SAS Macro Problem

[ballardw]

Data  not pictures. I'm not going to create data from pictures to write code.

Instructions here: https://communities.sas.com/t5/SAS-Communities-Library/How-to-create-a-data-step-version-of-your-dat... will show how to turn an existing SAS data set into data step code that can be pasted into a forum code box using the <> icon or attached as text to show exactly what you have and that we can test code against.

[FreelanceReinh]

---

@cougar300 wrote:

Why is &&Name&i resolving fine for the put statement, but not for the if then comparison line?

---

Hi @cougar300,

While the main problems with your code have been pointed out already, let me just answer the above question: In the case of &&Name&i resolving to A/O Stuff the reason was that the macro processor interpreted the slash in "A/O Stuff" as the division operator, which is valid only with numeric operands, not with character operands such as A and O. This is what the error message "A character operand was found in the %EVAL function or %IF condition where a numeric operand is required." referred to. (This would not have occurred with &&Name&i resolving to Brand A, which is what your initial post suggested.) It is indeed somewhat confusing that the error message ("The condition was: brand = &&Name&i") does not resolve the macro variable reference and thus hides the culprit (the slash in this case).

If the comparison made sense in an %IF condition, double quotes or macro quoting functions (such as %BQUOTE) would need to be applied in order to avoid the wrong interpretation of the slash.

[cougar300]

Thank you

the "/" making it attempting a numeric conversion makes sense.  That clarifies that part, at least.

[cougar300]

If anyone can easily turn this starting data into the blue part shown in the above picture, please let me know.

thanks

data WORK.start;
infile datalines dsd truncover delimiter=',';
input Brand:$8. Amount:32. Ratio:32.;
datalines;
Brand A,20,.4
Brand B,16,.5
Brand C,12,.25
Brand D,8,.25
Brand E,14,0
;

[Tom · Accepted solution]

Like this:

data WORK.start;
  infile datalines dsd truncover ;
  input Brand $ Amount Ratio ;
datalines;
Brand A,20,.4
Brand B,16,.5
Brand C,12,.25
Brand D,8,.25
Brand E,14,0
;
proc sql;
  create table want as 
  select a.brand,a.amount,a.ratio
       , b.brand as brand2,b.amount*a.ratio as new_amount
  from start a
     , start b
  ;
quit;
proc report data=want ;
  columns brand amount ratio new_amount,brand2;
  define brand / group;
  define amount / group;
  define ratio / group;
  define brand2 / across ' ';
  define new_amount / sum;
run;

[cougar300]

Thank you so much!

If I want this to be different by store, I assume I can just add "by store" in the SQL?

Either way.

Thanks!

[cougar300]

One other Q if you are here - 

The report is what I want, but I need it in a dataset.

When I put out=dataset into the proc report, everything is there, but instead of the var names being Brand A through Brand E, the var names are _C4_ through _c8_.

Can I get a dataset to look like the report does?

Thanks again!

[Tom]

Why would you want a dataset with data (BRAND name) in metadata (variable name)?
Use PROC TRANSPOSE.

[cougar300]

I will try to take the output and make it what I need it to be.

The reason for the var name is that I need to sum each of them across all stores, so if one store has 5 brands and another store has a different 5 brands I cant have them both being labeled c4-c8.  There can be from 1-20 brands and in the end I need to sum across all of them, so having the actual brand there would be best.

[Tom]

Nope. Leave the data in the tall format and that issues doesn't exist.