It means 'not equal to' ... View more

Good to know, If you are satisfied with the solution and if it actually worked could you mark this solution as accepted so it will help other sas community users. ... View more

Please try the untested code proc sort data=have; by family name descending code; run; data want; set have(rename=(code=code2)); by family name; retain code; if first.name then code=.; if code2 ne . then code=code2; run;   ... View more

You can try below code   data welcome; infile datalines dlm=","; input ID $ DeptID $ Name :$20. svcdate :yymmdd10. Fiscal_Year Hours Amount ; format svcdate yymmddd10.; datalines; 1234, 7, East, 2019-01-01, 2019, 1.50, 109.7 1234, 7, East, 2019-01-01, 2019, 7.50, 509.7 6543, 7, East, 2019-01-1, 2019, 2.75, 340.82 7847, 8, West, 2019-02-01, 2019, 5, 7000 5818, 10, North, 2019-03-01, 2019, 6, 1032 0844, 3, South, 2019-04-01, 2019, 40, 855 4499, 5, West, 2018-09-01, 2019, 10, 790 1234, 7, East, 2019-09-01, 2019, 6.50, 1900.7 3000, 2, North, 2018-12-01, 2019, 11.75, 874.79 3785, 27, West, 2018-11-01, 2019, 20.3, 450.64 ; run; proc sql; create table want as select DeptID, Name, svcdate, Fiscal_Year , sum(Hours) as hours, sum(amount) as amount, count(id) as count from welcome group by DeptID, Name, svcdate, Fiscal_Year ; quit; ... View more

Are you fine with this   legend1 position=(right bottom inside) across=1 label=none shape=bar(.10in,.10in) mode=protect; proc gchart data=sashelp.prdsale; hbar quarter / DISCRETE type=sum sumvar=actual sum sumlabel='Counts' descending group=product subgroup=region maxis=axis1 raxis=axis2 gaxis=axis3 legend=legend1 noframe PATTERNID=subgroup ; run;   ... View more

Sorry I am not clear, could you please give more details ... View more

Hope you are fine with the below, used annotation dataset as below   goptions reset=global; data barlabel; length function color $8; retain xsys ysys '2' color 'magenta' when 'a'; set sashelp.prdsale; function='label'; midpoint=region; group=product; subgroup=quarter; text=' '||strip(put(quarter,best.)); run; legend1 position=(right bottom inside) across=1 label=none shape=bar(.10in,.10in) mode=protect; proc gchart data=sashelp.prdsale; hbar region / DISCRETE type=sum sumvar=actual sum sumlabel='Counts' descending group=product subgroup=quarter maxis=axis1 raxis=axis2 gaxis=axis3 legend=legend1 noframe PATTERNID=midpoint annotate=barlabel; run;     ... View more

Please check now, I used the subgroup=quarter in the code below   legend1 position=(right bottom inside) across=1 label=none shape=bar(.10in,.10in); proc gchart data=sashelp.prdsale; hbar region / DISCRETE type=sum sumvar=actual sum sumlabel='Counts' descending group=product subgroup=quarter maxis=axis1 raxis=axis2 gaxis=axis3 legend=legend1 noframe PATTERNID=midpoint ; run;   ... View more

If you are expecting the graph as below then try the code   legend1 position=(right bottom inside) across=1 label=none shape=bar(.10in,.10in); proc gchart data=sashelp.prdsale; hbar region / DISCRETE type=sum sumvar=actual sum sumlabel='Counts' descending group=product maxis=axis1 raxis=axis2 gaxis=axis3 legend=legend1 noframe PATTERNID=midpoint ; run;   ... View more

Here is a dynamic approach by which we can derive n number of variables as you expected   data have; input T1 $ T2 T3 $; datalines; Alice 10 Female Tim 30 Male Bob 50 Male ; proc sql noprint; /*create macro variables for character and numeric variables*/ select name into: chars separated by ' ' from dictionary.columns where libname='WORK' and memname='HAVE' and type='char' order by type, name; select name into: nums separated by ' ' from dictionary.columns where libname='WORK' and memname='HAVE' and type='num' order by type, name; /*create macro variables for character and numeric variables with names changed i.e to have names like T_1... instead of T1*/ select prxchange('s/(\w[a-z]*)(\d*)/$1_$2/i',1,name) into: newn separated by ' ' from dictionary.columns where libname='WORK' and memname='HAVE' and type='num' order by type, name; select prxchange('s/(\w[a-z]*)(\d*)/$1_$2/i',1,name) into: newc separated by ' ' from dictionary.columns where libname='WORK' and memname='HAVE' and type='char' order by type, name; quit; %put &chars &nums &newn &newc; data want; set have; array var1(*) &chars; array var2(*) &nums; array var3(*) $100. &newn; array var4(*) $100. &newc; do i = 1 to dim(var1); var4(i)=catx(' - ', vname(var1(i)), var1(i)); end; do i = 1 to dim(var2); var3(i)=catx(' - ', vname(var2(i)), var2(i)); end; run; ... View more