1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
|
#include "purify/types.h"
#include <array>
#include <random>
#include "benchmarks/utilities.h"
#include "purify/convergence_factory.h"
#include "purify/directories.h"
#include "purify/distribute.h"
#include "purify/mpi_utilities.h"
#include "purify/operators.h"
#include "purify/utilities.h"
#include <sopt/imaging_padmm.h>
#include <sopt/mpi/communicator.h>
#include <sopt/mpi/session.h>
#include <sopt/relative_variation.h>
#include <sopt/utilities.h>
#include <sopt/wavelets.h>
#include <sopt/wavelets/sara.h>
using namespace purify;
class PadmmFixtureMPI : public ::benchmark::Fixture {
public:
void SetUp(const ::benchmark::State &state) {
// Reading image from file and update related quantities
bool newImage = b_utilities::updateImage(state.range(0), m_image, m_imsizex, m_imsizey);
// Generating random uv(w) coverage
bool newMeasurements = m_uv_data.size() != state.range(1);
if (newMeasurements) {
t_real const sigma_m = constant::pi / 3;
m_uv_data = utilities::random_sample_density(state.range(1), 0, sigma_m);
}
bool newKernel = m_kernel != state.range(2);
if (newImage || newMeasurements || newKernel) {
m_kernel = state.range(2);
// creating the measurement operator
const t_real FoV = 1; // deg
const t_real cellsize = FoV / m_imsizex * 60. * 60.;
const bool w_term = false;
// algorithm 1
if (state.range(4) == 1)
m_measurements1 = measurementoperator::init_degrid_operator_2d_mpi<Vector<t_complex>>(
m_world, m_uv_data, m_image.rows(), m_image.cols(), cellsize, cellsize, 2,
kernels::kernel::kb, m_kernel, m_kernel, w_term);
// algorithm 3
if (state.range(4) == 3)
m_measurements3 = measurementoperator::init_degrid_operator_2d<Vector<t_complex>>(
m_world, m_uv_data, m_image.rows(), m_image.cols(), cellsize, cellsize, 2,
kernels::kernel::kb, m_kernel, m_kernel, w_term);
}
}
void TearDown(const ::benchmark::State &state) {}
sopt::mpi::Communicator m_world;
const sopt::wavelets::SARA m_sara{
std::make_tuple("Dirac", 3u), std::make_tuple("DB1", 3u), std::make_tuple("DB2", 3u),
std::make_tuple("DB3", 3u), std::make_tuple("DB4", 3u), std::make_tuple("DB5", 3u),
std::make_tuple("DB6", 3u), std::make_tuple("DB7", 3u), std::make_tuple("DB8", 3u)};
Image<t_complex> m_image;
t_uint m_imsizex;
t_uint m_imsizey;
utilities::vis_params m_uv_data;
t_real m_epsilon;
t_uint m_kernel;
std::shared_ptr<sopt::LinearTransform<Vector<t_complex>> const> m_measurements1;
std::shared_ptr<sopt::LinearTransform<Vector<t_complex>> const> m_measurements3;
};
BENCHMARK_DEFINE_F(PadmmFixtureMPI, ApplyAlgo1)(benchmark::State &state) {
// Create the algorithm - somehow doesn't work if done in the fixture...
sopt::wavelets::SARA saraDistr = sopt::wavelets::distribute_sara(m_sara, m_world);
auto const Psi =
sopt::linear_transform<t_complex>(saraDistr, m_image.rows(), m_image.cols(), m_world);
t_real gamma =
utilities::step_size(m_uv_data.vis, m_measurements1,
std::make_shared<sopt::LinearTransform<Vector<t_complex>> const>(Psi),
saraDistr.size()) *
1e-3;
gamma = m_world.all_reduce(gamma, MPI_MAX);
std::shared_ptr<sopt::algorithm::ImagingProximalADMM<t_complex>> padmm =
std::make_shared<sopt::algorithm::ImagingProximalADMM<t_complex>>(m_uv_data.vis);
padmm->itermax(state.range(3) + 1)
.gamma(gamma)
.relative_variation(1e-3)
.l2ball_proximal_epsilon(m_epsilon)
// communicator ensuring l1 norm in l1 proximal is global
.l1_proximal_adjoint_space_comm(m_world)
.tight_frame(false)
.l1_proximal_tolerance(1e-2)
.l1_proximal_nu(1)
.l1_proximal_itermax(2)
.l1_proximal_positivity_constraint(true)
.l1_proximal_real_constraint(true)
.residual_tolerance(m_epsilon)
.lagrange_update_scale(0.9)
.nu(1e0)
.Psi(Psi)
.Phi(*m_measurements1);
std::weak_ptr<decltype(padmm)::element_type> const padmm_weak(padmm);
padmm->residual_convergence(
factory::l2_convergence_factory<t_complex>(factory::ConvergenceType::mpi_local, padmm_weak));
padmm->objective_convergence(
factory::l1_convergence_factory<t_complex>(factory::ConvergenceType::mpi_local, padmm_weak));
// Benchmark the application of the algorithm
while (state.KeepRunning()) {
auto start = std::chrono::high_resolution_clock::now();
auto result = (*padmm)();
auto end = std::chrono::high_resolution_clock::now();
// std::cout << "Converged? " << result.good << " , niters = " << result.niters << std::endl;
state.SetIterationTime(b_utilities::duration(start, end, m_world));
}
}
BENCHMARK_DEFINE_F(PadmmFixtureMPI, ApplyAlgo3)(benchmark::State &state) {
// Create the algorithm - somehow doesn't work if done in the fixture...
sopt::wavelets::SARA saraDistr = sopt::wavelets::distribute_sara(m_sara, m_world);
auto const Psi =
sopt::linear_transform<t_complex>(saraDistr, m_image.rows(), m_image.cols(), m_world);
t_real gamma =
utilities::step_size(m_uv_data.vis, m_measurements3,
std::make_shared<sopt::LinearTransform<Vector<t_complex>> const>(Psi),
saraDistr.size()) *
1e-3;
gamma = m_world.all_reduce(gamma, MPI_MAX);
std::shared_ptr<sopt::algorithm::ImagingProximalADMM<t_complex>> padmm =
std::make_shared<sopt::algorithm::ImagingProximalADMM<t_complex>>(m_uv_data.vis);
padmm->itermax(state.range(3) + 1)
.gamma(gamma)
.relative_variation(1e-3)
.l2ball_proximal_epsilon(m_epsilon)
// communicator ensuring l1 norm in l1 proximal is global
.l1_proximal_adjoint_space_comm(m_world)
.tight_frame(false)
.l1_proximal_tolerance(1e-2)
.l1_proximal_nu(1)
.l1_proximal_itermax(2)
.l1_proximal_positivity_constraint(true)
.l1_proximal_real_constraint(true)
.residual_tolerance(m_epsilon)
.lagrange_update_scale(0.9)
.nu(1e0)
.Psi(Psi)
.Phi(*m_measurements3);
std::weak_ptr<decltype(padmm)::element_type> const padmm_weak(padmm);
padmm->residual_convergence(
factory::l2_convergence_factory<t_complex>(factory::ConvergenceType::mpi_local, padmm_weak));
padmm->objective_convergence(
factory::l1_convergence_factory<t_complex>(factory::ConvergenceType::mpi_local, padmm_weak));
// Benchmark the application of the algorithm
while (state.KeepRunning()) {
auto start = std::chrono::high_resolution_clock::now();
auto result = (*padmm)();
auto end = std::chrono::high_resolution_clock::now();
// std::cout << "Converged? " << result.good << " , niters = " << result.niters << std::endl;
state.SetIterationTime(b_utilities::duration(start, end, m_world));
}
}
BENCHMARK_REGISTER_F(PadmmFixtureMPI, ApplyAlgo1)
//->Apply(b_utilities::Arguments)
->Args({1024, static_cast<t_int>(1e6), 4, 10, 1})
->Args({1024, static_cast<t_int>(5e6), 4, 10, 1})
->Args({1024, static_cast<t_int>(1e7), 4, 10, 1})
->Args({1024, static_cast<t_int>(5e7), 4, 10, 1})
->Args({1024, static_cast<t_int>(1e8), 4, 10, 1})
->Args({1024, static_cast<t_int>(5e8), 4, 10, 1})
//->Args({128, 1000, 4})
->UseManualTime()
->Repetitions(3) //->ReportAggregatesOnly(true)
->Unit(benchmark::kMillisecond);
BENCHMARK_REGISTER_F(PadmmFixtureMPI, ApplyAlgo3)
//->Apply(b_utilities::Arguments)
->Args({1024, static_cast<t_int>(1e6), 4, 10, 3})
->Args({1024, static_cast<t_int>(5e6), 4, 10, 3})
->Args({1024, static_cast<t_int>(1e7), 4, 10, 3})
->Args({1024, static_cast<t_int>(5e7), 4, 10, 3})
->Args({1024, static_cast<t_int>(1e8), 4, 10, 3})
->Args({1024, static_cast<t_int>(5e8), 4, 10, 3})
//->Args({128, 1000, 4})
->UseManualTime()
->Repetitions(3) //->ReportAggregatesOnly(true)
->Unit(benchmark::kMillisecond);
|
