| 1 | /* -*- mode: c++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */ |
| 2 | |
| 3 | /* |
| 4 | Copyright (C) 2010 Kakhkhor Abdijalilov |
| 5 | |
| 6 | This file is part of QuantLib, a free-software/open-source library |
| 7 | for financial quantitative analysts and developers - http://quantlib.org/ |
| 8 | |
| 9 | QuantLib is free software: you can redistribute it and/or modify it |
| 10 | under the terms of the QuantLib license. You should have received a |
| 11 | copy of the license along with this program; if not, please email |
| 12 | <quantlib-dev@lists.sf.net>. The license is also available online at |
| 13 | <http://quantlib.org/license.shtml>. |
| 14 | |
| 15 | This program is distributed in the hope that it will be useful, but WITHOUT |
| 16 | ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS |
| 17 | FOR A PARTICULAR PURPOSE. See the license for more details. |
| 18 | */ |
| 19 | |
| 20 | /*! \file zigguratrng.hpp |
| 21 | \brief Ziggurat random-number generator |
| 22 | */ |
| 23 | |
| 24 | #ifndef quantlib_ziggurat_generator_hpp |
| 25 | #define quantlib_ziggurat_generator_hpp |
| 26 | |
| 27 | #include <ql/math/randomnumbers/mt19937uniformrng.hpp> |
| 28 | #include <ql/math/randomnumbers/randomsequencegenerator.hpp> |
| 29 | |
| 30 | namespace QuantLib { |
| 31 | |
| 32 | //! Ziggurat random-number generator |
| 33 | /*! This generator returns standard normal variates using the |
| 34 | Ziggurat method. The underlying RNG is mt19937 (32 bit |
| 35 | version). The algorithm is described in Marsaglia and Tsang |
| 36 | (2000). "The Ziggurat Method for Generating Random |
| 37 | Variables". Journal of Statistical Software 5 (8). Note that |
| 38 | step 2 from the above paper reuses the rightmost 8 bits of the |
| 39 | random integer, which creates correlation between steps 1 and |
| 40 | 2. This implementation was written from scratch, following |
| 41 | Marsaglia and Tsang. It avoids the correlation by using only |
| 42 | the leftmost 24 bits of mt19937's output. |
| 43 | |
| 44 | Note that the GNU GSL implementation uses a different value |
| 45 | for the right-most step. The GSL value is somewhat different |
| 46 | from the one reported by Marsaglia and Tsang because GSL uses |
| 47 | a different tail. This implementation uses the same right-most |
| 48 | step as reported by Marsaglia and Tsang. The generator was |
| 49 | put through Marsaglia's Diehard battery of tests and didn't |
| 50 | exibit any abnormal behavior. |
| 51 | */ |
| 52 | class ZigguratRng { |
| 53 | public: |
| 54 | typedef Sample<Real> sample_type; |
| 55 | explicit ZigguratRng(unsigned long seed = 0); |
| 56 | sample_type next() const { return {nextGaussian(), 1.0}; } |
| 57 | |
| 58 | private: |
| 59 | mutable MersenneTwisterUniformRng mt32_; |
| 60 | Real nextGaussian() const; |
| 61 | }; |
| 62 | |
| 63 | // RNG traits for Ziggurat generator |
| 64 | struct Ziggurat { |
| 65 | // typedefs |
| 66 | typedef ZigguratRng rng_type; |
| 67 | typedef RandomSequenceGenerator<rng_type> rsg_type; |
| 68 | // more traits |
| 69 | enum { allowsErrorEstimate = 1 }; |
| 70 | // factory |
| 71 | static rsg_type make_sequence_generator(Size dimension, |
| 72 | BigNatural seed) { |
| 73 | return rsg_type(dimension, seed); |
| 74 | } |
| 75 | }; |
| 76 | |
| 77 | } |
| 78 | |
| 79 | #endif |
| 80 | |