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
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
|
/*
* This file is part of the MOSCA library
* Copyright (C) 2013 European Southern Observatory
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
#include "config.h"
#if defined HAVE_BOOST_UNIT_TEST_FRAMEWORK && HAVE_CXX11
#define BOOST_TEST_DYN_LINK
#define BOOST_TEST_MODULE image_normalisation-test
#define _GLIBCXX_USE_NANOSLEEP
#include <random>
#include <boost/test/unit_test.hpp>
#include <boost/test/test_case_template.hpp>
#include <boost/mpl/list.hpp>
#include "cpl.h"
#include "image_normalisation.h"
BOOST_AUTO_TEST_SUITE(image_normalisation_exceptions)
BOOST_AUTO_TEST_CASE(empty_image)
{
cpl_init(CPL_INIT_DEFAULT);
cpl_errorstate prestate = cpl_errorstate_get();
size_t nx = 100;
size_t ny = 10;
mosca::image empty(nx, ny);
mosca::image weight(nx, ny);
std::vector<double> spa_profile;
std::vector<double> spec_profile;
BOOST_CHECK_NO_THROW(image_normalise(empty, weight, 0, 0, 0, 0, 0.,
spa_profile, spec_profile));
BOOST_REQUIRE_EQUAL(prestate, cpl_errorstate_get() );
cpl_end();
}
BOOST_AUTO_TEST_CASE(size_mismatch)
{
cpl_init(CPL_INIT_DEFAULT);
cpl_errorstate prestate = cpl_errorstate_get();
size_t nx = 100;
size_t ny = 10;
mosca::image empty(nx, ny);
mosca::image weight(ny, nx);
std::vector<double> spa_profile;
std::vector<double> spec_profile;
BOOST_CHECK_THROW(image_normalise(empty, weight, 0, 0, 0, 0, 0.,
spa_profile, spec_profile),
std::invalid_argument);
BOOST_REQUIRE_EQUAL(prestate, cpl_errorstate_get() );
cpl_end();
}
BOOST_AUTO_TEST_SUITE_END()
BOOST_AUTO_TEST_SUITE(image_normalisation_templates)
typedef boost::mpl::list<float, double, int> test_types;
BOOST_AUTO_TEST_CASE_TEMPLATE(smooth, T, test_types)
{
cpl_init(CPL_INIT_DEFAULT);
cpl_errorstate prestate = cpl_errorstate_get();
size_t nx = 100;
size_t ny = 10;
T flux = 1000;
cpl_type type = mosca::type_trait<T>::cpl_eq_type;
mosca::image flat_image(nx, ny, type);
mosca::image weight(nx, ny,type);
std::vector<T> spa_profile;
std::vector<T> spec_profile;
std::fill(flat_image.get_data<T>(), flat_image.get_data<T>() + nx*ny, flux);
std::fill(weight.get_data<T>(), weight.get_data<T>() + nx*ny, 1);
//Loop on various smoothing values in both directions
//TODO: If spa_smooth_radius is larger than half the size in spatial dir,
//then it fails the test. Check it
for (int spa_smooth_radius = 0; spa_smooth_radius <= 3; spa_smooth_radius++)
for (int disp_smooth_radius = 0; disp_smooth_radius <= 12; disp_smooth_radius++)
{
mosca::image smoothed =
image_normalise(flat_image, weight, spa_smooth_radius,
disp_smooth_radius, 0, 0, 0.,
spa_profile, spec_profile);
//When doing smoothing and the input is already
//flat we should be getting the same thing
//Actually even if not doing normalisation,
//should get something flat everywhere
BOOST_CHECK_EQUAL_COLLECTIONS(flat_image.get_data<T>(),
flat_image.get_data<T>() + nx*ny,
smoothed.get_data<T>(),
smoothed.get_data<T>() + nx*ny);
}
BOOST_REQUIRE_EQUAL(prestate, cpl_errorstate_get() );
cpl_end();
}
typedef boost::mpl::list<float, double> test_types_fp;
BOOST_TEST_DECORATOR( * boost::unit_test::tolerance(1e-6))
BOOST_TEST_DECORATOR( * boost::unit_test::tolerance((float)1e-6))
BOOST_AUTO_TEST_CASE_TEMPLATE(fit, T, test_types_fp)
{
cpl_init(CPL_INIT_DEFAULT);
cpl_errorstate prestate = cpl_errorstate_get();
size_t nx = 100;
size_t ny = 10;
T flux = 10;
cpl_type type = mosca::type_trait<T>::cpl_eq_type;
mosca::image flat_image(nx, ny, type);
mosca::image weight(nx, ny,type);
std::vector<T> spa_profile;
std::vector<T> spec_profile;
std::fill(flat_image.get_data<T>(), flat_image.get_data<T>() + nx*ny, flux);
std::fill(weight.get_data<T>(), weight.get_data<T>() + nx * ny, 1.0);
//Loop on various fitting values in both directions
for (int spa_fit_degree = 0; spa_fit_degree <= 3; spa_fit_degree++)
for (int disp_fit_nknots = 2; disp_fit_nknots <= 10; disp_fit_nknots++)
{
mosca::image smoothed =
image_normalise(flat_image, weight, 0, 0,
spa_fit_degree, disp_fit_nknots, 0.,
spa_profile, spec_profile);
//Check for close enough fits
std::vector<T> differences;
std::transform (flat_image.get_data<T>(),
flat_image.get_data<T>() + nx*ny,
smoothed.get_data<T>(),
std::back_inserter(differences), std::minus<T>());
std::vector<T> zero(differences.size());
BOOST_TEST(differences == zero, boost::test_tools::per_element());
}
//Check for case disp_fit_nknots==1 which is not support by gsl:
BOOST_CHECK_THROW(image_normalise(flat_image, weight, 0, 0,
0, 1, 0.,
spa_profile, spec_profile), std::exception);
BOOST_REQUIRE_EQUAL(prestate, cpl_errorstate_get() );
cpl_end();
}
BOOST_TEST_DECORATOR( * boost::unit_test::tolerance(1e-4))
BOOST_TEST_DECORATOR( * boost::unit_test::tolerance((float)1e-4))
BOOST_AUTO_TEST_CASE_TEMPLATE(disp_y, T, test_types_fp)
{
cpl_init(CPL_INIT_DEFAULT);
cpl_errorstate prestate = cpl_errorstate_get();
size_t ny = 100;
size_t nx = 10;
T av_flux = 100;
T x_span_factor = 0.1;
T y_span_factor = 0.1;
cpl_type type = mosca::type_trait<T>::cpl_eq_type;
mosca::image flat_image(nx, ny, type, mosca::Y_AXIS);
mosca::image weight(nx, ny, type, mosca::Y_AXIS);
std::vector<T> spa_profile;
std::vector<T> spec_profile;
std::fill(weight.get_data<T>(), weight.get_data<T>() + nx*ny, 1);
size_t ipix = 0;
std::generate(flat_image.get_data<T>(), flat_image.get_data<T>() + nx*ny,
[nx, ny, av_flux, x_span_factor, y_span_factor, &ipix]()
{//Linear gradient
size_t i = ipix % nx;
size_t j = ipix / nx;
T fx = std::sqrt(av_flux) * (1 + x_span_factor * ((i-nx/2.)/nx));
T fy = std::sqrt(av_flux) * (1 + y_span_factor * ((j-ny/2.)/ny));
ipix++;
return fx * fy ;
});
//Loop on various normalisation parameters
for (int spa_smooth_radius = 0; spa_smooth_radius <= 3; spa_smooth_radius++)
for (int disp_smooth_radius = 0; disp_smooth_radius <= 3; disp_smooth_radius++)
for (int spa_fit_degree = 1; spa_fit_degree <= 3; spa_fit_degree++)
for (int disp_fit_nknots = 2; disp_fit_nknots <= 4; disp_fit_nknots++)
{
mosca::image smoothed =
image_normalise(flat_image, weight,
spa_smooth_radius, disp_smooth_radius,
spa_fit_degree, disp_fit_nknots, 0.,
spa_profile, spec_profile);
//Check for close enough fits
std::vector<T> differences;
std::transform (flat_image.get_data<T>(),
flat_image.get_data<T>() + nx*ny,
smoothed.get_data<T>(),
std::back_inserter(differences), std::minus<T>());
std::vector<T> zero(differences.size());
BOOST_TEST(differences == zero, boost::test_tools::per_element());
}
BOOST_REQUIRE_EQUAL(prestate, cpl_errorstate_get() );
cpl_end();
}
BOOST_TEST_DECORATOR( * boost::unit_test::tolerance(1e-4))
BOOST_TEST_DECORATOR( * boost::unit_test::tolerance((float)1e-4))
BOOST_AUTO_TEST_CASE_TEMPLATE(gradient, T, test_types_fp)
{
cpl_init(CPL_INIT_DEFAULT);
cpl_errorstate prestate = cpl_errorstate_get();
size_t nx = 100;
size_t ny = 10;
T av_flux = 100;
cpl_type type = mosca::type_trait<T>::cpl_eq_type;
mosca::image flat_image(nx, ny, type);
mosca::image weight(nx, ny,type);
std::vector<T> spa_profile;
std::vector<T> spec_profile;
std::fill(weight.get_data<T>(), weight.get_data<T>() + nx*ny, 1);
//Loop on various gradient values in both directions
for (T x_span_factor = 0.1; x_span_factor <= 0.3; x_span_factor+=0.1)
for (T y_span_factor = 0.1; y_span_factor <= 0.3; y_span_factor+=0.1)
{
size_t ipix = 0;
std::generate(flat_image.get_data<T>(), flat_image.get_data<T>() + nx*ny,
[nx, ny, av_flux, x_span_factor, y_span_factor, &ipix]()
{//Linear gradient
size_t i = ipix % nx;
size_t j = ipix / nx;
T fx = std::sqrt(av_flux) * (1 + x_span_factor * ((i-nx/2.)/nx));
T fy = std::sqrt(av_flux) * (1 + y_span_factor * ((j-ny/2.)/ny));
ipix++;
return fx * fy ;
});
//Loop on various normalisation parameters
for (int spa_smooth_radius = 1; spa_smooth_radius <= 3; spa_smooth_radius++)
for (int disp_smooth_radius = 1; disp_smooth_radius <= 5; disp_smooth_radius++)
for (int spa_fit_degree = 1; spa_fit_degree <= 3; spa_fit_degree++)
for (int disp_fit_nknots = 2; disp_fit_nknots <= 6; disp_fit_nknots++)
{
mosca::image smoothed =
image_normalise(flat_image, weight,
spa_smooth_radius, disp_smooth_radius,
spa_fit_degree, disp_fit_nknots, 0.,
spa_profile, spec_profile);
//Check for close enough fits
std::vector<T> differences;
std::transform (flat_image.get_data<T>(),
flat_image.get_data<T>() + nx*ny,
smoothed.get_data<T>(),
std::back_inserter(differences), std::minus<T>());
std::vector<T> zero(differences.size());
BOOST_TEST(differences == zero, boost::test_tools::per_element());
}
}
BOOST_REQUIRE_EQUAL(prestate, cpl_errorstate_get() );
cpl_end();
}
BOOST_AUTO_TEST_SUITE_END()
BOOST_AUTO_TEST_SUITE(image_normalisation_weights)
typedef boost::mpl::list<float, double> test_types_fp;
BOOST_AUTO_TEST_CASE_TEMPLATE(weight_flat_smooth, T, test_types_fp)
{
cpl_init(CPL_INIT_DEFAULT);
cpl_errorstate prestate = cpl_errorstate_get();
size_t nx = 100;
size_t ny = 10;
T flux = 10;
cpl_type type = mosca::type_trait<T>::cpl_eq_type;
mosca::image flat_image(nx, ny, type);
mosca::image weight(nx, ny,type);
std::vector<T> spa_profile;
std::vector<T> spec_profile;
std::fill(flat_image.get_data<T>(), flat_image.get_data<T>() + nx*ny, flux);
std::fill(weight.get_data<T>(), weight.get_data<T>() + nx*ny, 1);
//Set the outer borders of the weights to 0 and the image to something bad
std::fill(weight.get_data<T>(), weight.get_data<T>() + nx, 0);
std::fill(weight.get_data<T>() + nx*(ny-1), weight.get_data<T>() + nx*ny, 0);
std::fill(flat_image.get_data<T>(), flat_image.get_data<T>() + nx, 100);
std::fill(flat_image.get_data<T>() + nx*(ny-1), flat_image.get_data<T>() + nx*ny, 50);
for(int iy=1;iy<=ny;iy++)
{
cpl_image_set(weight.get_cpl_image(), 1, iy, 0.);
cpl_image_set(weight.get_cpl_image(), nx, iy, 0.);
cpl_image_set(flat_image.get_cpl_image(), 1, iy, 20.);
cpl_image_set(flat_image.get_cpl_image(), nx, iy, 40.);
}
//One pixel in the corner is actually valid...
cpl_image_set(weight.get_cpl_image(), nx, ny, 1.);
cpl_image_set(flat_image.get_cpl_image(), nx, ny, flux);
//Loop on various smoothing values in both directions
//TODO: If spa_smooth_radius is larger than half the size in spatial dir,
//then it fails the test. Check it
for (int spa_smooth_radius = 0; spa_smooth_radius <= 3; spa_smooth_radius++)
for (int disp_smooth_radius = 0; disp_smooth_radius <= 12; disp_smooth_radius++)
{
mosca::image smoothed =
image_normalise(flat_image, weight, spa_smooth_radius,
disp_smooth_radius, 0, 0, 0.,
spa_profile, spec_profile);
//When doing smoothing and the input is already
//flat we should be getting the same thing
//Actually even if not doing normalisation,
//should get something flat everywhere
//Here we check only the regions with weights != 0
for(int iy=1;iy<ny-1;iy++)
{
BOOST_CHECK_EQUAL_COLLECTIONS(flat_image.get_data<T>() + 1 + iy * nx,
flat_image.get_data<T>() + iy * nx + nx -1,
smoothed.get_data<T>() + 1 + iy * nx,
smoothed.get_data<T>() + iy * nx + nx -1);
}
}
BOOST_REQUIRE_EQUAL(prestate, cpl_errorstate_get() );
cpl_end();
}
BOOST_TEST_DECORATOR( * boost::unit_test::tolerance(0.7))
BOOST_TEST_DECORATOR( * boost::unit_test::tolerance((float)0.7))
BOOST_AUTO_TEST_CASE_TEMPLATE(random_weight_gradient_smooth, T, test_types_fp)
{
cpl_init(CPL_INIT_DEFAULT);
cpl_errorstate prestate = cpl_errorstate_get();
size_t nx = 100;
size_t ny = 10;
T av_flux = 100;
T random_stddev = 1000;
float bad_frac = 0.20;
cpl_type type = mosca::type_trait<T>::cpl_eq_type;
mosca::image flat_image(nx, ny, type);
mosca::image weight(nx, ny,type);
std::vector<T> spa_profile;
std::vector<T> spec_profile;
std::fill(weight.get_data<T>(), weight.get_data<T>() + nx*ny, 1);
//Loop on various gradient values in both directions
for (T x_span_factor = 0.0; x_span_factor <= 0.3; x_span_factor+=0.1)
for (T y_span_factor = 0.0; y_span_factor <= 0.3; y_span_factor+=0.1)
{
size_t ipix = 0;
std::generate(flat_image.get_data<T>(), flat_image.get_data<T>() + nx*ny,
[nx, ny, av_flux, x_span_factor, y_span_factor, &ipix]()
{//Linear gradient
size_t i = ipix % nx;
size_t j = ipix / nx;
T fx = std::sqrt(av_flux) * (1 + x_span_factor * ((i-nx/2.)/nx));
T fy = std::sqrt(av_flux) * (1 + y_span_factor * ((j-ny/2.)/ny));
ipix++;
return fx * fy;
});
//Distribute randomly invalid pixels
std::fill(weight.get_data<T>(), weight.get_data<T>() + nx*ny, 1);
std::default_random_engine rnd;
std::uniform_int_distribution<> xpix_dist(1, nx);
std::uniform_int_distribution<> ypix_dist(1, ny);
std::normal_distribution<> val_dist(av_flux, random_stddev);
for(size_t ibad = 0; (float)ibad < nx*ny*bad_frac; ++ibad)
{
size_t xpix = xpix_dist(rnd);
size_t ypix = ypix_dist(rnd);
T im_val = val_dist(rnd);
cpl_image_set(weight.get_cpl_image(), xpix, ypix, 0.);
cpl_image_set(flat_image.get_cpl_image(), xpix, ypix, im_val);
}
//Loop on various normalisation parameters
//Since this is not a flat image, it doesn't make much sense
//to try with the smoothing parameters, since they are not
//really able to cope with the gradient.
for (int spa_fit_degree = 1; spa_fit_degree <= 2; spa_fit_degree++)
for (int disp_fit_nknots = 3; disp_fit_nknots <= 6; disp_fit_nknots++)
{
mosca::image smoothed =
image_normalise(flat_image, weight,
0, 0,
spa_fit_degree, disp_fit_nknots, 0.,
spa_profile, spec_profile);
//Test in the pixels where the mask is valid
mosca::image flat_w = flat_image;
std::transform (flat_image.get_data<T>(), flat_image.get_data<T>() + nx*ny,
weight.get_data<T>(),
flat_w.get_data<T>(), std::multiplies<T>());
mosca::image smooth_w = smoothed;
std::transform (smooth_w.get_data<T>(), smooth_w.get_data<T>() + nx*ny,
weight.get_data<T>(),
smooth_w.get_data<T>(), std::multiplies<T>());
//Check for close enough fits. The tolerance is
//actually quite high 0.7, but that's the best that can
//be done with 20% of the points masked
std::vector<T> differences;
std::transform (flat_w.get_data<T>(),
flat_w.get_data<T>() + nx*ny,
smooth_w.get_data<T>(),
std::back_inserter(differences), std::minus<T>());
std::vector<T> zero(differences.size());
BOOST_TEST(differences == zero, boost::test_tools::per_element());
}
}
BOOST_REQUIRE_EQUAL(prestate, cpl_errorstate_get() );
cpl_end();
}
BOOST_AUTO_TEST_SUITE_END()
#else
int main(void)
{
return 0;
}
#endif
|
