Buzz Burhans
> I'm having trouble accomplishing the following:
>
> In a dataset with multiple panels that is tsset, and I have
> three event
> variables containing integers in the same scale as the time
> variable (days); for any given panel either 1, 2 or all 3 of these
> variables may be
> missing. I have programmed a routine whereby I repeatedly
> generate xtline
> graphs to inspect these panels values, and would like to
> use the 1,2, or 3
> values ( if not missing) in the three event variables as
> xline() values in
> those xtline graphs. I have been trying to get the three
> values into a
> macro which I could use inside the xline option, but have been
> unsuccessful. Can anyone suggest how to capture three
> values of three
> variables in a single macro? Or, alternatively, another
> way to get the
> xtline values I need in a manner that can be automated to
> reflect the unique values for each panel?
Buzz later found another solution;
nevertheless, a direct solution may be of interest.
As I understand it, Buzz has something like this
id
time
event1
event2
event3
and for each panel -event1-, -event2-,
-event3- contain the times of events 1 to 3
if and only if those events occur
and (presumably) there is at least one
other response variable.
I'd try something like this:
levels id, local(levels)
foreach l of local levels {
// next line clears macro
local events
forval i = 1/3 {
su event`i' if id == `l', meanonly
if r(max) < . local events "`events'`r(max)' "
}
if "`events'" != "" {
xtline response if id == `l', ///
subtitle(panel `l') xline(`events')
}
else xtline response if id == `l', subtitle(panel `l')
more
}
Nick
[email protected]
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
