Nick,
Thanks for the solution below: it worked! (As did the comma-less -egen
max- solution suggested by Matt and yourself - that was a very *slick*
piece of code!).
The time variable *did* generate and *did* tsset (although the time
variable I'd generated previously was coded in exactly the same way and
yet didn't -tsset-: strange!). I've also generated a 'year' variable.
However, -xtpcse- still won't run:
no time periods are common to all panels, cannot estimate disturbance
covariance matrix using casewise inclusion
r(459);
It's not fair to bug the list for the umpteenth time on this ongoing saga,
so I'll admit defeat here and continue running FGLS models. Professor Beck
and Dr Katz won't be happy!! *lol*
Thanks once again! :-)
C.
> No reason why it should.
>
> . bysort group : gen time = _n
>
> will in general produce quite
> arbitrary values for -time-,
> just as would
>
> . sysuse auto
> . bysort rep78 : gen time = _n
>
> You're more likely to want something like
>
> . bysort group (year) : gen time = _n
>
> That is, for this to make sense, some
> other variable must be available
> to indicate the order _within_ groups.
>
> Nick
> [email protected]
Yours,
CLIVE NICHOLAS,
Politics Building,
School of Geography, Politics and Sociology,
University of Newcastle-upon-Tyne,
Newcastle-upon-Tyne,
NE1 7RU,
United Kingdom.
*
* 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/
