### 10-03-03 Deleted bug in calculation of Log Likelihood sum

A mistake was discovered in the calculation of the Log
Likelihood sum of the CoupModel.

The two possible methods availabe based on the Gaussian
distribution or a non Gaussian distribution according to Sivias

include both a constant value

-0.5*Log(2*PI)

This constant has not been defined in version of the
CoupModel prior March 3th. This mean that the LogLikelihood values that

are reported are to high = n*the constant. The good news is
that this does not have any direct impact on the calibration. However

for comparision with other calculation we have to include
this constant number.