Subject: [Boost-commit] svn:boost r75199 - sandbox/multiprecision/libs/multiprecision/example/simple_example_1
From: pbristow_at_[hidden]
Date: 2011-10-31 13:38:37
Author: pbristow
Date: 2011-10-31 13:38:36 EDT (Mon, 31 Oct 2011)
New Revision: 75199
URL: http://svn.boost.org/trac/boost/changeset/75199
Log:
simple example used in draft docs. bat will need changes for CK's environment, removed after use - do not commit changed version!
Added:
sandbox/multiprecision/libs/multiprecision/example/simple_example_1/
sandbox/multiprecision/libs/multiprecision/example/simple_example_1/jamfile.v2 (contents, props changed)
sandbox/multiprecision/libs/multiprecision/example/simple_example_1/simple_example_1.bat (contents, props changed)
sandbox/multiprecision/libs/multiprecision/example/simple_example_1/simple_example_1.cpp (contents, props changed)
Added: sandbox/multiprecision/libs/multiprecision/example/simple_example_1/jamfile.v2
==============================================================================
--- (empty file)
+++ sandbox/multiprecision/libs/multiprecision/example/simple_example_1/jamfile.v2 2011-10-31 13:38:36 EDT (Mon, 31 Oct 2011)
@@ -0,0 +1,67 @@
+# multiprecision/libs/multiprecision/example/simple_example_1/jamfile.v2
+
+# Runs some simple examples using bjam.
+
+# Copyright Paul A. Bristow 2011
+
+# Distributed under 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)
+
+import testing ;
+
+project
+ : requirements
+ #<library>/boost/test//boost_test_exec_monitor/<link>static
+ # But this original version does not use Boost.Test.
+ # (See elsewhere for tests using Boost.Test)
+ <include>../../../..
+ <include>../../../../boost/ # MP Headers.
+ <define>BOOST_ALL_NO_LIB=1 # Why?
+ <define>BOOST_MULTIPRECISION_BACKEND_MP_FLOAT_TYPE_EFX # Could also be BOOST_MULTIPRECISION_BACKEND_MP_FLOAT_TYPE_GMP or BOOST_MULTIPRECISION_BACKEND_MP_FLOAT_TYPE_MFPR.
+ <define>BOOST_MULTIPRECISION_BACKEND_MP_FLOAT_DIGITS10=50 # Overriding the default of 100 for testing purposes.
+ # But has no effect without bjam -a option?
+
+ <toolset>gcc:<cxxflags>-Wextra
+ <toolset>gcc:<cxxflags>-Wshadow
+ <toolset>gcc:<warnings-as-errors>on
+ <toolset>gcc:<cxxflags>-Wno-long-long
+ <toolset>gcc:<cxxflags>-Wcast-align
+ <toolset>intel:<warnings-as-errors>on
+
+ -<toolset>msvc:<cxxflags>/Za # Requires MS extensions, so explictly cancel /Za to *enable extensions*.
+ <toolset>msvc:<asynch-exceptions>on # Needed for Boost.Test
+ <toolset>msvc:<warnings>all
+ <toolset>msvc:<warnings-as-errors>on
+ <toolset>msvc:<define>_SCL_SECURE_NO_WARNINGS
+ <toolset>msvc:<define>_SCL_SECURE_NO_DEPRECATE
+ <toolset>msvc:<define>_CRT_SECURE_NO_DEPRECATE
+ <toolset>msvc:<define>_CRT_SECURE_NO_WARNINGS
+ # <toolset>msvc:<define>BOOST_MULTIPRECISION_BACKEND_MP_FLOAT_DIGITS10=100 # Overriding the default of 50 (for testing MSVC only)?
+
+ # Warning suppression:
+ <toolset>msvc:<cxxflags>/wd4224 # nonstandard extension used : formal parameter 'arg' was previously defined as a type.
+ <toolset>msvc:<cxxflags>/wd4127 # expression is constant.
+ ;
+
+ # Simple example 1 from Paul Bristow.
+ test-suite "multiprecision examples" :
+ [
+ run ./simple_example_1.cpp
+
+ # Source files required .
+ ../../src/backends/float/mp_float.cpp
+ ../../src/backends/float/mp_float_base.cpp
+ ../../src/backends/float/efx/mp_float_efx.cpp
+ ../../src/functions/constants/constants.cpp
+ ../../src/functions/elementary/elementary_complex.cpp
+ ../../src/functions/elementary/elementary_math.cpp
+ ../../src/functions/elementary/elementary_hyper_g.cpp
+ ../../src/functions/elementary/elementary_trans.cpp
+ ../../src/functions/elementary/elementary_trig.cpp
+
+ ]
+ ;
+
+# Output is at
+
+/boost-sandbox/multiprecision/libs/multiprecision/example/simple_example_1/bin/simple_example_1.test/msvc-10.0/debug/asynch-exceptions-on/threading-multi/simple_example_1.output
\ No newline at end of file
Added: sandbox/multiprecision/libs/multiprecision/example/simple_example_1/simple_example_1.bat
==============================================================================
--- (empty file)
+++ sandbox/multiprecision/libs/multiprecision/example/simple_example_1/simple_example_1.bat 2011-10-31 13:38:36 EDT (Mon, 31 Oct 2011)
@@ -0,0 +1,12 @@
+echo off
+rem The DOS time format is assumed 12:34 and the : separator is not used.
+set t=%time% /T
+set tim=%t:~0,2%%t:~3,2%
+rem pick just hours and minutes.
+rem time may include leading space, like " 915", so remove space.
+set tim=%tim: =%
+i:
+cd \boost-sandbox\multiprecision\libs\multiprecision\example\simple_example_1
+b2 > simple_example_1_%date%_%tim%.log
+if not ERRORLEVEL 0 (echo Errorlevel is %ERRORLEVEL%) else (echo OK)
+pause
Added: sandbox/multiprecision/libs/multiprecision/example/simple_example_1/simple_example_1.cpp
==============================================================================
--- (empty file)
+++ sandbox/multiprecision/libs/multiprecision/example/simple_example_1/simple_example_1.cpp 2011-10-31 13:38:36 EDT (Mon, 31 Oct 2011)
@@ -0,0 +1,194 @@
+
+//! \file
+
+// Copyright Paul A. Bristow 2011.
+
+// 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)
+
+// Some really simple examples of using mp_float.
+
+#include <iostream>
+using std::cout;
+using std::cin;
+using std::endl;
+using std::dec;
+using std::hex;
+using std::boolalpha;
+using std::showpoint;
+
+#include <iomanip>
+using std::setprecision;
+using std::setw;
+#include <string>
+using std::string;
+
+#include <cassert>
+#include <limits> // for std::numeric_limits
+
+#include <boost/multiprecision/mp_float.hpp>
+
+#include <boost/multiprecision/mp_float_functions.hpp> // for constants.
+
+// Select the mp_float back-end big-number type
+// by defining BOOST_MULTIPRECISION_BACKEND_MP_FLOAT_TYPE_xxx, for example,
+// as a compiler defined preprocessor macro.
+
+// For example #define BOOST_MULTIPRECISION_BACKEND_MP_FLOAT_TYPE_EFX; in project properties.
+// Ensure that MS language extensions are NOT disabled.
+
+// Need to include .cpp files:
+// constants.cpp
+
+int main()
+{
+
+ using boost::multiprecision::mp_float;
+
+ // Display all the std::numeric_limits values.
+
+ cout << boolalpha << endl;
+ cout << "std::numeric_limits<mp_float>::max_digits10 is " << std::numeric_limits<mp_float>::max_digits10 << endl;
+ cout << "std::numeric_limits<mp_float>::digits10 is " << std::numeric_limits<mp_float>::digits10 << endl;
+ cout << "std::numeric_limits<mp_float>::digits is " << std::numeric_limits<mp_float>::digits << endl;
+ cout << "std::numeric_limits<mp_float>::radix is " << std::numeric_limits<mp_float>::radix << endl;
+ cout << "std::numeric_limits<mp_float>::epsilon() is " << std::numeric_limits<mp_float>::epsilon() << endl;
+
+ cout << "std::numeric_limits<mp_float>::has_infinity is " << std::numeric_limits<mp_float>::has_infinity << endl;
+ cout << "std::numeric_limits<mp_float>::infinity() is " << std::numeric_limits<mp_float>::infinity() << endl;
+ cout << "std::numeric_limits<mp_float>::has_quiet_NaN is " << std::numeric_limits<mp_float>::has_quiet_NaN << endl;
+ cout << "std::numeric_limits<mp_float>::quiet_NaN() is " << std::numeric_limits<mp_float>::quiet_NaN() << endl;
+
+
+ cout << "std::numeric_limits<mp_float>::round_error() is " << std::numeric_limits<mp_float>::round_error() << endl;
+ cout << "std::numeric_limits<mp_float>::max() is " << std::numeric_limits<mp_float>::max() << endl;
+ cout << "std::numeric_limits<mp_float>::min() is " << std::numeric_limits<mp_float>::min() << endl;
+ cout << "std::numeric_limits<mp_float>::max_exponent is " << std::numeric_limits<mp_float>::max_exponent << endl;
+ cout << "std::numeric_limits<mp_float>::min_exponent is " << std::numeric_limits<mp_float>::min_exponent << endl;
+
+ cout << "std::numeric_limits<mp_float>::max_exponent10 is " << std::numeric_limits<mp_float>::max_exponent10 << endl;
+ cout << "std::numeric_limits<mp_float>::min_exponent10 is " << std::numeric_limits<mp_float>::min_exponent10 << endl;
+
+ mp_float my_float;
+ cout << "mp_float default my_float; = " << my_float << endl; // Default construction is zero.
+
+ my_float = 1.23456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890;
+ // my_float = 1.2345678901234566904321354741114191710948944091796875
+ // ^
+ // BUT see the catastrophic loss of accuracy starting at the 17th decimal digit!
+ // (std::numeric_limits<double>>::max_digits10 = 17)
+
+ // Change to maximum precision to show all significant digits and then show again.
+ cout.precision(std::numeric_limits<mp_float>::max_digits10);
+ cout << "cout.precision(std::numeric_limits<mp_float>::max_digits10); is " << cout.precision() << endl;
+ cout << "my_float = " << my_float << endl; //
+
+ // Construct from string with far more digits than max_digits10.
+ mp_float d("1.23456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890");
+
+ cout << "mp_float d constructed from string " << d << endl; // 1.234567890123456789012345678901234567890123456789012345679
+ // No loss of accuracy until the 57th digit.
+
+ my_float = 2.0; // IS allowed - implicit conversion from double (by only the accuracy of double).
+ // And the accuracy will not necessarily be the same on all platforms and implementations!
+
+ my_float = 2; // IS allowed - implicit conversion from int.
+
+ mp_float my_float_double = mp_float(2.); // Construct from double.
+ mp_float my_float_int = mp_float(2); // Construct from int.
+
+ { // Show a few display features.
+ cout.precision(6); // Reset to default precision = 6.
+ cout << "my_float from long string = " << d << endl; // my_float from long string = 1.23457
+ // Default precision = 6, so doesn't show much of the potential accuracy!
+
+ // Return to showing all possibly significant digits.
+ cout.precision(std::numeric_limits<mp_float>::max_digits10);
+
+ // Force display of trailing zero decimal digits with showpoint manipulator.
+ mp_float my_float_int(mp_float(2));
+ cout << my_float_int << endl; // Just "2", no decimal point and no trailing zeros.
+ cout << "mp_float my_float_int(mp_float(2)); = " << showpoint << my_float << endl;
+ // 2.000000000000000000000000000000000000000000000000000000000
+
+ my_float *= 2;
+ cout <<"my_float *= 2; = " << showpoint << my_float << endl;
+
+ BOOST_ASSERT(my_float == 4.);
+
+ } // Display features.
+
+ { // multiprecision Constants.
+
+ using boost::multiprecision::half;
+
+ cout << "half = " << half << endl; // 01369B60 - address not value!
+ cout << "half() = " << half() << endl; //
+
+ cout << boost::multiprecision::mp_float_efx::mp_radix << endl; // 10
+ cout << boost::multiprecision::mp_float_efx::mp_elem_digits10 << endl; // 8
+ cout << boost::multiprecision::mp_float_efx::mp_float_digits10_extra << endl; // 8
+ cout << boost::multiprecision::mp_float_efx::mp_float_max_digits10 << endl; // 8
+
+ using boost::multiprecision::two;
+
+ cout << "two = " << two << endl; // NO! - it's a function call that you want - this returns the address!
+
+ cout << "boost::multiprecision::two() = " << two() << endl;
+
+ cout << "boost::multiprecision::golden_ratio() "<< boost::multiprecision::golden_ratio() << endl;
+
+ mp_float my_pi;
+
+ my_pi = boost::multiprecision::pi();
+
+ cout << "boost::multiprecision::pi() " << my_pi << endl;
+ }
+
+
+} // int main()
+
+/*
+
+Output is
+
+ simple_example_1.vcxproj -> I:\boost-sandbox\multiprecision\libs\multiprecision\build\Release\simple_example_1.exe
+
+ std::numeric_limits<mp_float>::max_digits10 is 58
+ std::numeric_limits<mp_float>::digits10 is 50
+ std::numeric_limits<mp_float>::digits is 50
+ std::numeric_limits<mp_float>::radix is 10
+ std::numeric_limits<mp_float>::epsilon() is 1e-049
+ std::numeric_limits<mp_float>::has_infinity is true
+ std::numeric_limits<mp_float>::infinity() is inf
+ std::numeric_limits<mp_float>::has_quiet_NaN is true
+ std::numeric_limits<mp_float>::quiet_NaN() is nan
+ std::numeric_limits<mp_float>::round_error() is 0.5
+ std::numeric_limits<mp_float>::max() is 1e+003063937869882635617
+ std::numeric_limits<mp_float>::min() is 1e-003063937869882635616
+ std::numeric_limits<mp_float>::max_exponent is 9223372036854775795
+ std::numeric_limits<mp_float>::min_exponent is -9223372036854775795
+ std::numeric_limits<mp_float>::max_exponent10 is 3063937869882635616
+ std::numeric_limits<mp_float>::min_exponent10 is -3063937869882635616
+ mp_float default my_float; = 0
+ cout.precision(std::numeric_limits<mp_float>::max_digits10); is 58
+ my_float = 1.2345678901234566904321354741114191710948944091796875
+ mp_float d constructed from string 1.234567890123456789012345678901234567890123456789012345679
+ my_float from long string = 1.23457
+ 2
+ mp_float my_float_int(mp_float(2)); = 2.000000000000000000000000000000000000000000000000000000000
+ my_float *= 2; = 4.000000000000000000000000000000000000000000000000000000000
+ half = 00D3A070
+ half() = 0.5000000000000000000000000000000000000000000000000000000000
+ 10
+ 8
+ 8
+ 58
+ two = 00D3BA60
+ boost::multiprecision::two() = 2.000000000000000000000000000000000000000000000000000000000
+ boost::multiprecision::golden_ratio() 1.618033988749894848204586834365638117720309179805762862135
+ boost::multiprecision::pi() 3.141592653589793238462643383279502884197169399375105820975
+
+ */
\ No newline at end of file