Could I use if-then statement in the proc expand as follows?   %macro all16; %local j_list k_list j k j_index k_index; %let j_list = 3 6 9 12; %let k_list =3 3 3 3; %do j_index=1 %to 4; %let j = %scan(&j_list, &j_index); %do k_index=1 %to 4; %let k = %scan(&k_list, &k_index); data vickyhave; input date :yymmdd10. gdp country; format date yymmdds10.; cards; 19800131 28 222 19800229 29.375 222 19800331 27.125 222 19800430 24.25 222 19800530 26.75 222 19800630 30.75 222 19800731 37.75 222 19800829 40.75 222 19800930 43.375 222 19801031 51.5 222 19801128 56 222 19801231 24.5 222 19810130 22.25 222 19810227 20 222 19810331 22.75 222 19810430 20.5 222 19810529 19.5 222 19810630 18 222 19810731 18.5 222 19810831 17.5 222 19810930 14.25 222 19811030 14.375 222 19811130 15.75 222 19811231 14.625 222 19820129 14.625 222 19820226 12.875 222 19820331 12.625 222 19820430 12.625 222 19820528 12 222 19820630 12.125 222 19820730 10.875 222 19820831 11.375 222 19820930 11.75 222 19821029 14.25 222 19821130 13.75 222 19821231 13 222 19830131 14.25 222 19830228 13.5 222 19830331 14 222 19830429 14.625 222 19830531 14.5 222 19830630 15.75 222 19830729 15.375 222 19830831 14.625 222 19830930 16.625 222 19831031 15 222 19831130 15 222 19831230 15.5 222 19840131 15.375 222 19840229 14 222 19840330 13.625 222 19840430 12.125 222 19840531 11.625 222 19840629 11.5 222 19840731 10.125 222 19840831 11.875 222 19840928 12.125 222 19841031 12.25 222 19841130 11.5 222 19841231 11.125 222 19850131 13.5 222 19850228 14.875 222 19850329 14.25 222 19850430 13.625 222 19850531 13.25 222 19850628 13.375 222 19850731 13.75 222 19850830 12.75 222 19850930 12.25 222 19851031 11.25 222 19851129 12 222 19851231 12.75 222 19860131 13.125 222 19860228 14 222 19860331 14.5 222 19860430 13.25 222 19860530 13.875 222 19860630 13.5 222 19860731 11.875 222 19860829 12.25 222 19860930 11.25 222 19861031 13.25 222 19861128 13 222 19861231 13.125 222 19870130 17 222 19870227 17 222 19870331 16 222 19870430 15.25 222 19870529 19.5 222 19870630 20.25 222 19870731 22 222 19870831 22 222 19870930 21.625 222 19871030 13.125 222 19871130 10.875 222 19871231 13.75 222 19880129 14.5 222 19880229 14.75 222 19880331 13.125 222 19880429 12 222 19880531 13.375 222 19880630 16.625 222 19880729 15.5 222 19880831 15.875 222 19880930 15.25 222 19881031 14.375 222 19881130 13.375 222 19881230 14.5 222 19890131 16.125 222 19890228 16.625 222 19890331 17 222 19890428 18.25 222 19890531 20 222 19890630 18.625 222 19890731 21 222 19890831 21.25 222 19890929 20 222 19891031 17.375 222 19891130 19.875 222 19891229 20.75 222 19800131 25.75 257 19800229 25.625 257 19800331 20.375 257 19800430 20.5 257 19800530 20.75 257 19800630 22 257 19800731 21.625 257 19800829 22.875 257 19800930 24.375 257 19801031 24.5 257 19801128 24.5 257 19801231 22.875 257 19810130 23.75 257 19810227 24.875 257 19810331 28.5 257 19810430 27.75 257 19810529 27.5 257 19810630 27.125 257 19810731 29.5 257 19810831 26.875 257 19810930 27.5 257 19811030 26.75 257 19811130 25.75 257 19811231 23.25 257 19820129 23.5 257 19820226 22.375 257 19820331 22.625 257 19820430 22.125 257 19820528 23.125 257 19820630 22.75 257 19820730 20.5 257 19820831 23.5 257 19820930 24.25 257 19821029 27.75 257 19821130 31 257 19821231 30 257 19830131 31.25 257 19830228 30.875 257 19830331 33.75 257 19830429 32.875 257 19830531 34.375 257 19830630 36.25 257 ; run; proc expand data= vickyhave out=vickywant (where=((gdp>=max_gdp_hist&j) and (centered_window_range=2*&J+1))) method=none ; by country; id date; convert gdp=max_gdp_hist&j / transformin=(lag 1) transformout=(nomiss movmax &j); convert gdp=centered_window_range / transformin=(*0 +1 cmovsum 2*&J+1 ) ;   if j=3 then convert gdp=gmean_future_rate_3 / transformin=(ratio 1 log cmovave ( 0 0 0 0 1 1 0) exp -1); else if j=6 then convert gdp=gmean_future_rate_3 / transformin=(ratio 1 log cmovave ( 0 0 0 0 0 0 0 1 1 0) exp -1); else if j=9 then convert gdp=gmean_future_rate_3 / transformin=(ratio 1 log cmovave ( 0 0 0 0 0 0 0 0 0 0 1 1 0) exp -1); else convert gdp=gmean_future_rate_3 / transformin=(ratio 1 log cmovave ( 0 0 0 0 0 0 0 0 0 0 1 1 0) exp -1); run; %end; %end; %mend all16; %all16;   ... View more

How can I convert daily returns  to monthly cumulative returns with proc expand convert? Please note that some of the daily returns might be zero. Part of the data is attached.   ... View more

How can I obtain the dataset if the requirement is as follows? If the time between one obs and the succeeding obs is shorter than 6 months, the succeeding obs will be deleted. For example the obs (19810630, 19811130, 19811231, 19820129…) will be deleted, and so on. Part of the data is as shown below: data have; input entity date :yymmdd10. sales; format date yymmdds10.; cards; entity DATE sales abc 19800731 77.875 abc 19810430 71.375 abc 19810630 71.75 abc 19811030 75.75 abc 19811130 78.25 abc 19811231 80.25 abc 19820129 80.375 abc 19820430 85.25 abc 19820831 85.625 abc 19820930 95.125 abc 19821029 98 abc 19821130 106.125 abc 19831130 55.625 abc 19840731 52.375 abc 19840831 52.625 abc 19841031 56 abc 19850731 56.875 abc 19850830 57.25 abc 19851031 57.5 abc 19851129 63.375 abc 19851231 67.125 abc 19860331 69.75 abc 19860530 73.625 abc 19860630 77.875 abc 19870130 80 abc 19870331 88 abc 19870630 92 abc 19870831 98 abc 19880930 77.875 abc 19881031 80 abc 19881230 82.625 abc 19890131 87 abc 19890331 92.125 abc 19890428 92.5 abc 19890531 95.625 abc 19890630 102.75 xyz 19800930 17 xyz 19801031 17.375 xyz 19801128 17.375 xyz 19810227 17.375 xyz 19810331 18.75 xyz 19810430 20.375 xyz 19810529 23 xyz 19810630 24.25 xyz 19820831 19.75 xyz 19820930 22.125 xyz 19821029 23 xyz 19821130 23 xyz 19830331 24.75 xyz 19830429 25.875 xyz 19830630 30.375 xyz 19830729 30.75 xyz 19830930 33.375 xyz 19850131 16.25 xyz 19850430 18.875 xyz 19850531 19.625 xyz 19850628 19.75 xyz 19850731 19.75 xyz 19850830 19.75 ; Run; ... View more

How can I create 25 my_class&j&k files, if one of the file (my_class36) from (j,k)=(3,6) is as follows?   %LET j=3; %LET k=6; data my_class&j&k; set sashelp.class; height&j=height*&j; weight&k=weight*&k; run;   There 25 combinations of j and k, if j=1,3,6,9, & 12 and k=1,3,6,9,& 12. ... View more