I use MAXLIK library to estimate multiple discrete-continuous extreme value (MDCEV) model.
After convergence, I got a mean log-likelihood value, -26.4448.
But I can find what it means. Is it different from a general log-likelihood value?
Please, answer this question.
===============================================================================
MAXLIK Version 4.0.15 1/25/2016 12:55 pm
===============================================================================
Data Set: c:\my_data
-------------------------------------------------------------------------------
return code = 2
maximum number of iterations exceeded
Mean log-likelihood -26.4448
Number of cases 21750
Covariance matrix of the parameters computed by the following method:
Inverse of computed Hessian
Parameters Estimates Std. err. Est./s.e. Prob. Gradient
------------------------------------------------------------------
P01 -9.0217 0.0790 -114.164 0.0000 0.0000
P02 -9.2502 0.1231 -75.142 0.0000 0.0000
P03 -8.5399 0.2080 -41.067 0.0000 0.0000
P04 -2.5965 0.0710 -36.566 0.0000 0.0000
P05 -5.8711 0.0590 -99.581 0.0000 0.0000
P06 -5.3901 0.0616 -87.510 0.0000 0.0000
1 Answer
0
The mean log-likelihood is being used to avoid precision problems because it can get very large sometimes, and it is best to keep numbers small on a computer.
To get the log-likelihood, multiply by the number of observations.
Your Answer
1 Answer
The mean log-likelihood is being used to avoid precision problems because it can get very large sometimes, and it is best to keep numbers small on a computer.
To get the log-likelihood, multiply by the number of observations.