Trivial Example of a Likelihood Profile
Here is a trivial example
of how the profile likelihood gives the correct interpretation of skewness. This example
can be extended for problems where projected biomass is desired. The following template
works (and demonstrates that the data file is not necessary):
DATA_SECTION
PARAMETER_SECTION
init_number rec_dev
objective_function_value f
likeprof_number recruit;
PROCEDURE_SECTION
f=4.0 * rec_dev*rec_dev
recruit=100*exp(rec_dev);
Note that the single "parameter" rec_dev has an estimate of zero and a standard deviation of 0.35 (since 1/sqrt(2*lambda) =
0.35 where lambda=4.0).
The results of this can be plotted as
follows:
 |
