Creating new variables by dividing two others that include zero

[Nanna Brandt]

Hi,

I have the following table, which is a little output from the data that I have.

I would like to create a new variable (new_var_5) by dividing var_1 and var_2 and also create another new variable (new_var_6) by dividing var_3 and var_4. I'm doing this so I'll be able to run different t-tests afterwards comparing the mean of new_var_5 and new_var_6. I know how to create new variables by dividing the two others, but the problem is that there are several zero-values in the columns that I want to divide resulting in a lot of missing data.

Therefore, my question is - how would you deal with this, so I'll be able to compare the means of new_var_5 and new_var_6 without missing a lot of data that is still relevant for my research-question?

var_1	var_2	var_3	var_4	new_var_5	new_var_6
2	4	2	2	0.5	1
5	10	5	20	0.5	0.25
0	10	0	5	.	.
10	10	0	2	1	.

Thank you very much!

Regards Nanna

[Ken Chui]

Could you double check your math? 0 divided by 10 is equal to 0, not undefined, so the result should not be missing:

clear
input v1 v2 v3 v4
2 4 2 2
5 10 5 20
0 10 0 5
10 10 0 2
end

gen v5 = v1/v2
gen v6 = v3/v4

list

Results:

     +------------------------------+
     | v1   v2   v3   v4   v5    v6 |
     |------------------------------|
  1. |  2    4    2    2   .5     1 |
  2. |  5   10    5   20   .5   .25 |
  3. |  0   10    0    5    0     0 |
  4. | 10   10    0    2    1     0 |
     +------------------------------+

[Nanna Brandt]

Thank you vey much for your answer.

You are of course right, I'm sorry. However, I also meant it the opposite way around. Sometimes var_2 is 0 and var_1 is 10. This will give a missing value.

var_1	var_2	var_3	var_4	new_var_5	new_var_6
2	4	2	2	0.5	1
5	10	5	20	0.5	0.25
10	0	0	5	.	0
10	10	2	0	1	.