[prev in list] [next in list] [prev in thread] [next in thread] 

List:       boost-commit
Subject:    [Boost-commit] svn:boost r39070 -
From:       pbristow () hetp ! u-net ! com
Date:       2007-08-30 14:04:20
Message-ID: 20070830140420.667852F81FB () wowbagger ! osl ! iu ! edu
[Download RAW message or body]

Author: pbristow
Date: 2007-08-30 10:04:19 EDT (Thu, 30 Aug 2007)
New Revision: 39070
URL: http://svn.boost.org/trac/boost/changeset/39070

Log:
tests now pass.
Text files modified: 
   sandbox/math_toolkit/libs/math/test/test_find_location.cpp |   145 \
+++++++++++++++++++++++++--------------   1 files changed, 94 insertions(+), 51 \
deletions(-)

Modified: sandbox/math_toolkit/libs/math/test/test_find_location.cpp
==============================================================================
--- sandbox/math_toolkit/libs/math/test/test_find_location.cpp	(original)
+++ sandbox/math_toolkit/libs/math/test/test_find_location.cpp	2007-08-30 10:04:19 \
EDT (Thu, 30 Aug 2007) @@ -1,46 +1,42 @@
 // test_find_location.cpp
 
-// Copyright John Maddock 2006.
-// Copyright  Paul A. Bristow 2007.
+// Copyright John Maddock 2007.
+// Copyright Paul A. Bristow 2007.
 
 // 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)
 
-// Basic sanity test for find_location Function.
+// Basic sanity test for find_location function.
 
 // Default domain error policy is
 // #define BOOST_MATH_DOMAIN_ERROR_POLICY throw_on_error
 
 #include <boost/math/distributions/normal.hpp> // for normal_distribution
-  using boost::math::normal; 
-  using boost::math::normal_distribution; 
+  using boost::math::normal; // Default type double.
+  using boost::math::normal_distribution; // All floating-point types.
 #include <boost/math/distributions/cauchy.hpp> // for cauchy_distribution
   using boost::math::cauchy;
 #include <boost/math/distributions/pareto.hpp> // for cauchy_distribution
   using boost::math::pareto;
 #include <boost/math/distributions/find_location.hpp>
   using boost::math::find_location;
-
+  using boost::math::complement;// will be needed by users who want complement,
+#include <boost/math/policies/policy.hpp>
   using boost::math::policies::policy;
 
 #include <boost/test/included/test_exec_monitor.hpp> // for test_main
 #include <boost/test/floating_point_comparison.hpp> // for \
BOOST_CHECK_CLOSE_FRACTION, BOOST_CHECK_EQUAL... +#include \
<boost/math/concepts/real_concept.hpp> // for real_concept  
 #include <iostream>
-using std::cout;
-using std::endl;
-using std::fixed;
-using std::right;
-using std::left;
-using std::showpoint;
-using std::showpos;
-using std::setw;
-using std::setprecision;
+  using std::cout; using std::endl; using std::fixed;
+  using std::right; using std::left; using std::showpoint;
+  using std::showpos; using std::setw; using std::setprecision;
 
 #include <limits>
-using std::numeric_limits;
+  using std::numeric_limits;
 
 template <class RealType> // Any floating-point type RealType.
 void test_spots(RealType)
@@ -54,57 +50,95 @@
   RealType tolerance = (std::max)
       (boost::math::tools::epsilon<RealType>(),
       static_cast<RealType>(std::numeric_limits<double>::epsilon()));
-   tolerance *= 100;
+   tolerance *= 100; // 100 eps as a fraction.
 
   cout << "Tolerance for type " << typeid(RealType).name()  << " is "
     << setprecision(3) << tolerance  << " (or " << tolerance * 100 << "%)." << endl;
 
-  BOOST_CHECK_THROW( //  probability outside 0 to 1.
-       find_location<normal_distribution<RealType> >(static_cast<RealType>(0.), \
static_cast<RealType>(-1.), static_cast<RealType>(0.) ), std::domain_error); +  \
BOOST_CHECK_THROW( // Probability outside 0 to 1. +       \
find_location<normal_distribution<RealType> >( +       static_cast<RealType>(0.), \
static_cast<RealType>(-1.), static_cast<RealType>(0.) ), +       std::domain_error);
   
-} // template <class RealType>void test_spots(RealType)
+  normal_distribution<RealType> n; // standard N(0,1)
+  BOOST_CHECK_EQUAL(n.location(), 0); // aka mean.
+  BOOST_CHECK_EQUAL(n.scale(), 1); // aka standard_deviation.
 
-int test_main(int, char* [])
-{
 	// Check for 'bad' arguments.
-  BOOST_CHECK_THROW(find_location<normal>(0., -1., 0.), std::domain_error); // p \
                outside 0 to 1.
-  BOOST_CHECK_THROW(find_location<normal>(0., 2., 0.), std::domain_error); // p \
                outside 0 to 1.
-  BOOST_CHECK_THROW(find_location<normal>(numeric_limits<double>::infinity(), 0.5, \
                0.), std::domain_error); // z not finite.
-  BOOST_CHECK_THROW(find_location<normal>(numeric_limits<double>::quiet_NaN(), -1., \
                0.), std::domain_error); // z not finite
-  BOOST_CHECK_THROW(find_location<normal>(0., -1., \
                numeric_limits<double>::quiet_NaN()), std::domain_error); // scale \
                not finite
-
-  //// Check for ab-use with unsuitable distribution(s).
-  //BOOST_CHECK_THROW(find_location<pareto>(0., 0.5, 0.), std::domain_error); // 
-
-  double tol5eps = std::numeric_limits<double>::epsilon() * 5; // 5 eps as a \
                fraction.
-  double tol100eps = std::numeric_limits<double>::epsilon() * 100; // 100 eps as a \
                fraction.
-
-  double mean = 3.; // kg
-  double standard_deviation = 0.1; // kg
-  normal n(mean, standard_deviation);
-  double z = 2.9;
-  double p = 0.05;
-  double l = find_location<normal>(z, p, n.scale());
-  BOOST_CHECK_CLOSE_FRACTION(l, 3.0644853626951472, tol100eps);
+  BOOST_CHECK_THROW(find_location<normal>(0., -1., 0.), std::domain_error); // p \
below 0 to 1. +  BOOST_CHECK_THROW(find_location<normal>(0., 2., 0.), \
std::domain_error); // p above 0 to 1. +  \
BOOST_CHECK_THROW(find_location<normal>(numeric_limits<double>::infinity(), 0.5, 0.), \
+    std::domain_error); // z not finite. +  \
BOOST_CHECK_THROW(find_location<normal>(numeric_limits<double>::quiet_NaN(), -1., \
0.), +    std::domain_error); // z not finite
+  BOOST_CHECK_THROW(find_location<normal>(0., -1., \
numeric_limits<double>::quiet_NaN()), +    std::domain_error); // scale not finite
 
-  //l = find_location<normal>(complement(z, 1 - p, n.scale()));
-  BOOST_CHECK_CLOSE_FRACTION(l, 3.0644853626951472, tol100eps);
+  //// Check for ab-use with unsuitable distribution(s) when concept check \
implemented. +  // BOOST_CHECK_THROW(find_location<pareto>(0., 0.5, 0.), \
std::domain_error); // pareto can't be used with find_location.  
+  // Check doesn't throw when an ignore_error for domain_error policy is used.
+  using boost::math::policies::policy;
+  using boost::math::policies::domain_error;
+  using boost::math::policies::ignore_error;
 
-  ////using namespace boost::math; or 
-  //using boost::math::normal;
+  // Define a (bad?) policy to ignore domain errors ('bad' arguments):
+  typedef policy<domain_error<ignore_error> > ignore_domain_policy;
+  // Using a typedef is convenient, especially if it is re-used.
+
+  BOOST_CHECK_NO_THROW(find_location<normal>(0, -1, 1,
+    ignore_domain_policy())); // probability outside [0, 1]
+  BOOST_CHECK_NO_THROW(find_location<normal>(numeric_limits<double>::infinity(), -1, \
1, +    ignore_domain_policy())); // z not finite.
+
+  // Find location to give a probability p (0.05) of z (-2)
+  RealType sd = static_cast<RealType>(1); // normal default standard deviation = 1.
+  RealType z = static_cast<RealType>(-2); // z to give prob p
+  RealType p = static_cast<RealType>(0.05); // only 5% will be below z
+  RealType l = find_location<normal_distribution<RealType> >(z, p, sd);
+  // cout << z << " " << p << " " << sd << " " << l << endl;
+
+  normal_distribution<RealType> np05pc(l, sd); // Same standard_deviation (scale) \
but with mean(location) shifted. +  //  cout << "Normal distribution with mean = " << \
l << " has " << "fraction <= " << z << " = "  << cdf(np05pc, z) << endl; +  // Check \
cdf such that only fraction p really is below offset mean l.  +  \
BOOST_CHECK_CLOSE_FRACTION(p, cdf(np05pc, z), tolerance); +
+  // Check that some policies can be applied (though not used here).
+  l = find_location<normal_distribution<RealType> >(z, p, sd, policy<>()); // \
Default policy, needs using boost::math::policies::policy; +  l = \
find_location<normal_distribution<RealType> >(z, p, sd, \
boost::math::policies::policy<>()); // Default policy, fully specified. +  l = \
find_location<normal_distribution<RealType> >(z, p, sd, ignore_domain_policy()); // \
find_location with new policy, using typedef. +  l = \
find_location<normal_distribution<RealType> >(z, p, sd, \
policy<domain_error<ignore_error> >()); // New policy, without typedef. +
+  // Check that can use the complement version.
+  RealType q = 1 - p; // complement.
+  // cout << "find_location<normal_distribution<RealType> >(complement(z, q, sd)) = \
" << endl; +  l = find_location<normal_distribution<RealType> >(complement(z, q, \
sd)); +
+  normal_distribution<RealType> np95pc(l, sd); // Same standard_deviation (scale) \
but with mean(location) shifted +  //  cout << "Normal distribution with mean = " << \
l << " has " << "fraction <= " << z << " = "  << cdf(np95pc, z) << endl; +  \
BOOST_CHECK_CLOSE_FRACTION(q, cdf(np95pc, z), tolerance);  
+} // template <class RealType>void test_spots(RealType)
 
+int test_main(int, char* [])
+{
   // Basic sanity-check spot values.
 
-  // (Parameter value, arbitrarily zero, only communicates the floating point type).
+  // (Parameter value, arbitrarily zero, only communicates the floating-point type).
   test_spots(0.0F); // Test float.
   test_spots(0.0); // Test double.
-  //test_spots(0.0L); // Test long double.
-//#if !BOOST_WORKAROUND(__BORLANDC__, BOOST_TESTED_AT(0x582))
-//  test_spots(boost::math::concepts::real_concept(0.)); // Test real concept.
-//#endif
-
+  test_spots(0.0L); // Test long double.
+  #ifndef BOOST_MATH_NO_LONG_DOUBLE_MATH_FUNCTIONS
+    test_spots(0.0L); // Test long double.
+  #if !BOOST_WORKAROUND(__BORLANDC__, BOOST_TESTED_AT(0x0582))
+    test_spots(boost::math::concepts::real_concept(0.)); // Test real concept.
+  #endif
+  #else
+     std::cout << "<note>The long double tests have been disabled on this platform "
+        "either because the long double overloads of the usual math functions are "
+        "not available at all, or because they are too inaccurate for these tests "
+        "to pass.</note>" << std::cout;
+  #endif
   return 0;
 } // int test_main(int, char* [])
 
@@ -112,6 +146,15 @@
 
 Output is:
 
+Autorun "i:\boost-06-05-03-1300\libs\math\test\Math_test\debug\test_find_location.exe"
 +Running 1 test case...
+Tolerance for type float is 1.19e-005 (or 0.00119%).
+Tolerance for type double is 2.22e-014 (or 2.22e-012%).
+Tolerance for type long double is 2.22e-014 (or 2.22e-012%).
+Tolerance for type long double is 2.22e-014 (or 2.22e-012%).
+Tolerance for type class boost::math::concepts::real_concept is 2.22e-014 (or \
2.22e-012%). +*** No errors detected
+
 
 */
 
_______________________________________________
Boost-commit mailing list
Boost-commit@lists.boost.org
http://lists.boost.org/mailman/listinfo.cgi/boost-commit


[prev in list] [next in list] [prev in thread] [next in thread] 

Configure | About | News | Add a list | Sponsored by KoreLogic