I see your problem, but I believe your solution is quite
disproportionate. As said, the consequences
for other programs and documentation would be
considerable. Of course, that's not your problem,
but it certainly would add to the disclination
at StataCorp to do this, and I guess that is
massive already.
If you have this problem repeatedly, a fix
should be programmable, so that with a single
command line you can fill in the gaps after -tsfill-.
I guess it would be much easier to do that for you
than to program your solution.
Nick
[email protected]
Charles Thibault
> The reason mainly revolves around tsfill. When doing -
>
> egen groupvar, group()
>
> tsset groupvar timevar
>
> tsfill, full
>
> - the information that identifies the panels, for example
> counties and
> households jointly, is lost.
>
> The information needs to be replaced manually according to
> the method you
> described, namely:
>
> replace county = county[_n-1] if mi(county)
> replace household = household[_n-1] if mi(household)
>
> Allowing multiple variables to identify the panel would avoid
> this step and
> the first egen command, which not only saves time but also
> avoids potential
> danger for a novice if the user first sorts on county and
> household and THEN
> replaces the missing data.
*
* 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/
