@ Cynthia: This is indeed something for VVALUEX function but unfortunately this function was not available on SAS 8 (as mentioned by Tim). @Ksharp: your code does generate the expected result but requires that we update the code everytime we have a new value coming in (e.g. we add variable F and variable can reference to this new variable). Again, thanks for your contributions ! Regards, Florent ... View more

Hi Tim, Thank you for the very quick answer: it's exactly what I was looking for ! 🙂 Do you know whether there is a similar function available for SAS 8 ? My application is executing remotely on SAS 9.2 (so, no problem) but if the server is down or the connection can not be established then it's running on the users' local session (where SAS 8 - release 8.02 is installed...). Thanks again for your help ! Regards, Florent ... View more

Hi Valérie, Could you please check whether the following code match your expectations ? 🙂 I hope it helps ! Kind regards, Florent proc sql noprint; update decava.person a set death_date = (select iden_mortaldate from ident_sel_1 where iden_inss = a.social_security_nr), smals_last_update_date = (select iden_lastupdt_datetime from ident_sel_1 where iden_inss = a.social_security_nr) where a.social_security_nr in (select iden_inss from ident_sel_1) ; quit; ... View more

Hi, Indeed, the macro definition (%macro send_mail) should appear before the datastep in which you make the calling to this macro. Kind regards, Florent ... View more

Hello, What if you replace your 'full outer join' by a 'cross join' ? It gives the following query: PROC SQL; CREATE TABLE TEST AS SELECT A.DATE,A.DAY,A.HANDLED,A.LEVEL1DESC, P.UCL,P.LCL FROM MEAN1 A CROSS JOIN DAILY_MEAN P ON A.LEVEL1DESC = P.LEVEL1DESC AND A.DAY = P.DAY ORDER BY A.LEVEL1DESC, A.DATE; QUIT; Regards, Florent ... View more

Hi, You could also create a remote library pointing the WORK library of your server, and then use the datasets procedure in order to move all datasets from that remote library to a local library. Code is to be executed on the local session (outside of the RSUBMIT/ENDRSUBMIT) and before signing off your connection with the server. libname rwork server= slibref=work; proc datasets nolist; copy in=rwork out=work; quit; run; I hope it helps. Regards, Florent ... View more

Hi Tommy, You should first make sure there is a spawner running on the Unix server. This is probably why the TCP link establishment failed (see the ERROR in your log). Also I don't really get why you are creating a filename if you prompt for the username when signing on (see the WARNING in your log). Regards, Florent ... View more

Hi, You should have a look at the CALL EXECUTE statement. In your example it would be used that way: data Temp_Email; set temp.email_list; format Email $quote150.; CALL EXECUTE('%Send_Mail('||Email||')'); run; I hope it helps. Regards, Florent ... View more

Thanks to both of you for the feedback ! Regards, Florent ... View more

Hi Sachin, I don't know if this is what you are really expecting but I've created the following sample program in order to show what are the possibilities. %macro get_calcs(time_interval, cnt_intervals); %global %quote(&cnt_intervals); %local word; %let %quote(&cnt_intervals)=1; %let word=%qscan(&time_interval,&&&cnt_intervals,%str( )); %do %while(&word ne); %global word&&&cnt_intervals; %let word&&&cnt_intervals = &word; %*put %superq(word&&&cnt_intervals); %let %quote(&cnt_intervals)=%eval(&&&cnt_intervals+1); %let word=%qscan(&time_interval,&&&cnt_intervals,%str( )); %end; %let %quote(&cnt_intervals)=%eval(&&&cnt_intervals-1); %*put "&time_interval" contains &&&cnt_intervals words.; data test; set dummy; %do i=1 %to &&&cnt_intervals.; gross_combined_ratio_%left(&&word&i) = (inc_%left(&&word&i) + comearn_%left(&&word&i)) / earn_%left(&&word&i); %end; run; %mend get_calcs; data dummy; inc_month = 1; inc_month3 = 2; inc_annual = 3; inc_ytd = 4; comearn_month = 5; comearn_month3 = 6; comearn_annual = 7; comearn_ytd = 8; earn_month = 2; earn_month3 = 2; earn_annual = 2; earn_ytd = 2; run; %get_calcs(month month3 annual ytd, cntWrd); I hope it helps. Regards, Florent ... View more

Hello JFroberg, Could you please execute the following code and let me know whether it fits your needs or not? Watch out: I did not check if all cases are taken into account but I think it could be a good starting point. proc sort data=sashelp.shoes out=shoes_1; by region product subsidiary; run; data shoes_2; set shoes_1; by region product subsidiary; id = _N_; run; proc sql; create table shoes_3 as select distinct t1.region, t1.product, trim(left(t1.subsidiary)) || ' ' || trim(left(t2.subsidiary)) as Combinations from (select id, region, product, subsidiary from shoes_2) t1 , (select id, region, product, subsidiary from shoes_2) t2 where t1.region = t2.region and t1.product = t2.product and t1.subsidiary >= t2.subsidiary and t1.id > t2.id; quit; I hope it helps. Regards, Florent. ... View more

Hello, You should use the %by option of the %do loop. Your macro becomes: %macro RunMP; %do RunMP = 1 %to 5 %by 3; %let p1=%eval(&RunMP); %let p2=%eval(&RunMP+1); %let p3=%eval(&RunMP+2); %put &p1 &p2 &p3; %end; %mend; %RunMP *; You should also keep in mind that if you have a number of elements to display which is not a multiple of 3, you could see unresolved macro-variables displayed (text '&p1', '&p2' or '&p3' ) in your log... So you should check that each macro-variable you want to display exists prior including its reference in the %put statement. I hope it helps. Regards, Florent ... View more