survey for tipping point analysis?   Thanks ... View more

  Hi All,   Would you please let me know the way to write out an error or warning message along with the name given in the varlist which is not in the input dataset. For example, %varlist= a b c d; and variable d is not in the input 'CLASS' dataset I would like to the display in the log the following message after running names1 datasets. 'Variable 'D'  is not in the CLASS dataset.' or if variable D and C are not in the class dataset then the message 'Variable C and D are not in the CLASS dataset.' And after that system should abort or go to the end instead running next data step.           %let varlist= a b c d; %let varlist_quotesAndcomma="%sysfunc(prxchange(s/\s+/%nrbquote(",")/i,-1,&varlist))";  %put &=varlist_quotesAndcomma; DATA checkdata; %DO _t = 1 %TO %sysfunc(countw("&varlist",' ')); name = strip(scan("&&varlist",&_t," ")); OUTPUT; %END; RUN; data names (keep=name); set dictionary.columns( where =( libname='SASHELP' and memname='CLASS' and upcase(name) in (%upcase(&varlist_quotesAndcomma)))); run; data names1; merge names(in=names) checkdata(in=checkdata); if checkdata and not names; run;   Thank you! ... View more

Dear All,   Would you please help me to achieve something as below?   list1 contains a list of continuous and categorical variables altogether & sublist contains only continuous variables.  If the sublist contains more then 2 variables then 'else do' step runs twice or more. Would it be possible to put a stop or condition on the 'J'  loop that it should only run once for list1 variables?     %do i = 1 %to %sysfunc(countw(&list1)); %do j = 1 %to %sysfunc(countw(&sublist));   %let main&i. = %scan(&list1,&i,%str( )); %let sub&j. = %scan(&sublist,&j,%str( ));   %if &&main&. = &&sub&j. %then %do; /*Some steps */ %end; %else %do;   /**other steps**/   %end; %end; %end;   Thank you!   ... View more