The effective dimension of the simulation is the number of deviates needed to compute a single observation of the random quantity. If as usual the deviates are obtained by application of the inverse cumulative normal distribution to independent uniform deviates in , these uniform deviates must be equidistributed in for Monte Carlo methods to be theoretically sound.
If the dimension is high random number generators might not (do not) have this property. In this case one is well advised to rely on low discrepancy sequences which have this property in all dimensions.
This file declares the interface for a stochastic driver and some drivers which we need for our stochastic processes. We have drivers generating pseudorandom numbers based on the Mersenne Twister and drivers generating quasirandom numbers based on the Sobol sequence.
Definition in file StochasticGenerator.h.
#include "TypedefsMacros.h"
#include "Utils.h"
#include "Random.h"
#include "Matrix.h"
Go to the source code of this file.
Compounds | |
class | StochasticGenerator |
class | MonteCarloLiborDriver |
class | SobolLiborDriver |
class | MonteCarloVectorDriver |
class | SobolVectorDriver |
class | MonteCarloScalarDriver |
class | SobolScalarDriver |