All,
Although I've discovered that -xtgee- models may be a solution to my
problems (posted earlier), my panel is unbalanced and I have gaps in the
time series, which means that specifying an AR-1 structure is invalid, for
various reasons. Thus, I have tried to specify a -fixed- correlation
structure. I used Shannon Driver's -corrmat- package (downloadable from
-ssc-, I believe), saved the correlation matrix as -lib-, and then:
. xtgee edldmch edround2 edround3 edtime edtimesq edyear edpollch lagconch
laglabch lagldmch clmargin ldmargin conplace labplace ldmplace edenp class
wales prvnorth midlands swest if edmarker==1, corr(fixed lib) robust
correlation matrix must be 1676x1676
r(198);
...and then...
. set matsize 11000
op. sys. refuses to provide memory
r(909);
. set matsize 10000
op. sys. refuses to provide memory
r(909);
Difficult one. I use Stata 8.2 (up-to-date on Windows 2000).
CLIVE NICHOLAS |t: 0(044)191 222 5969
Politics |e: [email protected]
Newcastle University |http://www.ncl.ac.uk/geps
*
* 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/