📄 test_roots.cpp
字号:
// (C) Copyright John Maddock 2006.// Use, modification and distribution are subject to the// Boost Software License, Version 1.0. (See accompanying file// LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)#include <boost/test/included/test_exec_monitor.hpp>#include <boost/test/floating_point_comparison.hpp>#include <boost/test/results_collector.hpp>#include <boost/math/special_functions/beta.hpp>#include <boost/math/tools/roots.hpp>#include <boost/array.hpp>#define BOOST_CHECK_CLOSE_EX(a, b, prec, i) \ {\ unsigned int failures = boost::unit_test::results_collector.results( boost::unit_test::framework::current_test_case().p_id ).p_assertions_failed;\ BOOST_CHECK_CLOSE(a, b, prec); \ if(failures != boost::unit_test::results_collector.results( boost::unit_test::framework::current_test_case().p_id ).p_assertions_failed)\ {\ std::cerr << "Failure was at row " << i << std::endl;\ std::cerr << std::setprecision(35); \ std::cerr << "{ " << data[i][0] << " , " << data[i][1] << " , " << data[i][2];\ std::cerr << " , " << data[i][3] << " , " << data[i][4] << " , " << data[i][5] << " } " << std::endl;\ }\ }//// Implement various versions of inverse of the incomplete beta// using different root finding algorithms, and deliberately "bad"// starting conditions: that way we get all the pathological cases// we could ever wish for!!!//template <class T, class Policy>struct ibeta_roots_1 // for first order algorithms{ ibeta_roots_1(T _a, T _b, T t, bool inv = false) : a(_a), b(_b), target(t), invert(inv) {} T operator()(const T& x) { return boost::math::detail::ibeta_imp(a, b, x, Policy(), invert, true) - target; }private: T a, b, target; bool invert;};template <class T, class Policy>struct ibeta_roots_2 // for second order algorithms{ ibeta_roots_2(T _a, T _b, T t, bool inv = false) : a(_a), b(_b), target(t), invert(inv) {} std::tr1::tuple<T, T> operator()(const T& x) { typedef typename boost::math::lanczos::lanczos<T, Policy>::type L; T f = boost::math::detail::ibeta_imp(a, b, x, Policy(), invert, true) - target; T f1 = invert ? -boost::math::detail::ibeta_power_terms(b, a, 1 - x, x, L(), true, Policy()) : boost::math::detail::ibeta_power_terms(a, b, x, 1 - x, L(), true, Policy()); T y = 1 - x; if(y == 0) y = boost::math::tools::min_value<T>() * 8; f1 /= y * x; // make sure we don't have a zero derivative: if(f1 == 0) f1 = (invert ? -1 : 1) * boost::math::tools::min_value<T>() * 64; return std::tr1::make_tuple(f, f1); }private: T a, b, target; bool invert;};template <class T, class Policy>struct ibeta_roots_3 // for third order algorithms{ ibeta_roots_3(T _a, T _b, T t, bool inv = false) : a(_a), b(_b), target(t), invert(inv) {} std::tr1::tuple<T, T, T> operator()(const T& x) { typedef typename boost::math::lanczos::lanczos<T, Policy>::type L; T f = boost::math::detail::ibeta_imp(a, b, x, Policy(), invert, true) - target; T f1 = invert ? -boost::math::detail::ibeta_power_terms(b, a, 1 - x, x, L(), true, Policy()) : boost::math::detail::ibeta_power_terms(a, b, x, 1 - x, L(), true, Policy()); T y = 1 - x; if(y == 0) y = boost::math::tools::min_value<T>() * 8; f1 /= y * x; T f2 = f1 * (-y * a + (b - 2) * x + 1) / (y * x); if(invert) f2 = -f2; // make sure we don't have a zero derivative: if(f1 == 0) f1 = (invert ? -1 : 1) * boost::math::tools::min_value<T>() * 64; return std::tr1::make_tuple(f, f1, f2); }private: T a, b, target; bool invert;};double inverse_ibeta_bisect(double a, double b, double z){ typedef boost::math::policies::policy<> pol; bool invert = false; int bits = std::numeric_limits<double>::digits; // // special cases, we need to have these because there may be other // possible answers: // if(z == 1) return 1; if(z == 0) return 0; // // We need a good estimate of the error in the incomplete beta function // so that we don't set the desired precision too high. Assume that 3-bits // are lost each time the arguments increase by a factor of 10: // using namespace std; int bits_lost = static_cast<int>(ceil(log10((std::max)(a, b)) * 3)); if(bits_lost < 0) bits_lost = 3; else bits_lost += 3; int precision = bits - bits_lost; double min = 0; double max = 1; boost::math::tools::eps_tolerance<double> tol(precision); return boost::math::tools::bisect(ibeta_roots_1<double, pol>(a, b, z, invert), min, max, tol).first;}double inverse_ibeta_newton(double a, double b, double z){ double guess = 0.5; bool invert = false; int bits = std::numeric_limits<double>::digits; // // special cases, we need to have these because there may be other // possible answers: // if(z == 1) return 1; if(z == 0) return 0; // // We need a good estimate of the error in the incomplete beta function // so that we don't set the desired precision too high. Assume that 3-bits // are lost each time the arguments increase by a factor of 10: // using namespace std; int bits_lost = static_cast<int>(ceil(log10((std::max)(a, b)) * 3)); if(bits_lost < 0) bits_lost = 3; else bits_lost += 3; int precision = bits - bits_lost; double min = 0; double max = 1; return boost::math::tools::newton_raphson_iterate(ibeta_roots_2<double, boost::math::policies::policy<> >(a, b, z, invert), guess, min, max, precision);}double inverse_ibeta_halley(double a, double b, double z){ double guess = 0.5; bool invert = false; int bits = std::numeric_limits<double>::digits; // // special cases, we need to have these because there may be other // possible answers: // if(z == 1) return 1; if(z == 0) return 0; // // We need a good estimate of the error in the incomplete beta function // so that we don't set the desired precision too high. Assume that 3-bits // are lost each time the arguments increase by a factor of 10: // using namespace std; int bits_lost = static_cast<int>(ceil(log10((std::max)(a, b)) * 3)); if(bits_lost < 0) bits_lost = 3; else bits_lost += 3; int precision = bits - bits_lost; double min = 0; double max = 1; return boost::math::tools::halley_iterate(ibeta_roots_3<double, boost::math::policies::policy<> >(a, b, z, invert), guess, min, max, precision);}double inverse_ibeta_schroeder(double a, double b, double z){ double guess = 0.5; bool invert = false; int bits = std::numeric_limits<double>::digits; // // special cases, we need to have these because there may be other // possible answers: // if(z == 1) return 1; if(z == 0) return 0; // // We need a good estimate of the error in the incomplete beta function // so that we don't set the desired precision too high. Assume that 3-bits // are lost each time the arguments increase by a factor of 10: // using namespace std; int bits_lost = static_cast<int>(ceil(log10((std::max)(a, b)) * 3)); if(bits_lost < 0) bits_lost = 3; else bits_lost += 3; int precision = bits - bits_lost; double min = 0; double max = 1; return boost::math::tools::schroeder_iterate(ibeta_roots_3<double, boost::math::policies::policy<> >(a, b, z, invert), guess, min, max, precision);}template <class T>void test_inverses(const T& data){ using namespace std; typedef typename T::value_type row_type; typedef typename row_type::value_type value_type; value_type precision = static_cast<value_type>(ldexp(1.0, 1-boost::math::policies::digits<value_type, boost::math::policies::policy<> >()/2)) * 100; if(boost::math::policies::digits<value_type, boost::math::policies::policy<> >() < 50) precision = 1; // 1% or two decimal digits, all we can hope for when the input is truncated for(unsigned i = 0; i < data.size(); ++i) { // // These inverse tests are thrown off if the output of the // incomplete beta is too close to 1: basically there is insuffient // information left in the value we're using as input to the inverse // to be able to get back to the original value. // if(data[i][5] == 0) { BOOST_CHECK_EQUAL(inverse_ibeta_halley(data[i][0], data[i][1], data[i][5]), value_type(0)); BOOST_CHECK_EQUAL(inverse_ibeta_schroeder(data[i][0], data[i][1], data[i][5]), value_type(0)); BOOST_CHECK_EQUAL(inverse_ibeta_newton(data[i][0], data[i][1], data[i][5]), value_type(0)); BOOST_CHECK_EQUAL(inverse_ibeta_bisect(data[i][0], data[i][1], data[i][5]), value_type(0)); } else if((1 - data[i][5] > 0.001) && (fabs(data[i][5]) > 2 * boost::math::tools::min_value<value_type>()) && (fabs(data[i][5]) > 2 * boost::math::tools::min_value<double>())) { value_type inv = inverse_ibeta_halley(data[i][0], data[i][1], data[i][5]); BOOST_CHECK_CLOSE_EX(data[i][2], inv, precision, i); inv = inverse_ibeta_schroeder(data[i][0], data[i][1], data[i][5]); BOOST_CHECK_CLOSE_EX(data[i][2], inv, precision, i); inv = inverse_ibeta_newton(data[i][0], data[i][1], data[i][5]); BOOST_CHECK_CLOSE_EX(data[i][2], inv, precision, i); inv = inverse_ibeta_bisect(data[i][0], data[i][1], data[i][5]); BOOST_CHECK_CLOSE_EX(data[i][2], inv, precision, i); } else if(1 == data[i][5]) { BOOST_CHECK_EQUAL(inverse_ibeta_halley(data[i][0], data[i][1], data[i][5]), value_type(1)); BOOST_CHECK_EQUAL(inverse_ibeta_schroeder(data[i][0], data[i][1], data[i][5]), value_type(1)); BOOST_CHECK_EQUAL(inverse_ibeta_newton(data[i][0], data[i][1], data[i][5]), value_type(1)); BOOST_CHECK_EQUAL(inverse_ibeta_bisect(data[i][0], data[i][1], data[i][5]), value_type(1)); } }}template <class T>void test_beta(T, const char* /* name */){ // // The actual test data is rather verbose, so it's in a separate file // // The contents are as follows, each row of data contains // five items, input value a, input value b, integration limits x, beta(a, b, x) and ibeta(a, b, x): //# include "ibeta_small_data.ipp" test_inverses(ibeta_small_data);# include "ibeta_data.ipp" test_inverses(ibeta_data);# include "ibeta_large_data.ipp" test_inverses(ibeta_large_data);}int test_main(int, char* []){ test_beta(0.1, "double"); return 0;}
⌨️ 快捷键说明
复制代码
Ctrl + C
搜索代码
Ctrl + F
全屏模式
F11
切换主题
Ctrl + Shift + D
显示快捷键
?
增大字号
Ctrl + =
减小字号
Ctrl + -