Counting distinct values by group

[Nacho Man]

Hello,

I have a data-set with a unique id (permno) and time (date). I have taken the year of this date using year(date) to create a new variable. What I want to do is calculate a variable called count, which counts the number of distinct 'years' by the permno.
I have an example attached below. My original data is exactly the same, just without the "count" column.

I tried using

by permno (year): gen count = _n
but it is giving me a count for each observation, despite being in the same year.


Intended output:

permno date year count
10001 281986 1986 1
10001 311986 1986 1
10001 301986 1986 1
10001 301986 1986 1
10001 301986 1986 1
10001 311986 1986 1
10001 301987 1987 2
10001 271987 1987 2
10001 311987 1987 2
10001 301987 1987 2
10001 291987 1987 2
10001 291988 1988 3
10001 291988 1988 3

Any help is appreciated.

Thanks

[Nick Cox]

In Stata count is called a variable; calling it a column is spreadsheet-speak best reserved for private use.

egen wanted = group(permno year)

should help.

In your earlier thread we reminded you of the strong preference here for full real names.

[Nacho Man]

Thanks for your response Nick. and apologies, but I'm unaware of how to change my name; will I have to make a new account?

[Nick Cox]

No; see the FAQ Advice for how to contact the administrators.