QuantLib: a free/open-source library for quantitative finance
fully annotated source code - version 1.34
|
Sobol low-discrepancy sequence generator. More...
#include <sobolrsg.hpp>
Public Types | |
enum | DirectionIntegers { Unit , Jaeckel , SobolLevitan , SobolLevitanLemieux , JoeKuoD5 , JoeKuoD6 , JoeKuoD7 , Kuo , Kuo2 , Kuo3 } |
typedef Sample< std::vector< Real > > | sample_type |
Public Member Functions | |
SobolRsg (Size dimensionality, unsigned long seed=0, DirectionIntegers directionIntegers=Jaeckel, bool useGrayCode=true) | |
const std::vector< std::uint32_t > & | skipTo (std::uint32_t n) const |
const std::vector< std::uint32_t > & | nextInt32Sequence () const |
const SobolRsg::sample_type & | nextSequence () const |
const sample_type & | lastSequence () const |
Size | dimension () const |
Private Attributes | |
Size | dimensionality_ |
std::uint32_t | sequenceCounter_ = 0 |
bool | firstDraw_ = true |
sample_type | sequence_ |
std::vector< std::uint32_t > | integerSequence_ |
std::vector< std::vector< std::uint32_t > > | directionIntegers_ |
bool | useGrayCode_ |
Sobol low-discrepancy sequence generator.
A Gray code counter and bitwise operations are used for very fast sequence generation.
The implementation relies on primitive polynomials modulo two from the book "Monte Carlo Methods in Finance" by Peter Jäckel.
21 200 primitive polynomials modulo two are provided in QuantLib. Jäckel has calculated 8 129 334 polynomials: if you need that many dimensions you can replace the primitivepolynomials.cpp file included in QuantLib with the one provided in the CD of the "Monte Carlo Methods in Finance" book.
The choice of initialization numbers (also know as free direction integers) is crucial for the homogeneity properties of the sequence. Sobol defines two homogeneity properties: Property A and Property A'.
The unit initialization numbers suggested in "Numerical Recipes in C", 2nd edition, by Press, Teukolsky, Vetterling, and Flannery (section 7.7) fail the test for Property A even for low dimensions.
Bratley and Fox published coefficients of the free direction integers up to dimension 40, crediting unpublished work of Sobol' and Levitan. See Bratley, P., Fox, B.L. (1988) "Algorithm 659: Implementing Sobol's quasirandom sequence generator," ACM Transactions on Mathematical Software 14:88-100. These values satisfy Property A for d<=20 and d = 23, 31, 33, 34, 37; Property A' holds for d<=6.
Jäckel provides in his book (section 8.3) initialization numbers up to dimension 32. Coefficients for d<=8 are the same as in Bradley-Fox, so Property A' holds for d<=6 but Property A holds for d<=32.
The implementation of Lemieux, Cieslak, and Luttmer includes coefficients of the free direction integers up to dimension
For more info on Sobol' sequences see also "Monte Carlo Methods in Financial Engineering," by P. Glasserman, 2004, Springer, section 5.2.3
The Joe–Kuo numbers and the Kuo numbers are due to Stephen Joe and Frances Kuo.
S. Joe and F. Y. Kuo, Constructing Sobol sequences with better two-dimensional projections, preprint Nov 22 2007
See http://web.maths.unsw.edu.au/~fkuo/sobol/ for more information.
The Joe-Kuo numbers are available under a BSD-style license available at the above link.
Note that the Kuo numbers were generated to work with a different ordering of primitive polynomials for the first 40 or so dimensions which is why we have the Alternative Primitive Polynomials.
Definition at line 110 of file sobolrsg.hpp.
typedef Sample<std::vector<Real> > sample_type |
Definition at line 112 of file sobolrsg.hpp.
enum DirectionIntegers |
Enumerator | |
---|---|
Unit | |
Jaeckel | |
SobolLevitan | |
SobolLevitanLemieux | |
JoeKuoD5 | |
JoeKuoD6 | |
JoeKuoD7 | |
Kuo | |
Kuo2 | |
Kuo3 |
Definition at line 113 of file sobolrsg.hpp.
|
explicit |
The so called generating integer is chosen to be \(\gamma(n) = n\) if useGrayCode is set to false and \(\gamma(n) = G(n)\) where \(G(n)\) is the Gray code of \(n\) otherwise. The Sobol integers are then constructed using formula 8.20 resp. 8.23, see "Monte Carlo Methods in Finance" by Peter Jäckel. The default is to use the Gray code since this allows a faster sequence generation. The Burley2020SobolRsg relies on an underlying SobolRsg not using the Gray code on the other hand due to its specific way of constructing the integer sequence.
Definition at line 78477 of file sobolrsg.cpp.
const std::vector< std::uint32_t > & skipTo | ( | std::uint32_t | n | ) | const |
skip to the n-th sample in the low-discrepancy sequence
Definition at line 78775 of file sobolrsg.cpp.
const std::vector< std::uint32_t > & nextInt32Sequence | ( | ) | const |
Definition at line 78807 of file sobolrsg.cpp.
const SobolRsg::sample_type & nextSequence | ( | ) | const |
Definition at line 134 of file sobolrsg.hpp.
const sample_type & lastSequence | ( | ) | const |
Definition at line 141 of file sobolrsg.hpp.
Size dimension | ( | ) | const |
Definition at line 142 of file sobolrsg.hpp.
|
private |
Definition at line 144 of file sobolrsg.hpp.
|
mutableprivate |
Definition at line 145 of file sobolrsg.hpp.
|
mutableprivate |
Definition at line 146 of file sobolrsg.hpp.
|
mutableprivate |
Definition at line 147 of file sobolrsg.hpp.
|
mutableprivate |
Definition at line 148 of file sobolrsg.hpp.
|
private |
Definition at line 149 of file sobolrsg.hpp.
|
private |
Definition at line 150 of file sobolrsg.hpp.