-
Notifications
You must be signed in to change notification settings - Fork 0
/
malmconvergence.h
227 lines (190 loc) · 6.22 KB
/
malmconvergence.h
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
/**********************************************************************
* Regression and stability estimation.
* Copyright (C) 2013 Georg Rudoy
*
* Boost Software License - Version 1.0 - August 17th, 2003
*
* Permission is hereby granted, free of charge, to any person or organization
* obtaining a copy of the software and accompanying documentation covered by
* this license (the "Software") to use, reproduce, display, distribute,
* execute, and transmit the Software, and to prepare derivative works of the
* Software, and to permit third-parties to whom the Software is furnished to
* do so, all subject to the following:
*
* The copyright notices in the Software and this entire statement, including
* the above license grant, this restriction and the following disclaimer,
* must be included in all copies of the Software, in whole or in part, and
* all derivative works of the Software, unless such copies or derivative
* works are solely in the form of machine-executable object code generated by
* a source language processor.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE, TITLE AND NON-INFRINGEMENT. IN NO EVENT
* SHALL THE COPYRIGHT HOLDERS OR ANYONE DISTRIBUTING THE SOFTWARE BE LIABLE
* FOR ANY DAMAGES OR OTHER LIABILITY, WHETHER IN CONTRACT, TORT OR OTHERWISE,
* ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
* DEALINGS IN THE SOFTWARE.
**********************************************************************/
#pragma once
#include <random>
#include "defs.h"
#include "threadpool.h"
DType_t clamp (DType_t from, DType_t to, DType_t x, DType_t xDev)
{
if (x + xDev <= to && x + xDev >= from)
return x + xDev;
if (x + xDev > to)
{
if (x - xDev >= from)
return x - xDev;
return 2 * x >= to - from ? to : from;
}
else
{
if (x - xDev >= to)
return x - xDev;
return 2 * x >= to - from ? to : from;
}
}
template<
typename Model,
typename YSigmaGetterT,
typename XSigmasGetterT
>
TrainingSet_t<> genSample (size_t size, DType_t from, DType_t to,
const YSigmaGetterT& ySigma,
const XSigmasGetterT& xSigma,
const Params_t<Model::ParamsCount>& params)
{
std::mt19937_64 generator { std::random_device {} () };
std::uniform_real_distribution<DType_t> rawXDistr { from, to };
TrainingSet_t<> result;
for (size_t i = 0; i < size; ++i)
{
const auto rawX = rawXDistr (generator);
const double xArr [] = { rawX };
const SampleType_t<> pseudoSample { xArr };
auto preprocessed = Model::preprocess ({ { pseudoSample, 0 } });
const auto rawY = Model::residual (preprocessed [0], params);
const TrainingSetInstance_t<> pair { pseudoSample, rawY };
const auto yDev = std::normal_distribution<DType_t> { 0, ySigma (pair) } (generator);
const auto xDev = std::normal_distribution<DType_t> { 0, xSigma (pair) } (generator);
SampleType_t<> sample;
sample (0) = rawX + xDev;
result.push_back ({ sample, rawY + yDev });
}
return result;
}
template<size_t ParamsCount>
struct SingleCompareResult
{
Params_t<ParamsCount> m_classicalParams;
Params_t<ParamsCount> m_modifiedParams;
SingleCompareResult ()
{
for (size_t i = 0; i < ParamsCount; ++i)
{
m_classicalParams (i) = 0;
m_modifiedParams (i) = 0;
}
}
SingleCompareResult (const Params_t<ParamsCount>& c, const Params_t<ParamsCount>& m)
: m_classicalParams { c }
, m_modifiedParams { m }
{
}
SingleCompareResult& operator+= (const SingleCompareResult& other)
{
m_classicalParams += other.m_classicalParams;
m_modifiedParams += other.m_modifiedParams;
return *this;
}
SingleCompareResult operator- () const
{
return { -m_classicalParams, -m_modifiedParams };
}
SingleCompareResult& operator-= (const SingleCompareResult& other)
{
return (*this) += (-other);
}
friend SingleCompareResult operator+ (SingleCompareResult left, const SingleCompareResult& right)
{
left += right;
return left;
}
friend SingleCompareResult operator- (SingleCompareResult left, const SingleCompareResult& right)
{
left -= right;
return left;
}
SingleCompareResult& abs ()
{
for (size_t i = 0; i < ParamsCount; ++i)
{
m_classicalParams (i) = std::abs (m_classicalParams (i));
m_modifiedParams (i) = std::abs (m_modifiedParams (i));
}
return *this;
}
};
template<
typename Model,
typename YSigmaGetterT,
typename XSigmasGetterT
>
SingleCompareResult<Model::ParamsCount> compareFunctionals (size_t size, DType_t from, DType_t to,
const YSigmaGetterT& ySigma,
const XSigmasGetterT& xSigma,
const Params_t<Model::ParamsCount>& params,
double multiplier)
{
const auto& trainingSet = genSample<Model> (size, from, to, ySigma, xSigma, params);
const auto& preprocessed = Model::preprocess (trainingSet);
const auto& classicP = solve<Model::ParamsCount> (preprocessed,
Model::residual, Model::residualDer, Model::initial ());
const auto& fixedP = solve<Model::ParamsCount> (preprocessed,
Model::residual, Model::residualDer, Model::varsDer,
ySigma,
xSigma,
Model::initial (),
multiplier);
return { classicP, fixedP };
}
template<
typename Model,
typename YSigmaGetterT,
typename XSigmasGetterT
>
auto compareFunctionals (size_t sizeFrom, size_t sizeTo,
int repetitions,
DType_t pointFrom, DType_t pointTo,
const YSigmaGetterT& ySigma,
const XSigmasGetterT& xSigma,
const Params_t<Model::ParamsCount>& params,
double multiplier)
{
using SingleResult_t = SingleCompareResult<Model::ParamsCount>;
std::vector<SingleResult_t> result;
result.resize (sizeTo - sizeFrom + 1);
const SingleResult_t reference { params, params };
{
ThreadPool pool;
std::mutex outMutex;
for (auto size = sizeFrom; size <= sizeTo; ++size)
pool << [&, size]
{
{
std::lock_guard<std::mutex> lock { outMutex };
std::cout << "\tdoing " << size << std::endl;
}
SingleResult_t subres;
for (size_t i = 0; i < repetitions; ++i)
subres += (compareFunctionals<Model> (size, pointFrom, pointTo, ySigma, xSigma, params, multiplier) - reference).abs ();
subres.m_classicalParams /= repetitions;
subres.m_modifiedParams /= repetitions;
result [size - sizeFrom] = subres;
};
}
return result;
}