mirror of
				https://github.com/saitohirga/WSJT-X.git
				synced 2025-10-27 11:00:32 -04:00 
			
		
		
		
	
		
			
				
	
	
		
			214 lines
		
	
	
		
			8.6 KiB
		
	
	
	
		
			C++
		
	
	
	
	
	
			
		
		
	
	
			214 lines
		
	
	
		
			8.6 KiB
		
	
	
	
		
			C++
		
	
	
	
	
	
| // Copyright Paul A. Bristow 2014, 2015.
 | |
| 
 | |
| // 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)
 | |
| 
 | |
| // Note that this file contains Quickbook mark-up as well as code
 | |
| // and comments, don't change any of the special comment mark-ups!
 | |
| 
 | |
| // Example of finding nth root using 1st and 2nd derivatives of x^n.
 | |
| 
 | |
| #include <boost/math/tools/roots.hpp>
 | |
| //using boost::math::policies::policy;
 | |
| //using boost::math::tools::newton_raphson_iterate;
 | |
| //using boost::math::tools::halley_iterate;
 | |
| //using boost::math::tools::eps_tolerance; // Binary functor for specified number of bits.
 | |
| //using boost::math::tools::bracket_and_solve_root;
 | |
| //using boost::math::tools::toms748_solve;
 | |
| 
 | |
| #include <boost/math/special_functions/next.hpp>
 | |
| #include <boost/multiprecision/cpp_dec_float.hpp>
 | |
| #include <boost/math/special_functions/pow.hpp>
 | |
| #include <boost/math/constants/constants.hpp>
 | |
| 
 | |
| #include <boost/multiprecision/cpp_dec_float.hpp> // For cpp_dec_float_50.
 | |
| #include <boost/multiprecision/cpp_bin_float.hpp> // using boost::multiprecision::cpp_bin_float_50;
 | |
| #ifndef _MSC_VER  // float128 is not yet supported by Microsoft compiler at 2013.
 | |
| # include <boost/multiprecision/float128.hpp>
 | |
| #endif
 | |
| 
 | |
| #include <iostream>
 | |
| // using std::cout; using std::endl;
 | |
| #include <iomanip>
 | |
| // using std::setw; using std::setprecision;
 | |
| #include <limits>
 | |
| using std::numeric_limits;
 | |
| #include <tuple>
 | |
| #include <utility> // pair, make_pair
 | |
| 
 | |
| 
 | |
| //[root_finding_nth_functor_2deriv
 | |
| template <int N, class T = double>
 | |
| struct nth_functor_2deriv
 | |
| { // Functor returning both 1st and 2nd derivatives.
 | |
|   BOOST_STATIC_ASSERT_MSG(boost::is_integral<T>::value == false, "Only floating-point type types can be used!");
 | |
|   BOOST_STATIC_ASSERT_MSG((N > 0) == true, "root N must be > 0!");
 | |
| 
 | |
|   nth_functor_2deriv(T const& to_find_root_of) : a(to_find_root_of)
 | |
|   { /* Constructor stores value a to find root of, for example: */ }
 | |
| 
 | |
|   // using boost::math::tuple; // to return three values.
 | |
|   std::tuple<T, T, T> operator()(T const& x)
 | |
|   { 
 | |
|     // Return f(x), f'(x) and f''(x).
 | |
|     using boost::math::pow;
 | |
|     T fx = pow<N>(x) - a;                  // Difference (estimate x^n - a).
 | |
|     T dx = N * pow<N - 1>(x);              // 1st derivative f'(x).
 | |
|     T d2x = N * (N - 1) * pow<N - 2 >(x);  // 2nd derivative f''(x).
 | |
| 
 | |
|     return std::make_tuple(fx, dx, d2x);   // 'return' fx, dx and d2x.
 | |
|   }
 | |
| private:
 | |
|   T a;                                     // to be 'nth_rooted'.
 | |
| };
 | |
| 
 | |
| //] [/root_finding_nth_functor_2deriv]
 | |
| 
 | |
| /*
 | |
| To show the progress, one might use this before the return statement above?
 | |
| #ifdef BOOST_MATH_ROOT_DIAGNOSTIC
 | |
| std::cout << " x = " << x << ", fx = " << fx << ", dx = " << dx << ", dx2 = " << d2x << std::endl;
 | |
| #endif
 | |
| */
 | |
| 
 | |
| // If T is a floating-point type, might be quicker to compute the guess using a built-in type,
 | |
| // probably quickest using double, but perhaps with float or long double, T.
 | |
| 
 | |
| // If T is a type for which frexp and ldexp are not defined,
 | |
| // then it is necessary to compute the guess using a built-in type,
 | |
| // probably quickest (but limited range) using double,
 | |
| // but perhaps with float or long double, or a multiprecision T for the full range of T.
 | |
| // typedef double guess_type; is used to specify the this.
 | |
| 
 | |
| //[root_finding_nth_function_2deriv
 | |
| 
 | |
| template <int N, class T = double>
 | |
| T nth_2deriv(T x)
 | |
| { // return nth root of x using 1st and 2nd derivatives and Halley.
 | |
| 
 | |
|   using namespace std;  // Help ADL of std functions.
 | |
|   using namespace boost::math::tools; // For halley_iterate.
 | |
| 
 | |
|   BOOST_STATIC_ASSERT_MSG(boost::is_integral<T>::value == false, "Only floating-point type types can be used!");
 | |
|   BOOST_STATIC_ASSERT_MSG((N > 0) == true, "root N must be > 0!");
 | |
|   BOOST_STATIC_ASSERT_MSG((N > 1000) == false, "root N is too big!");
 | |
| 
 | |
|   typedef double guess_type; // double may restrict (exponent) range for a multiprecision T?
 | |
| 
 | |
|   int exponent;
 | |
|   frexp(static_cast<guess_type>(x), &exponent);                 // Get exponent of z (ignore mantissa).
 | |
|   T guess = ldexp(static_cast<guess_type>(1.), exponent / N);   // Rough guess is to divide the exponent by n.
 | |
|   T min = ldexp(static_cast<guess_type>(1.) / 2, exponent / N); // Minimum possible value is half our guess.
 | |
|   T max = ldexp(static_cast<guess_type>(2.), exponent / N);     // Maximum possible value is twice our guess.
 | |
| 
 | |
|   int digits = std::numeric_limits<T>::digits * 0.4;            // Accuracy triples with each step, so stop when
 | |
|                                                                 // slightly more than one third of the digits are correct.
 | |
|   const boost::uintmax_t maxit = 20;
 | |
|   boost::uintmax_t it = maxit;
 | |
|   T result = halley_iterate(nth_functor_2deriv<N, T>(x), guess, min, max, digits, it);
 | |
|   return result;
 | |
| }
 | |
| 
 | |
| //] [/root_finding_nth_function_2deriv]
 | |
| 
 | |
| 
 | |
| template <int N, typename T = double>
 | |
| T show_nth_root(T value)
 | |
| { // Demonstrate by printing the nth root using all possibly significant digits.
 | |
|   //std::cout.precision(std::numeric_limits<T>::max_digits10);
 | |
|   // or use   cout.precision(max_digits10 = 2 + std::numeric_limits<double>::digits * 3010/10000);
 | |
|   // Or guaranteed significant digits:
 | |
|    std::cout.precision(std::numeric_limits<T>::digits10);
 | |
| 
 | |
|   T r = nth_2deriv<N>(value);
 | |
|   std::cout << "Type " << typeid(T).name() << " value = " << value << ", " << N << "th root = " << r << std::endl;
 | |
|   return r;
 | |
| } // print_nth_root
 | |
| 
 | |
| 
 | |
| int main()
 | |
| {
 | |
|   std::cout << "nth Root finding Example." << std::endl;
 | |
|   using boost::multiprecision::cpp_dec_float_50; // decimal.
 | |
|   using boost::multiprecision::cpp_bin_float_50; // binary.
 | |
| #ifndef _MSC_VER  // Not supported by Microsoft compiler.
 | |
|   using boost::multiprecision::float128; // Requires libquadmath
 | |
| #endif
 | |
|   try
 | |
|   { // Always use try'n'catch blocks with Boost.Math to get any error messages.
 | |
| 
 | |
| //[root_finding_n_example_1
 | |
|     double r1 = nth_2deriv<5, double>(2); // Integral value converted to double.
 | |
| 
 | |
|     // double r2 = nth_2deriv<5>(2); // Only floating-point type types can be used!
 | |
| 
 | |
| //] [/root_finding_n_example_1
 | |
| 
 | |
|     //show_nth_root<5, float>(2); // Integral value converted to float.
 | |
|     //show_nth_root<5, float>(2.F); // 'initializing' : conversion from 'double' to 'float', possible loss of data
 | |
| 
 | |
| //[root_finding_n_example_2
 | |
| 
 | |
| 
 | |
|     show_nth_root<5, double>(2.);
 | |
|     show_nth_root<5, long double>(2.);
 | |
| #ifndef _MSC_VER  // float128 is not supported by Microsoft compiler 2013.
 | |
|     show_nth_root<5, float128>(2);
 | |
| #endif
 | |
|     show_nth_root<5, cpp_dec_float_50>(2); // dec
 | |
|     show_nth_root<5, cpp_bin_float_50>(2); // bin
 | |
| //] [/root_finding_n_example_2
 | |
| 
 | |
|     // show_nth_root<1000000>(2.); // Type double value = 2, 555th root = 1.00124969405651
 | |
|     // Type double value = 2, 1000th root = 1.00069338746258
 | |
|     // Type double value = 2, 1000000th root = 1.00000069314783
 | |
|   }
 | |
|   catch (const std::exception& e)
 | |
|   { // Always useful to include try & catch blocks because default policies
 | |
|     // are to throw exceptions on arguments that cause errors like underflow, overflow.
 | |
|     // Lacking try & catch blocks, the program will abort without a message below,
 | |
|     // which may give some helpful clues as to the cause of the exception.
 | |
|     std::cout <<
 | |
|       "\n""Message from thrown exception was:\n   " << e.what() << std::endl;
 | |
|   }
 | |
|   return 0;
 | |
| } // int main()
 | |
| 
 | |
| 
 | |
| /*
 | |
| //[root_finding_example_output_1
 | |
|  Using MSVC 2013
 | |
| 
 | |
| nth Root finding Example.
 | |
| Type double value = 2, 5th root = 1.14869835499704
 | |
| Type long double value = 2, 5th root = 1.14869835499704
 | |
| Type class boost::multiprecision::number<class boost::multiprecision::backends::cpp_dec_float<50,int,void>,1> value = 2,
 | |
|   5th root = 1.1486983549970350067986269467779275894438508890978
 | |
| Type class boost::multiprecision::number<class boost::multiprecision::backends::cpp_bin_float<50,10,void,int,0,0>,0> value = 2,
 | |
|   5th root = 1.1486983549970350067986269467779275894438508890978
 | |
| 
 | |
| //] [/root_finding_example_output_1]
 | |
| 
 | |
| //[root_finding_example_output_2
 | |
| 
 | |
|  Using GCC 4.91  (includes float_128 type)
 | |
| 
 | |
|  nth Root finding Example.
 | |
| Type d value = 2, 5th root = 1.14869835499704
 | |
| Type e value = 2, 5th root = 1.14869835499703501
 | |
| Type N5boost14multiprecision6numberINS0_8backends16float128_backendELNS0_26expression_template_optionE0EEE value = 2, 5th root = 1.148698354997035006798626946777928
 | |
| Type N5boost14multiprecision6numberINS0_8backends13cpp_dec_floatILj50EivEELNS0_26expression_template_optionE1EEE value = 2, 5th root = 1.1486983549970350067986269467779275894438508890978
 | |
| Type N5boost14multiprecision6numberINS0_8backends13cpp_bin_floatILj50ELNS2_15digit_base_typeE10EviLi0ELi0EEELNS0_26expression_template_optionE0EEE value = 2, 5th root = 1.1486983549970350067986269467779275894438508890978
 | |
| 
 | |
| RUN SUCCESSFUL (total time: 63ms)
 | |
| 
 | |
| //] [/root_finding_example_output_2]
 | |
| */
 | |
| 
 | |
| /*
 | |
| Throw out of range using GCC release mode :-(
 | |
| 
 | |
|  */
 |