Help with problem in creating variable that describes one event occurring at least once

[Rui Agostinho]

Hello. I'm new to this forum, and quite new to Stata in general, as I just started using it for the work in my thesis.

I have come across a problem that I can't seem to solve. In my data, i have, among a lot more data, a collection of firms and data about if they innovated or not. Something like this:


Year ID innovation
2007 1 0
2008 1 23
2009 1 0
2007 2 45
2008 2 48
2007 3 0
2008 3 0
2009 3 0
2007 4 34
2007 5 0

What I need to do is to create a variable, innovator, that would be 0 if a firm never innovated, and 1 if it innovated at least one time during the time it is in the database. In this case, innovator would be 1 for company 1, 2 and 4 and it woukd be 0 for company 3 and 5.
How can I solve this? I've tried to do

generate innovator = 0
replace innovator = 1 if innovation >0

but this will give values of innovator==1 only for the year in which the firm innovated, and that's not what I need.

Any help would be much appreciated !

[Rich Goldstein]

there are different ways to do this; one "easy" way to just continue with what you have done is:

egen byte innovator2=max(innovator), by(ID)

you could then drop your innovator variable if you want

[Nick Cox]

A direct way to proceed is

bysort ID (innovation) : gen wanted = innovation[_N] > 0

although this needs modification if missing values are present. That's why the egen route recommended by Rich is justifiably often considered better.