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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263
|
// ************************************************************************************************
//
// BornAgain: simulate and fit reflection and scattering
//
//! @file Sim/Simulation/ISimulation.cpp
//! @brief Implements interface ISimulation.
//!
//! @homepage http://www.bornagainproject.org
//! @license GNU General Public License v3 or higher (see COPYING)
//! @copyright Forschungszentrum Jülich GmbH 2018
//! @authors Scientific Computing Group at MLZ (see CITATION, AUTHORS)
//
// ************************************************************************************************
#include "Sim/Simulation/ISimulation.h"
#include "Base/Progress/ProgressHandler.h"
#include "Base/Util/Assert.h"
#include "Base/Util/StringUtil.h"
#include "Device/Data/Datafield.h"
#include "Param/Distrib/DistributionHandler.h"
#include "Resample/Option/SimulationOptions.h"
#include "Resample/Processed/ReSample.h"
#include "Sample/Multilayer/Sample.h"
#include "Sim/Background/IBackground.h"
#include <algorithm>
#include <gsl/gsl_errno.h>
#include <iomanip>
#include <iostream>
#include <mutex>
#include <thread>
namespace {
size_t indexStep(size_t total_size, size_t n_handlers)
{
ASSERT(total_size > 0);
ASSERT(n_handlers > 0);
size_t result = total_size / n_handlers;
return total_size % n_handlers ? ++result : result;
}
size_t startIndex(size_t n_handlers, size_t current_handler, size_t n_elements)
{
const size_t handler_size = indexStep(n_elements, n_handlers);
const size_t start_index = current_handler * handler_size;
if (start_index >= n_elements)
return n_elements;
return start_index;
}
size_t batchSize(size_t n_handlers, size_t current_handler, size_t n_elements)
{
const size_t handler_size = indexStep(n_elements, n_handlers);
const size_t start_index = current_handler * handler_size;
if (start_index >= n_elements)
return 0;
return std::min(handler_size, n_elements - start_index);
}
} // namespace
// ************************************************************************************************
// class implementation
// ************************************************************************************************
ISimulation::ISimulation(const Sample& sample)
: m_sample(sample.clone())
, m_options(std::make_unique<SimulationOptions>())
, m_distribution_handler(std::make_unique<DistributionHandler>())
, m_progress(std::make_unique<ProgressHandler>())
{
ASSERT(m_sample);
}
ISimulation::~ISimulation() = default;
//... Setters:
void ISimulation::setBackground(const IBackground& bg)
{
m_background.reset(bg.clone());
}
void ISimulation::subscribe(const std::function<bool(size_t)>& inform)
{
ASSERT(m_progress);
m_progress->subscribe(inform);
}
//! Initializes a progress monitor that prints to stdout.
void ISimulation::setTerminalProgressMonitor()
{
#ifndef SILENT_PROGRESS
subscribe([](size_t percentage_done) -> bool {
if (percentage_done < 100)
std::cout << std::setprecision(2) << "\r... " << percentage_done << "%" << std::flush;
else // wipe out
std::cout << "\r... 100%\n";
return true;
});
#endif
}
const SimulationOptions& ISimulation::options() const
{
ASSERT(m_options);
return *m_options;
}
//... Executor:
//! Runs simulation with possible averaging over parameter distributions; returns result.
Datafield ISimulation::simulate()
{
ASSERT(m_sample);
const std::string errs = m_sample->validateAmbientSubstrate();
if (!errs.empty())
throw std::runtime_error("Invalid sample model: " + errs + ".");
gsl_set_error_handler_off();
prepareSimulation();
m_cache = std::vector<double>(nOutChannels(), 0.);
const ReSample re_sample = ReSample::make(*m_sample, options(), force_polarized());
const size_t total_size = nElements();
if (total_size == 0)
throw std::runtime_error("No output pixels. All masked? Invalid axis?");
size_t n_combinations = distributionHandler().nParamSamples();
m_progress->reset();
m_progress->setExpectedNTicks(n_combinations * total_size);
// restrict calculation to current batch // TODO: clarify
const size_t n_batches = m_options->getNumberOfBatches();
const size_t current_batch = m_options->getCurrentBatch();
const size_t batch_start = startIndex(n_batches, current_batch, total_size);
const size_t batch_size = batchSize(n_batches, current_batch, total_size);
ASSERT(batch_size);
if (n_combinations == 1) {
runSingleSimulation(re_sample, batch_start, batch_size, 1.);
} else {
// only GISAS
initDistributionHandler();
for (size_t index = 0; index < n_combinations; ++index) {
double weight = distributionHandler().setParameterValues(index);
runSingleSimulation(re_sample, batch_start, batch_size, weight);
}
}
return packResult();
}
//... Getters:
std::vector<const INode*> ISimulation::nodeChildren() const
{
std::vector<const INode*> result;
if (m_sample)
result << m_sample.get();
return result;
}
const Sample* ISimulation::sample() const
{
return m_sample.get();
}
const IBackground* ISimulation::background() const
{
return m_background.get();
}
const std::vector<ParameterDistribution>& ISimulation::paramDistributions() const
{
return m_distribution_handler->paramDistributions();
}
SimulationOptions& ISimulation::options()
{
ASSERT(m_options);
return *m_options;
}
//... Protected accessors:
ProgressHandler& ISimulation::progress()
{
ASSERT(m_progress);
return *m_progress;
}
DistributionHandler& ISimulation::distributionHandler()
{
ASSERT(m_distribution_handler);
return *m_distribution_handler;
}
//... Private executor:
//! Runs a single simulation with fixed parameter values.
//! If desired, the simulation is run in several threads.
void ISimulation::runSingleSimulation(const ReSample& re_sample, size_t batch_start,
size_t batch_size, double weight)
{
initScanElementVector();
const size_t n_threads = m_options->getNumberOfThreads();
ASSERT(n_threads > 0);
if (n_threads == 1) {
// Run computation in current thread.
try {
for (size_t i = 0; i < batch_size; ++i) {
if (!m_progress->alive())
break;
runComputation(re_sample, batch_start + i, weight);
}
} catch (const std::exception& ex) {
throw std::runtime_error(std::string("Unexpected error in simulation:\n") + ex.what());
}
} else {
// Launch computation threads.
std::vector<std::unique_ptr<std::thread>> threads;
std::vector<std::string> failure_messages;
std::mutex mutex;
for (size_t i_thread = 0; i_thread < n_threads; ++i_thread) {
const size_t thread_start = batch_start + startIndex(n_threads, i_thread, batch_size);
const size_t thread_size = batchSize(n_threads, i_thread, batch_size);
if (thread_size == 0)
break;
threads.emplace_back(new std::thread(
[this, &re_sample, &weight, &failure_messages, &mutex, thread_start, thread_size] {
try {
for (size_t i = 0; i < thread_size; ++i) {
if (!m_progress->alive())
break;
runComputation(re_sample, thread_start + i, weight);
}
} catch (const std::exception& ex) {
mutex.lock();
if (std::find(failure_messages.begin(), failure_messages.end(), ex.what())
== failure_messages.end())
failure_messages.emplace_back(ex.what());
mutex.unlock();
}
}));
}
// Wait for threads to complete.
for (auto& thread : threads)
thread->join();
// Check successful completion.
if (!failure_messages.empty())
throw std::runtime_error("Unexpected error(s) in simulation thread(s):\n"
+ Base::String::join(failure_messages, "\n"));
}
}
|