I've got several distribution objects in Matlab like this one:
% gravity [m/s2]
Uncer.Param.gravity.LB = 9.801;
Uncer.Param.gravity.UB = 9.867;
Uncer.Param.gravity.pd = makedist('Uniform','Lower',Uncer.Param.gravity.LB, 'Upper', Uncer.Param.gravity.UB);
Uncer.Param.gravity.value=0;
I know I can generate a random sample with the random-function, but I want to generate a sample made of quasi-random-numbers (Sobol).
I get a matrix filled with these quasi-random-numbers like this:
set = net(sobolset(countParameter*2, 'Skip',1), countSimulation);
And I know furthermore, that I can interpolate the distribution values with the function interp1 and the corresponding CDF.
The problem is that my matrix-dimensions are round about 1000x20 and the interpolation would cost a ton ammount of time.
Is there a faster way to do this?
Aucun commentaire:
Enregistrer un commentaire