mirror of
https://github.com/boostorg/ublas.git
synced 2026-01-24 06:22:13 +00:00
231ba5f730
Auxiliary functions for extents and strides were using different functions. Additionally, many tags were used to distinguish between different tensor types. This patch simplifies interfaces of different core functions and unifies functions that can process different types of extent and stride types.
263 lines
8.0 KiB
C++
263 lines
8.0 KiB
C++
// Copyright (c) 2018-2019 Cem Bassoy
|
|
//
|
|
// 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)
|
|
//
|
|
// The authors gratefully acknowledge the support of
|
|
// Fraunhofer and Google in producing this work
|
|
// which started as a Google Summer of Code project.
|
|
//
|
|
|
|
|
|
|
|
#include <boost/numeric/ublas/tensor.hpp>
|
|
|
|
#include <boost/test/unit_test.hpp>
|
|
#include <boost/multiprecision/cpp_bin_float.hpp>
|
|
#include "utility.hpp"
|
|
|
|
BOOST_AUTO_TEST_SUITE(test_tensor_arithmetic_operations/*, * boost::unit_test::depends_on("test_tensor")*/)
|
|
|
|
using double_extended = boost::multiprecision::cpp_bin_float_double_extended;
|
|
|
|
using test_types = zip<int,float,double_extended>::with_t<boost::numeric::ublas::layout::first_order, boost::numeric::ublas::layout::last_order>;
|
|
|
|
struct fixture
|
|
{
|
|
using extents_type = boost::numeric::ublas::extents<>;
|
|
|
|
std::vector<extents_type> extents =
|
|
{
|
|
// extents_type{}, // 0
|
|
extents_type{1,1}, // 1
|
|
extents_type{1,2}, // 2
|
|
extents_type{2,1}, // 3
|
|
extents_type{2,3}, // 4
|
|
extents_type{2,3,1}, // 5
|
|
extents_type{4,1,3}, // 6
|
|
extents_type{1,2,3}, // 7
|
|
extents_type{4,2,3}, // 8
|
|
extents_type{4,2,3,5} // 9
|
|
};
|
|
};
|
|
|
|
|
|
|
|
|
|
BOOST_FIXTURE_TEST_CASE_TEMPLATE( test_tensor_binary_arithmetic_operations, value, test_types, fixture)
|
|
{
|
|
namespace ublas = boost::numeric::ublas;
|
|
using value_type = typename value::first_type;
|
|
using layout_type = typename value::second_type;
|
|
using tensor_type = ublas::tensor_dynamic<value_type,layout_type>;
|
|
|
|
|
|
auto check = [](auto const& e)
|
|
{
|
|
auto t = tensor_type (e);
|
|
auto t2 = tensor_type (e);
|
|
auto r = tensor_type (e);
|
|
auto v = value_type {};
|
|
|
|
std::iota(t.begin(), t.end(), v);
|
|
std::iota(t2.begin(), t2.end(), v+2);
|
|
|
|
r = t + t + t + t2;
|
|
|
|
for(auto i = 0ul; i < t.size(); ++i)
|
|
BOOST_CHECK_EQUAL ( r(i), 3*t(i) + t2(i) );
|
|
|
|
|
|
r = t2 / (t+3) * (t+1) - t2; // r = ( t2/ ((t+3)*(t+1)) ) - t2
|
|
|
|
for(auto i = 0ul; i < t.size(); ++i)
|
|
BOOST_CHECK_EQUAL ( r(i), t2(i) / (t(i)+3)*(t(i)+1) - t2(i) );
|
|
|
|
r = 3+t2 / (t+3) * (t+1) * t - t2; // r = 3+( t2/ ((t+3)*(t+1)*t) ) - t2
|
|
|
|
for(auto i = 0ul; i < t.size(); ++i)
|
|
BOOST_CHECK_EQUAL ( r(i), 3+t2(i) / (t(i)+3)*(t(i)+1)*t(i) - t2(i) );
|
|
|
|
r = t2 - t + t2 - t;
|
|
|
|
for(auto i = 0ul; i < r.size(); ++i)
|
|
BOOST_CHECK_EQUAL ( r(i), 4 );
|
|
|
|
|
|
r = t * t * t * t2;
|
|
|
|
for(auto i = 0ul; i < t.size(); ++i)
|
|
BOOST_CHECK_EQUAL ( r(i), t(i)*t(i)*t(i)*t2(i) );
|
|
|
|
r = (t2/t2) * (t2/t2);
|
|
|
|
for(auto i = 0ul; i < t.size(); ++i)
|
|
BOOST_CHECK_EQUAL ( r(i), 1 );
|
|
};
|
|
|
|
for(auto const& e : extents)
|
|
check(e);
|
|
|
|
|
|
BOOST_CHECK_NO_THROW ( tensor_type t = tensor_type(extents.at(0)) + tensor_type(extents.at(0)) );
|
|
BOOST_CHECK_THROW ( tensor_type t = tensor_type(extents.at(0)) + tensor_type(extents.at(2)), std::runtime_error );
|
|
BOOST_CHECK_THROW ( tensor_type t = tensor_type(extents.at(1)) + tensor_type(extents.at(2)), std::runtime_error );
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
BOOST_FIXTURE_TEST_CASE_TEMPLATE( test_tensor_unary_arithmetic_operations, value, test_types, fixture)
|
|
{
|
|
namespace ublas = boost::numeric::ublas;
|
|
using value_type = typename value::first_type;
|
|
using layout_type = typename value::second_type;
|
|
using tensor_type = ublas::tensor_dynamic<value_type,layout_type>;
|
|
|
|
|
|
auto check = [](auto const& e)
|
|
{
|
|
auto t = tensor_type (e);
|
|
auto t2 = tensor_type (e);
|
|
auto v = value_type {};
|
|
|
|
std::iota(t.begin(), t.end(), v);
|
|
std::iota(t2.begin(), t2.end(), v+2);
|
|
|
|
tensor_type r1 = t + 2 + t + 2;
|
|
|
|
for(auto i = 0ul; i < t.size(); ++i)
|
|
BOOST_CHECK_EQUAL ( r1(i), 2*t(i) + 4 );
|
|
|
|
tensor_type r2 = 2 + t + 2 + t;
|
|
|
|
for(auto i = 0ul; i < t.size(); ++i)
|
|
BOOST_CHECK_EQUAL ( r2(i), 2*t(i) + 4 );
|
|
|
|
tensor_type r3 = (t-2) + (t-2);
|
|
|
|
for(auto i = 0ul; i < t.size(); ++i)
|
|
BOOST_CHECK_EQUAL ( r3(i), 2*t(i) - 4 );
|
|
|
|
tensor_type r4 = (t*2) * (3*t);
|
|
|
|
for(auto i = 0ul; i < t.size(); ++i)
|
|
BOOST_CHECK_EQUAL ( r4(i), 2*3*t(i)*t(i) );
|
|
|
|
tensor_type r5 = (t2*2) / (2*t2) * t2;
|
|
|
|
for(auto i = 0ul; i < t.size(); ++i)
|
|
BOOST_CHECK_EQUAL ( r5(i), (t2(i)*2) / (2*t2(i)) * t2(i) );
|
|
|
|
tensor_type r6 = (t2/2+1) / (2/t2+1) / t2;
|
|
|
|
for(auto i = 0ul; i < t.size(); ++i)
|
|
BOOST_CHECK_EQUAL ( r6(i), (t2(i)/2+1) / (2/t2(i)+1) / t2(i) );
|
|
|
|
};
|
|
|
|
for(auto const& e : extents)
|
|
check(e);
|
|
|
|
BOOST_CHECK_NO_THROW ( tensor_type t = tensor_type(extents.at(0)) + 2 + tensor_type(extents.at(0)) );
|
|
BOOST_CHECK_THROW ( tensor_type t = tensor_type(extents.at(0)) + 2 + tensor_type(extents.at(2)), std::runtime_error );
|
|
BOOST_CHECK_THROW ( tensor_type t = tensor_type(extents.at(1)) + 2 + tensor_type(extents.at(2)), std::runtime_error );
|
|
BOOST_CHECK_THROW ( tensor_type t = tensor_type(extents.at(2)) + 2 + tensor_type(extents.at(2)) + tensor_type(extents.at(1)), std::runtime_error );
|
|
BOOST_CHECK_THROW ( tensor_type t = tensor_type(extents.at(2)) + 2 + tensor_type(extents.at(2)) + 2 + tensor_type(extents.at(1)), std::runtime_error );
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
BOOST_FIXTURE_TEST_CASE_TEMPLATE( test_tensor_assign_arithmetic_operations, value, test_types, fixture)
|
|
{
|
|
namespace ublas = boost::numeric::ublas;
|
|
using value_type = typename value::first_type;
|
|
using layout_type = typename value::second_type;
|
|
using tensor_type = ublas::tensor_dynamic<value_type,layout_type>;
|
|
|
|
|
|
auto check = [](auto const& e)
|
|
{
|
|
auto t = tensor_type (e);
|
|
auto t2 = tensor_type (e);
|
|
auto r = tensor_type (e);
|
|
auto v = value_type {};
|
|
|
|
std::iota(t.begin(), t.end(), v);
|
|
std::iota(t2.begin(), t2.end(), v+2);
|
|
|
|
r = t + 2;
|
|
r += t;
|
|
r += 2;
|
|
|
|
for(auto i = 0ul; i < t.size(); ++i)
|
|
BOOST_CHECK_EQUAL ( r(i), 2*t(i) + 4 );
|
|
|
|
r = 2 + t;
|
|
r += t;
|
|
r += 2;
|
|
(void)r;
|
|
for(auto i = 0ul; i < t.size(); ++i)
|
|
BOOST_CHECK_EQUAL ( r(i), 2*t(i) + 4 );
|
|
|
|
for(auto i = 0ul; i < t.size(); ++i)
|
|
BOOST_CHECK_EQUAL ( r(i), 2*t(i) + 4 );
|
|
|
|
r = (t-2);
|
|
r += t;
|
|
r -= 2;
|
|
|
|
for(auto i = 0ul; i < t.size(); ++i)
|
|
BOOST_CHECK_EQUAL ( r(i), 2*t(i) - 4 );
|
|
|
|
r = (t*2);
|
|
r *= 3;
|
|
r *= t;
|
|
|
|
for(auto i = 0ul; i < t.size(); ++i)
|
|
BOOST_CHECK_EQUAL ( r(i), 2*3*t(i)*t(i) );
|
|
|
|
r = (t2*2);
|
|
r /= 2;
|
|
r /= t2;
|
|
r *= t2;
|
|
|
|
for(auto i = 0ul; i < t.size(); ++i)
|
|
BOOST_CHECK_EQUAL ( r(i), (t2(i)*2) / (2*t2(i)) * t2(i) );
|
|
|
|
r = (t2/2+1);
|
|
r /= (2/t2+1);
|
|
r /= t2;
|
|
|
|
for(auto i = 0ul; i < t.size(); ++i)
|
|
BOOST_CHECK_EQUAL ( r(i), (t2(i)/2+1) / (2/t2(i)+1) / t2(i) );
|
|
|
|
tensor_type q = -r;
|
|
for(auto i = 0ul; i < t.size(); ++i)
|
|
BOOST_CHECK_EQUAL ( q(i), -r(i) );
|
|
|
|
tensor_type p = +r;
|
|
for(auto i = 0ul; i < t.size(); ++i)
|
|
BOOST_CHECK_EQUAL ( p(i), r(i) );
|
|
|
|
};
|
|
|
|
for(auto const& e : extents)
|
|
check(e);
|
|
|
|
auto r = tensor_type (extents.at(0));
|
|
|
|
BOOST_CHECK_NO_THROW ( r += tensor_type(extents.at(0)) + 2 + tensor_type(extents.at(0)) );
|
|
BOOST_CHECK_THROW ( r += tensor_type(extents.at(0)) + 2 + tensor_type(extents.at(2)), std::runtime_error );
|
|
BOOST_CHECK_THROW ( r += tensor_type(extents.at(1)) + 2 + tensor_type(extents.at(2)), std::runtime_error );
|
|
BOOST_CHECK_THROW ( r += tensor_type(extents.at(2)) + 2 + tensor_type(extents.at(2)) + tensor_type(extents.at(1)), std::runtime_error );
|
|
BOOST_CHECK_THROW ( r += tensor_type(extents.at(2)) + 2 + tensor_type(extents.at(2)) + 2 + tensor_type(extents.at(1)), std::runtime_error );
|
|
}
|
|
|
|
|
|
BOOST_AUTO_TEST_SUITE_END()
|