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
|
#include "rayleighfitter.h"
#include <iostream>
#include <limits>
#include <boost/numeric/conversion/bounds.hpp>
#ifdef HAVE_GSL
#include <gsl/gsl_vector.h>
#include <gsl/gsl_blas.h>
#include <gsl/gsl_multifit_nlin.h>
static int fit_f(const gsl_vector* xvec, void* data, gsl_vector* f) {
const double sigma = gsl_vector_get(xvec, 0);
const double n = gsl_vector_get(xvec, 1);
size_t t = 0;
const RayleighFitter& fitter = *static_cast<RayleighFitter*>(data);
const LogHistogram& hist = *fitter._hist;
const double minVal = fitter._minVal;
const double maxVal = fitter._maxVal;
for (LogHistogram::const_iterator i = hist.begin(); i != hist.end(); ++i) {
const double x = i.value();
if (x >= minVal && x < maxVal && std::isfinite(x)) {
const double val = i.normalizedCount();
// const double logval = log(val);
// const double weight = logval;
const double sigmaP2 = sigma * sigma;
const double Yi = x * exp(-(x * x) / (2 * sigmaP2)) * n / sigmaP2;
if (fitter.FitLogarithmic())
gsl_vector_set(f, t, log(Yi) - log(val));
else
gsl_vector_set(f, t, (Yi - val));
++t;
}
}
return GSL_SUCCESS;
}
int fit_df(const gsl_vector* xvec, void* data, gsl_matrix* J) {
const double sigma = gsl_vector_get(xvec, 0);
const double n = gsl_vector_get(xvec, 1);
size_t t = 0;
const RayleighFitter& fitter = *static_cast<RayleighFitter*>(data);
const LogHistogram& hist = *fitter._hist;
const double minVal = fitter._minVal;
const double maxVal = fitter._maxVal;
const double sigmaP2 = sigma * sigma;
const double sigmaP3 = sigma * sigma * sigma;
for (LogHistogram::const_iterator i = hist.begin(); i != hist.end(); ++i) {
const double x = i.value();
if (x >= minVal && x < maxVal && std::isfinite(x)) {
// const double val = i.normalizedCount();
// const double weight = log(val);
double dfdsigma, dfdn;
if (fitter.FitLogarithmic()) {
dfdsigma = (x * x - 2.0 * sigmaP2) / sigmaP3;
dfdn = 1.0 / n;
} else {
dfdsigma = -n * 2.0 * x * x * x / (sigmaP3 * sigmaP3) *
exp(-x * x / (2.0 * sigmaP2));
dfdn = x * exp(-(x * x) / (2 * sigmaP2)) / sigmaP2;
}
// diff to sigma
gsl_matrix_set(J, t, 0, dfdsigma);
// diff to n
gsl_matrix_set(J, t, 1, dfdn);
++t;
}
}
return GSL_SUCCESS;
}
int fit_fdf(const gsl_vector* x, void* data, gsl_vector* f, gsl_matrix* J) {
fit_f(x, data, f);
fit_df(x, data, J);
return GSL_SUCCESS;
}
void print_state(size_t iter, gsl_multifit_fdfsolver* s) {
const double sigma = gsl_vector_get(s->x, 0);
const double N = gsl_vector_get(s->x, 1);
std::cout << "iteration " << iter << ", sigma=" << sigma << ", N=" << N
<< "\n";
}
void RayleighFitter::Fit(double minVal, double maxVal, const LogHistogram& hist,
double& sigma, double& n) {
unsigned int iter = 0;
const size_t nVars = 2;
_hist = &hist;
if (minVal > 0)
_minVal = minVal;
else
_minVal = hist.MinPositiveAmplitude();
_maxVal = maxVal;
if (sigma < minVal) sigma = minVal;
size_t nData = 0;
for (LogHistogram::iterator i = hist.begin(); i != hist.end(); ++i) {
const double val = i.value();
if (val >= minVal && val < maxVal && std::isfinite(val)) ++nData;
}
std::cout << "ndata=" << nData << "\n";
double x_init[nVars] = {sigma, n};
const gsl_vector_view x = gsl_vector_view_array(x_init, nVars);
gsl_multifit_function_fdf f;
f.f = &fit_f;
f.df = &fit_df;
f.fdf = &fit_fdf;
f.n = nData;
f.p = nVars;
f.params = this;
const gsl_multifit_fdfsolver_type* T = gsl_multifit_fdfsolver_lmsder;
gsl_multifit_fdfsolver* s = gsl_multifit_fdfsolver_alloc(T, nData, nVars);
gsl_multifit_fdfsolver_set(s, &f, &x.vector);
print_state(iter, s);
int status;
do {
iter++;
status = gsl_multifit_fdfsolver_iterate(s);
std::cout << "status = " << gsl_strerror(status) << "\n";
print_state(iter, s);
if (status) break;
status = gsl_multifit_test_delta(s->dx, s->x, 1e-7, 1e-3);
} while (status == GSL_CONTINUE && iter < 500);
std::cout << "status = " << gsl_strerror(status) << "\n";
print_state(iter, s);
sigma = fabs(gsl_vector_get(s->x, 0));
n = fabs(gsl_vector_get(s->x, 1));
gsl_multifit_fdfsolver_free(s);
}
#else // No gsl...
void RayleighFitter::Fit(double minVal, double maxVal, const LogHistogram& hist,
double& sigma, double& n) {
sigma = 1.0;
n = 1.0;
}
#endif
double RayleighFitter::SigmaEstimate(const LogHistogram& hist) {
return hist.AmplitudeWithMaxNormalizedCount();
}
double RayleighFitter::SigmaEstimate(const LogHistogram& hist,
double rangeStart, double rangeEnd) {
double maxAmplitude = 0.0,
maxNormalizedCount = boost::numeric::bounds<double>::lowest();
for (LogHistogram::const_iterator i = hist.begin(); i != hist.end(); ++i) {
if (i.value() > rangeStart && i.value() < rangeEnd &&
std::isfinite(i.value())) {
if (std::isfinite(i.normalizedCount())) {
if (i.normalizedCount() > maxNormalizedCount) {
maxAmplitude = i.value();
maxNormalizedCount = i.normalizedCount();
}
}
}
}
return maxAmplitude;
}
void RayleighFitter::FindFitRangeUnderRFIContamination(
double minPositiveAmplitude, double sigmaEstimate, double& minValue,
double& maxValue) {
minValue = minPositiveAmplitude;
maxValue = sigmaEstimate * 1.5;
std::cout << "Found range " << minValue << " -- " << maxValue << "\n";
}
double RayleighFitter::ErrorOfFit(const LogHistogram& histogram,
double rangeStart, double rangeEnd,
double sigma, double n) {
double sum = 0.0;
size_t count = 0;
for (LogHistogram::const_iterator i = histogram.begin(); i != histogram.end();
++i) {
const double x = i.value();
if (x >= rangeStart && x < rangeEnd && std::isfinite(x)) {
const double val = i.normalizedCount();
const double sigmaP2 = sigma * sigma;
const double Yi = x * exp(-(x * x) / (2 * sigmaP2)) * n / sigmaP2;
const double error = (Yi - val) * (Yi - val);
sum += error;
++count;
}
}
return sum / (double)count;
}
double RayleighFitter::NEstimate(const LogHistogram& hist, double rangeStart,
double rangeEnd) {
double rangeSum = 0.0;
size_t count = 0;
for (LogHistogram::const_iterator i = hist.begin(); i != hist.end(); ++i) {
if (i.value() > rangeStart && i.value() < rangeEnd &&
std::isfinite(i.value())) {
if (std::isfinite(i.normalizedCount())) {
rangeSum += i.normalizedCount();
++count;
}
}
}
return rangeSum / (count * 10.0);
}
|
