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
|
// ************************************************************************************************
//
// BornAgain: simulate and fit reflection and scattering
//
//! @file Fit/Adapter/MinimizerAdapter.cpp
//! @brief Implements class MinimizerAdapter.
//!
//! @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 "Fit/Adapter/MinimizerAdapter.h"
#include "Fit/Adapter/ObjectiveFunctionAdapter.h"
#include "Fit/Adapter/Report.h"
#include "Fit/Residual/RootResidualFunction.h"
#include "Fit/Residual/RootScalarFunction.h"
#include "Fit/Tool/StringUtil.h"
#include <Math/Minimizer.h>
#include <utility>
using namespace mumufit;
MinimizerAdapter::MinimizerAdapter(const MinimizerInfo& minimizerInfo)
: m_minimizer_info(minimizerInfo)
, m_adapter(new mumufit::ObjectiveFunctionAdapter)
, m_status(false)
{
}
MinimizerAdapter::~MinimizerAdapter() = default;
MinimizerResult MinimizerAdapter::minimize_scalar(fcn_scalar_t fcn, Parameters parameters)
{
// Genetic minimizer requires SetFunction before setParameters, others don't care
rootMinimizer()->SetFunction(*m_adapter->rootObjectiveFunction(fcn, parameters));
return minimize(parameters);
}
MinimizerResult MinimizerAdapter::minimize_residual(fcn_residual_t fcn, Parameters parameters)
{
// Genetic minimizer requires SetFunction before setParameters, others don't care
rootMinimizer()->SetFunction(*m_adapter->rootResidualFunction(fcn, parameters));
return minimize(parameters);
}
MinimizerResult MinimizerAdapter::minimize(Parameters parameters)
{
setParameters(parameters);
propagateOptions();
m_status = rootMinimizer()->Minimize();
propagateResults(parameters);
MinimizerResult result;
result.setParameters(parameters);
result.setMinValue(minValue());
result.setReport(internal::reportToString(*this));
result.setNumberOfCalls(m_adapter->numberOfCalls());
result.setNumberOfGradientCalls(m_adapter->numberOfGradientCalls());
return result;
}
std::string MinimizerAdapter::minimizerName() const
{
return m_minimizer_info.name();
}
std::string MinimizerAdapter::algorithmName() const
{
return m_minimizer_info.algorithmName();
}
void MinimizerAdapter::setParameters(const mumufit::Parameters& parameters)
{
unsigned int index(0);
for (const auto& par : parameters)
setParameter(index++, par);
}
double MinimizerAdapter::minValue() const
{
return rootMinimizer()->MinValue();
}
std::string MinimizerAdapter::statusToString() const
{
return m_status ? "Minimum found" : "Error in solving";
}
bool MinimizerAdapter::providesError() const
{
return rootMinimizer()->ProvidesError();
}
std::map<std::string, std::string> MinimizerAdapter::statusMap() const
{
std::map<std::string, std::string> result;
result["Status"] = statusToString();
if (providesError())
result["ProvidesError"] = "Provides parameters error and error matrix";
else
result["ProvidesError"] = "Doesn't provide error calculation";
result["MinValue"] = mumufit::stringUtil::scientific(minValue());
return result;
}
void MinimizerAdapter::setOptions(const std::string& optionString)
{
options().setOptionString(optionString);
}
//! Propagates results of minimization to fit parameter set
void MinimizerAdapter::propagateResults(mumufit::Parameters& parameters)
{
parameters.setValues(parValuesAtMinimum());
parameters.setErrors(parErrorsAtMinimum());
// sets correlation matrix
if (providesError()) {
auto lambda = [&](size_t i, size_t j) -> double {
return rootMinimizer()->Correlation(static_cast<unsigned int>(i),
static_cast<unsigned int>(j));
};
double2d_t matrix = FieldUtil::make<double>(fitRank(), fitRank(), lambda);
parameters.setCorrelationMatrix(matrix);
}
}
void MinimizerAdapter::setParameter(unsigned int index, const mumufit::Parameter& par)
{
bool success;
if (par.limits().isFixed())
success = rootMinimizer()->SetFixedVariable(index, par.name().c_str(), par.value());
else if (par.limits().isLimited())
success =
rootMinimizer()->SetLimitedVariable(index, par.name().c_str(), par.value(), par.step(),
par.limits().min(), par.limits().max());
else if (par.limits().isLowerLimited())
success = rootMinimizer()->SetLowerLimitedVariable(index, par.name().c_str(), par.value(),
par.step(), par.limits().min());
else if (par.limits().isUpperLimited())
success = rootMinimizer()->SetUpperLimitedVariable(index, par.name().c_str(), par.value(),
par.step(), par.limits().max());
else if (par.limits().isLimitless())
success = rootMinimizer()->SetVariable(index, par.name().c_str(), par.value(), par.step());
else
throw std::runtime_error("BasicMinimizer::setParameter -> Error! Unexpected parameter.");
if (!success) {
std::ostringstream ostr;
ostr << "BasicMinimizer::setParameter -> Error! Cannot set minimizer's fit parameter";
ostr << "Index:" << index << " name '" << par.name() << "'";
throw std::runtime_error(ostr.str());
}
}
//! Returns number of fit parameters defined (i.e. dimension of the function to be minimized).
size_t MinimizerAdapter::fitRank() const
{
return rootMinimizer()->NDim();
}
//! Returns value of the variables at minimum.
std::vector<double> MinimizerAdapter::parValuesAtMinimum() const
{
std::vector<double> result;
result.resize(fitRank(), 0.0);
std::copy(rootMinimizer()->X(), rootMinimizer()->X() + fitRank(), result.begin());
return result;
}
//! Returns errors of the variables at minimum.
std::vector<double> MinimizerAdapter::parErrorsAtMinimum() const
{
std::vector<double> result;
result.resize(fitRank(), 0.0);
if (rootMinimizer()->Errors() != nullptr)
std::copy(rootMinimizer()->Errors(), rootMinimizer()->Errors() + fitRank(), result.begin());
return result;
}
MinimizerAdapter::root_minimizer_t* MinimizerAdapter::rootMinimizer()
{
return const_cast<root_minimizer_t*>(
static_cast<const MinimizerAdapter*>(this)->rootMinimizer());
}
|
