st: indexing for first non-missing value

[James Bernard <jamesstatalist@gmail.com>]

Hi all,

I have been dealing with a dataset where I need to copy the value of
the first non-missing value of one variable to create a new variable
called first year observed:

id             year                                 first year observed
-------      ------------
-----------------------------
A              .                                               1996
A              1996                                          1996
A              1997                                          1996
A              1998                                          1996
A                 .                                             1996
A                 .                                             1996
B               1998                                         1998
B               1999                                          1998
B                   .                                            1998
B                   .                                            1998


Going through Stata discussions, I found out that one way is to use indexing

bysort id: gen first_year_observed=year[1]

where the bracket indicates the first non-missing value

But this find only the first value, be it missing or non-missing. Is
there any simple indexing way?


Thanks,
James
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/