adding harmonic series convergence application

applications/accuracy/mathematics/harmonic_series.cpp

@@ -0,0 +1,162 @@
+// harmonic_series.cpp: experiments with mixed-precision representations of the Harmonic Series
+//
+// Copyright (C) 2017 Stillwater Supercomputing, Inc.
+// SPDX-License-Identifier: MIT
+//
+// This file is part of the UNIVERSAL project, which is released under an MIT Open Source license.
+#include <universal/utility/directives.hpp>
+#include <universal/traits/arithmetic_traits.hpp>
+#include <universal/number/cfloat/cfloat.hpp>
+#include <universal/number/posit/posit.hpp>
+#include <universal/number/dd/dd.hpp>
+#include <numbers>   // since C++20
+
+namespace sw {
+	namespace universal {
+
+		template<typename Scalar>
+		Scalar ForwardHarmonicSeries(unsigned long long terms) {
+			Scalar one{ 1 }, sum{ 0 };
+			for (unsigned long long i = 1; i <= terms; ++i) {
+				sum += one / Scalar(i);
+			}
+			return sum;
+		}
+
+		template<typename Scalar>
+		Scalar ReverseHarmonicSeries(unsigned long long terms) {
+			Scalar one{ 1 }, sum{ 0 };
+			for (unsigned long long i = terms; i > 0; --i) {
+				sum += one / Scalar(i);
+			}
+			return sum;
+		}
+
+
+		template<typename Scalar>
+		Scalar CompensatedHarmonicSeries(unsigned long long terms) {
+			Scalar one{ 1 }, sum{ 0 };
+			for (unsigned long long i = terms; i > 0; --i) {
+				sum += one / Scalar(i);
+			}
+			return sum;
+		}
+
+		template<typename Scalar>
+		void HarmonicSeriesConvergence(unsigned orderOfMagnitude = 6) {
+			using std::abs;
+
+			std::cout << "Harmonic Series Convergence for " << type_tag<Scalar>() << '\n';
+
+			std::vector<uint64_t> term;
+			uint64_t scale = 100;
+			term.push_back(scale);
+			for (unsigned i = 3; i < orderOfMagnitude; ++i) {
+				scale *= 10;
+				term.push_back(scale);
+			}
+
+			constexpr int max_digits10 = std::numeric_limits<Scalar>::max_digits10;
+			constexpr unsigned WIDTH = max_digits10 + 8;
+			auto oldPrec = std::cout.precision();
+			std::cout << std::setprecision(max_digits10) << std::scientific;
+
+			std::cout
+				<< std::setw(WIDTH) << "terms"
+				<< std::setw(WIDTH) << "forward"
+				<< std::setw(WIDTH) << "reverse"
+				<< std::setw(WIDTH) << "difference"
+//				<< std::setw(WIDTH) << "compensated"
+//				<< std::setw(WIDTH) << "diff"
+				<< '\n';
+			for (uint64_t terms : term) {
+				Scalar forwardSum = ForwardHarmonicSeries<Scalar>(terms);
+				Scalar reverseSum = ReverseHarmonicSeries<Scalar>(terms);
+//				Scalar compensatedSum = CompensatedHarmonicSeries<Scalar>(terms);
+				Scalar diff_fr = abs(forwardSum - reverseSum);
+//				Scalar diff_rc = abs(reverseSum - compensatedSum);
+
+				std::cout 
+					<< std::setw(WIDTH) << terms
+					<< std::setw(WIDTH) << forwardSum
+					<< std::setw(WIDTH) << reverseSum
+					<< std::setw(WIDTH) << diff_fr
+//					<< std::setw(WIDTH) << compensatedSum
+//					<< std::setw(WIDTH) << diff_rc
+					<< '\n';
+			}
+
+			std::cout << std::setprecision(oldPrec) << std::defaultfloat;
+		}
+
+	}
+}
+
+int main()
+try {
+	using namespace sw::universal;
+
+	HarmonicSeriesConvergence<float>();
+	HarmonicSeriesConvergence<double>();
+	HarmonicSeriesConvergence<dd>();
+
+	return EXIT_SUCCESS;
+}
+catch (char const* msg) {
+	std::cerr << "Caught ad-hoc exception: " << msg << std::endl;
+	return EXIT_FAILURE;
+}
+catch (const sw::universal::universal_arithmetic_exception& err) {
+	std::cerr << "Caught unexpected universal arithmetic exception: " << err.what() << std::endl;
+	return EXIT_FAILURE;
+}
+catch (const sw::universal::universal_internal_exception& err) {
+	std::cerr << "Caught unexpected universal internal exception: " << err.what() << std::endl;
+	return EXIT_FAILURE;
+}
+catch (std::runtime_error& err) {
+	std::cerr << "Caught unexpected runtime error: " << err.what() << std::endl;
+	return EXIT_FAILURE;
+}
+catch (...) {
+	std::cerr << "Caught unknown exception" << std::endl;
+	return EXIT_FAILURE;
+}
+
+
+/*
+Harmonic Series Convergence for float
+			terms          forward          reverse       difference
+			  100  5.187377930e+00  5.187376976e+00  9.536743164e-07
+			 1000  7.485478401e+00  7.485471725e+00  6.675720215e-06
+			10000  9.787612915e+00  9.787604332e+00  8.583068848e-06
+		   100000  1.209085083e+01  1.209015274e+01  6.980895996e-04
+		  1000000  1.435735798e+01  1.439265156e+01  3.529357910e-02
+		 10000000  1.540368271e+01  1.668603134e+01  1.282348633e+00
+		100000000  1.540368271e+01  1.880791855e+01  3.404235840e+00
+	   1000000000  1.540368271e+01  1.880791855e+01  3.404235840e+00
+	  10000000000  1.540368271e+01  1.880791855e+01  3.404235840e+00
+	 100000000000  1.540368271e+01  1.880791855e+01  3.404235840e+00
+Harmonic Series Convergence for double
+			terms                  forward                  reverse               difference
+			  100  5.18737751763962063e+00  5.18737751763962152e+00  8.88178419700125232e-16
+			 1000  7.48547086055034328e+00  7.48547086055034061e+00  2.66453525910037570e-15
+			10000  9.78760603604434820e+00  9.78760603604438550e+00  3.73034936274052598e-14
+		   100000  1.20901461298633350e+01  1.20901461298634079e+01  7.28306304154102691e-14
+		  1000000  1.43927267228649889e+01  1.43927267228657723e+01  7.83373366175510455e-13
+		 10000000  1.66953113658572718e+01  1.66953113658599648e+01  2.69295696853077970e-12
+		100000000  1.89978964138525548e+01  1.89978964138534465e+01  8.91731133378925733e-13
+	   1000000000  2.13004815023485499e+01  2.13004815023461482e+01  2.40163444686913863e-12
+	  10000000000  2.36030665949975003e+01  2.36030665948882685e+01  1.09231734768400202e-10
+	 100000000000  2.59056516865364301e+01  2.59056516878463441e+01  1.30991395508317510e-09
+Harmonic Series Convergence for double-double
+			terms                                forward                                reverse                             difference
+			 100  5.1873775176396202608051176756582e+00  5.1873775176396202608051176756583e+00  9.8607613152626475676466070660348e-32
+			1000  7.4854708605503449126565182043340e+00  7.4854708605503449126565182043338e+00  1.7256332301709633243381562365561e-31
+		   10000  9.7876060360443822641784779048557e+00  9.7876060360443822641784779048520e+00  3.7470892997998060757057106850932e-30
+		  100000  1.2090146129863427947363219363515e+01  1.2090146129863427947363219363505e+01  9.7621537021100210919701409953745e-30
+		 1000000  1.4392726722865723631381127493198e+01  1.4392726722865723631381127493196e+01  1.9721522630525295135293214132070e-30
+	    10000000  1.6695311365859851815399118939716e+01  1.6695311365859851815399118939541e+01  1.7562015902482775317978607184608e-28
+	   100000000  1.8997896413853898324417110394294e+01  1.8997896413853898324417110394212e+01  8.1745711303527348335790372577429e-29
+	  1000000000  2.1300481502347944016685101850059e+01  2.1300481502347944016685101848909e+01  1.1499619845859299593389473160410e-27
+ */

include/universal/number/dd/manipulators.hpp

@@ -14,7 +14,9 @@
 namespace sw { namespace universal {
 
 	// Generate a type tag for a doubledouble
-	inline std::string type_tag(const dd& = {}) {
+	template<typename DoubleDoubleType,
+		std::enable_if_t< is_dd<DoubleDoubleType>, bool> = true >
+	std::string type_tag(DoubleDoubleType = {}) {
 		return std::string("double-double");
 	}