Create table_ Use proc tabulate or something else?

[Chris_Loke]

Hello!

I have a dataset (simplified example) containing the following variables;

Data salary;

input age_group$ gender$ year$ salary;

25-29 M 2017 25000

25-29 M 2018 30000

25-29 W 2017 30000

25-29 W 2018 35000

30-34 M 2017 25000

30-34 M 2018 50000

30-34 W 2017 25000

30-34 W 2018 75000

I would like to create a table looking like this:

	25-29 (age_group)	30-34 (age_group)
	Gender	Gender
	M	W		M	W
	Salary	Salary		Salary	Salary
	Sum	Sum		Sum	Sum
2018	64000	35000		50000	75000
2017	34000	30000		25000	25000

M and W should be under agegroup and gender, not separeted ( there was an error in the HTML when I made this post).

Same table but from notepad;

  25-29 (age_group) 30-34 (age_group)
  Gender Gender
  M W   M W 
  Salary Salary   Salary Salary 
  Sum Sum   Sum Sum 
2018 64000 35000   50000 75000 
2017 34000 30000   25000 25000 

 I have tried proc tabulate and proc transpose, but probably I am doing somethin wrong as I don't get the result I want.  Would be super if someone could provide a code for this.

[PeterClemmensen]

The sums you posted are not your actual wanted sums, right?

[ed_sas_member · Accepted solution]

Hi @Chris_Loke 

You can try this:

proc tabulate data=salary;
	var salary;
	class age_group gender year;
	table year, (age_group*gender) * (salary*sum);
run;

[Chris_Loke]

Thanks!

It worked!

[ballardw]

---

@Chris_Loke wrote:

Hello!

M and W should be under agegroup and gender, not separeted ( there was an error in the HTML when I made this post).

Perhaps use a plain text editor, the SAS editor will work, to outline the positions instead of HTML. Paste into a code box opened on the forum using the {I} icon. There is more than one way to interpret your above requirement and some of the interpretations could be kind of wonky.