Stata Recoding Variable Macro

[Emily Janio]

Hi all,

I have a numeric variable, range is about from 100-400,000. I would like to recode this variable to group it into ranges of 1,000.
For example:
gen newvar = .
replace newvar = 1 if var >=100 & <1000

However, this will take a lot of time. I understand that a macro might help but I have zero experience with macros. I would appreciate any help in this. Thank you.

[Mike Lacy]

Here are a couple of ways to do this:

// Example data
clear
set seed 14876
set obs 1000
gen long x = trunc(100 + (400000-100) * runiform())
// When stumped, always check out -help egen-
egen trythis = cut(x), at(0(1000)500000) icodes  //500 to be safe
replace trythis = trythis + 1
// Do it yourself
gen trythat = 1+ trunc(x/1000)

[Nick Cox]

By "macro" you mean perhaps routine or script or something similar. Stata's sense of the term macro is different, with two flavours local and global.

Anyhow, I suggest one line of code

gen newvar = 1000 * ceil(var / 1000) 

The prefactor 1000 is optional but it makes bins self-describing;

See https://www.stata-journal.com/articl...article=dm0095 for an overview.

[Joro Kolev]

Apart from what Mike and Nick showed (which are the easy and more natural Stata ways to accomplish the task), the loop is in green below

. clear

. set obs 100000
number of observations (_N) was 0, now 100,000

. gen runiformint = runiformint(100,400000)

. gen newvar = .
(100,000 missing values generated)

. qui forvalues i = 1000(1000)400000 {
. replace newvar = `i' if runiformint<`i' & missing(newvar)
. }