Linear-Feedback Shift Register Generators

Detailed Description

An implementation of the small linear feedback shift registers (LFSR) for use in a Markov chain quasi-Monte Carlo context, as described in S. Chen, M. Matsumoto, T. Nishimura, and A. Owen: "New inputs and methods for Markov chain quasi-Monte Carlo", and S. Chen: "Consistence and convergence rate of Markov chain quasi-Monte Carlo with examples". Apart from the matrix that describes the recurrence, this class is stateless, and can therefore be easily shared in a multi-threaded context.

https://en.wikipedia.org/wiki/Linear-feedback_shift_register

Classes
class	cugar::LFSRGeneratorMatrix
struct	cugar::LFSRRandomStream

Functions
CUGAR_HOST_DEVICE CUGAR_FORCEINLINE	cugar::LFSRGeneratorMatrix::LFSRGeneratorMatrix (uint32 m, Offset_type offset_type=GOOD_PROJECTIONS)
CUGAR_HOST_DEVICE CUGAR_FORCEINLINE float	cugar::LFSRGeneratorMatrix::next (uint32 scramble, uint32 *state) const
CUGAR_HOST_DEVICE CUGAR_FORCEINLINE float	cugar::LFSRRandomStream::next ()