Dear Stata-users,
I am running an "streg" model with a Lognormal distribution and the
log-likelihood of the fitted model turns positive. I am not sure what
happened here.
Many thanks for your help in advance.
Amani
. streg haz if ~npointer2, distribution(lnormal)
failure _d: newtag2
analysis time _t: rmidobs2
Fitting constant-only model:
Iteration 0: log likelihood = -1615.5546 (not concave)
Iteration 1: log likelihood = -675.23785 (not concave)
Iteration 2: log likelihood = -214.2482
Iteration 3: log likelihood = 115.43239
...
Iteration 7: log likelihood = 178.42856
Fitting full model:
Iteration 0: log likelihood = 178.42856
Iteration 1: log likelihood = 178.51265
Iteration 2: log likelihood = 178.51266
Log-normal regression -- accelerated failure-time form
No. of subjects = 758 Number of obs =
758
No. of failures = 753
Time at risk = 197003.8033
LR chi2(1) =
0.17
Log likelihood = 178.51266 Prob > chi2 =
0.6817
----------------------------------------------------------------------------
--
_t | Coef. Std. Err. z P>|z| [95% Conf.
Interval]
-------------+--------------------------------------------------------------
--
haz | -.0032793 .0079953 -0.41 0.682 -.0189498
.0123913
_cons | 5.542043 .0070068 790.95 0.000 5.52831
5.555776
-------------+--------------------------------------------------------------
--
/ln_sig | -1.665713 .0258207 -64.51 0.000 -1.71632
-1.615105
-------------+--------------------------------------------------------------
--
sigma | .1890559 .0048816 .1797263
.1988698
----------------------------------------------------------------------------
--
*
* 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/
