Hi,
I 'm trying to implement NesonSiegel in quantib but the simplex optimization does'nt give me satisfaction.
I'm using the simplex algo to get the Nelson Siegel parameters
In the cost function I put the error generated by the model ( the sum of (the difference between the market prices and the Nelson Siegel prices generated by Nelson Siegel)^2 ) but what the simplex algo is checking to get out is the Real rtol = 2.0*QL_FABS(high - low)/
(QL_FABS(high) + QL_FABS(low) + QL_EPSILON);
with high the value of the highest error generated by a paramater
and with low the value of the hlowest error generated by a paramater
And this is not reflecting the error I wanted ...
please Help !!
perhaps I'm wrong to use the simplex algo to minimize the sum of (the difference between the market prices and the Nelson Siegel prices generated by Nelson Siegel )^2.
Découvrez le nouveau Yahoo! Mail :
250 Mo d'espace de stockage pour vos mails !
Créez votre Yahoo! Mail