Dividing variables

[ricardoabe]

Hello everybody.

I am trying to find a way to divide variables in a sequenced/ordered way.

incomeusa	incomechina	incomecanada	incomefrance	limitusa	limitchina	limitcanada	limitfrance
6000	1265	3776	0	15	23	2	13
4356	6785	3577	10089	18	4	34	11
7643	4456	1322	678	9	5	5	24
1087	365	0	778	8	8	14	11
4456	44	5789	676	6	6	17	1

I have to divide limit variables and income variables per country.

I tried 3 different commands but none of them worked properly.

1.Command: Don't create any new variable.

foreach var of varlist incomeusa-incomefrance {
foreach y of varlist limitusa-limitfrance {
generate `var'dif = `y' / `var'
}
}

incomeusadif already defined
r(110);
end of do-file

2. Command: Create variables but all income variables are using just limitusa as numerator

foreach var of varlist incomeusa-incomefrance {
foreach y of varlist limitusa-limitfrance {
generate `y'dif = `y' / `var'
}
}

limitusadif already defined
r(110);
end of do-file


3. Command: Tried to use the collon before the divison signal, but didnt worked either.

foreach var of varlist incomeusa-incomefrance {
foreach y of varlist limitusa-limitfrance {
generate `y'dif = `y' :/ `var'
}
}

limitusa: invalid name
r(198);
end of do-file

Anyone has an idea how to perfom division between variables in a sequenced way?

Thank you,
Ricardo.

[Jeph Herrin]

I think you want to retrieve the country name each iteration, and use that:

foreach var of varlist incomeusa-incomefrance {
        local country=subinstr("`var'","income","",1)
        gen diff`country'=limit`country'/income`country'
}

hth,
Jeph

[ricardoabe]

It worked.
Thank you!
Ricardo.

[Rachel Chan]

Hi, I have a slightly similar problem with this. I have variables Jan_13 to Mar_21, referring to months January 2013 to March 2021. I need to create new variables containing the results of current month/previous month. Ex. Feb_13_PR = Feb_13/Jan_13.

How can I automate this computation for all months through looping? Thank you.

[Nick Cox]

#4 was repeated at https://www.statalist.org/forums/for...hrough-looping