Aggregate data by TWO groups

[Tobias Winkelmann]

Hey everyone, I'm new to Stata and have a question.

I need to aggregate the data by year and company. With the collapse function, I can only sum up the values for the companies for all years.




The raw data look like this:

id year companyid sales
1 2010 555 50,000
2 2010 555 46,000
3 2010 555 40,000
4 2011 555 30,000
5 2011 555 51,000
6 2010 700 2,000
7 2010 700 4,000
8 2011 700 7,000
9 2011 700 11,000
10 2011 700 5,000
.....


The result I need, should look like this:

year companyid Sum
2010 555 136,000
2010 700 6,000
2011 555 81,000
2011 700 23,000
.....


I appreciate any thoughts, many thanks!

[Nick Cox]

I can follow what is wanted, but not why you can't get it. Strictly, collapse is a command, and not a function. In Stata function is not another word for command.

clear 
input id year companyid sales
1 2010 555 50.000
2 2010 555 46.000
3 2010 555 40.000
4 2011 555 30.000
5 2011 555 51.000
6 2010 700 2.000
7 2010 700 4.000
8 2011 700 7.000
9 2011 700 11.000
10 2011 700 5.000
end 

collapse (sum) sales, by(companyid year)

list 

     +-------------------------+
     | year   compan~d   sales |
     |-------------------------|
  1. | 2010        555     136 |
  2. | 2011        555      81 |
  3. | 2010        700       6 |
  4. | 2011        700      23 |
     +-------------------------+

[Tobias Winkelmann]

thank you very much! I did not know that I can use collapse with two variables at the same time...

[Nick Cox]

Take a look at

help collapse

which includes examples to that effect.