Here is some additional documentation:   https://blogs.sas.com/content/sgf/2015/02/13/sas-macro-variables-how-to-determine-scope/   ... View more

I am not sure how realistic this example is. In order to be able to compare lengths of variables across data sets, you need to be able to combine them first. I have done that by creating an ID variable and used that on a BY statement in a Merge. However, this assumes that the data sets have the same number of observations and that the X values you are comparing are on the same observation. Is this the case in the actual data? Here is a starting point that combines the data, renaming the X variables and assigning the length. If this is the case with the real data, then you can add IF logic to compare lengths and then reassign values. /* Create dataset1 */ data dataset1; input Product $ x $15.; datalines; via1 . via2 003 via3 014 via4 GA4 via5 GA015 via6 319 via7 23456 via8 10101010198765 ; run; /* Create dataset2 */ data dataset2; input Name $ x $15.; datalines; a 2 b 3 c 14 d 4 e 15 f GF319 g 23456 h 98765 ; run; data dataset1; set dataset1; id=_n_; rename x=x1; run; data dataset2; set dataset2; id=_n_; rename x=x2; run; data combine; merge dataset1 dataset2; by id; len1=length(x1); len2=length(x2); run; proc print data=combine; run; ... View more

Here is another possible way to do this: data have; input tbl $ field $; cards; tbl1 x tbl1 y tbl1 z tbl2 w tbl2 x tbl2 v tbl2 y ; run; data have; set have; by tbl; if first.tbl then flag=0; flag+1; run; proc sort data=have; by flag tbl; run; proc transpose data=have out=want(drop=flag _name_); by flag; id tbl; var field; run; proc print data=want; run; ... View more

From an efficiency/resource perspective, I would not expect a significant difference between sorting by 4 numeric variables or one character one. By "same length" I should clarify the same number of digits. A numeric length of 8 in SAS can mean up to 15 significant digits. If the numeric variable values are not the same number of digits, then you can get a different order from 4 numeric variables versus one concatenated character one. And I think order does matter based on this part of your comment: "most important that groups define by 4 columns be together for the next data analysis" As others have mentioned, testing this on your part is needed because only you can determine if you are getting the order you expect.   ... View more

Are the numeric values all the same length? If they are not, then you could get a different sort order comparing the numeric sort to the character sort. Consider the following example. data test; input x w z r; cards; 1 11 111 2 1 2 3 4 ; run; data test; set test; cat=cat(x, w, z, r); run; proc sort data=test out=nums; by x w z r; run; proc print data=nums; run; proc sort data=test out=chars; by cat; run; proc print data=chars; run; Remember that when sorting numeric values in a character variable, the 1's will sort before the 2's, etc. So 10 might come before 2 as in the following example: data test; input numc $; cards; 1 2 10 ; run; proc sort data=test out=sorted; by numc; run; proc print data=sorted; run; ... View more

You can try ABSOLUTE_COLUMN_WIDTH. The following will make all column widths 20. If you want to customize each column width, you would need to add more values separated by commas for each column. ods excel file='test.xlsx' options(absolute_column_width='20'); ... View more

There is also a SHEET_INTERVAL='NOW' option as described below:   https://support.sas.com/kb/57/766.html   ... View more

You can add a sum statement to accumulate the total. Is this what you are trying to do: data new; set orig_data; IF ICD = 10 then do; IF "A00" <= substr(death_cause,1,3) <= "A99" then do; IF substr(death_cause,1,3) in ("A20","A21","A25") then do; deathcause_2=2; end; Else deathcause_1=1; End; end; IF ICD = 9 then do; IF "001" <= substr(death_cause,1,3) <= "139" then do; IF substr(death_cause,1,3) in ("032","033","034") then do; deathcause_2=2; end; Else deathcause_1=1; End; end; total_deathcause_2+deathcause_2; total_deathcause_1+deathcause_1; run; ... View more

I realize there is already a solution for this, but the reason for the difference is the use of Class variables in PROC TABULATE. If any of the variables on the Class statement have a missing value, regardless of whether they are used in the TABLE statement, then those observations will be excluded from analysis. You can add the MISSING option to the PROC TABULATE statement to include those observations and see where the missing values exist: proc tabulate data=d202_raw MISSING;   In your PROC UNIVARIATE code, you are only using a VAR statement, so all observations are included in the analysis. ... View more

If you want the total non-missing count for each variable, the sample below may also provide an alternative: https://support.sas.com/kb/44/124.html   ... View more

Just to clarify, the default statistic for analysis variables in PROC REPORT is Sum. ... View more

You can also use macro logic to do this. Since your variables are numeric, it helps to define one as DISPLAY so that you get a row for every observation. Also in the Compute block, you must use the .sum suffix on the numeric variables if you are not defining them with another usage.   data have1; input CustID Wealth2501 wealth2502 wealth2503 Ind2501 ind2502 ind2503; cards; 1 10 20 30 1 0 0 2 30 20 10 0 1 1 3 50 40 20 0 1 1 ; Run; proc sql noprint; select name into :varlist separated by ' ' from dictionary.columns where libname='WORK' and memname='HAVE1' and upcase(name) like 'IND%'; quit; %put &varlist; %let count=&sqlobs; %put &count; %macro loop; ods listing close; ods excel file='test.xlsx'; proc report data=have1; column _All_ dummy; define custid / display; define dummy / computed noprint; compute dummy; %do i=1 %to &count; if %scan(&varlist,&i,%str( )).sum=1 then call define("%scan(&varlist,&i,%str( )).sum", 'style','style=[background=red]'); %end; endcomp; run; ods excel close; ods listing; %mend; options mprint mlogic symbolgen; %loop ... View more