I'm using XTGLS to estimate a panel data of about 280 cities and 360
time periods. I get a missing value for the log-likelihood, and the
xttest2 indicates that matrix has missing values. Please see regression
output below.
I'll appreciate any insights, and thanks.
John
.**************************
. *xtgls
. xtgls pmb_dev h_dev s_dev c_dev, /*
> */ i(city_id) t(edate) panels(c) corr(ar1)
Cross-sectional time-series FGLS regression
Coefficients: generalized least squares
Panels: heteroskedastic with cross-sectional correlation
Correlation: common AR(1) coefficient for all panels (0.8189)
Estimated covariances = 39903 Number of obs
= 101802
Estimated autocorrelations = 1 Number of groups
= 282
Estimated coefficients = 4 Time periods
= 361
Wald chi2(3) = 4104.11
Log likelihood = . Prob > chi2
= 0.0000
pmb_dev Coef. Std. Err. z P>z [95%
Conf. Interval]
h_dev .0000293 .0000297 0.98 0.325
-.000029 .0000875
s_dev -8.364303 .1310437 -63.83 0.000
-8.621144 -8.107463
c_dev .1018725 .0359491 2.83 0.005
.0314136 .1723313
_cons -.0083858 .2799171 -0.03 0.976
-.5570133 .5402417
. xttest2
matrix has missing values
r(504);
end of do-file
r(504);
*******************
*
* 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/
