-
Notifications
You must be signed in to change notification settings - Fork 95
/
Copy pathPoissonLogLikelihoodWithLinearModelForMeanAndProjData.cxx
1499 lines (1321 loc) · 61 KB
/
PoissonLogLikelihoodWithLinearModelForMeanAndProjData.cxx
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
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
/*
Copyright (C) 2000 PARAPET partners
Copyright (C) 2000-2011, Hammersmith Imanet Ltd
Copyright (C) 2014, 2016-2024 University College London
This file is part of STIR.
SPDX-License-Identifier: Apache-2.0 AND License-ref-PARAPET-license
See STIR/LICENSE.txt for details
*/
/*!
\file
\ingroup GeneralisedObjectiveFunction
\brief Declaration of class stir::PoissonLogLikelihoodWithLinearModelForMeanAndProjData
\author Kris Thielemans
\author Matthew Jacobson
\author Nikos Efthimiou
\author Elise Emond
\author Robert Twyman
\author Sanida Mustafovic
\author PARAPET project
*/
#include "stir/recon_buildblock/PoissonLogLikelihoodWithLinearModelForMeanAndProjData.h"
#include "stir/VoxelsOnCartesianGrid.h"
#include "stir/recon_buildblock/TrivialBinNormalisation.h"
#include "stir/Succeeded.h"
#include "stir/RelatedViewgrams.h"
#include "stir/stream.h"
#include "stir/info.h"
#include "stir/warning.h"
#include "stir/error.h"
#include "stir/recon_buildblock/ProjectorByBinPair.h"
#include "stir/recon_buildblock/BinNormalisationFromProjData.h"
#include "stir/DiscretisedDensity.h"
#ifdef STIR_MPI
# include "stir/recon_buildblock/DistributedCachingInformation.h"
#endif
#include "stir/recon_buildblock/distributable.h"
// for get_symmetries_ptr()
#include "stir/DataSymmetriesForViewSegmentNumbers.h"
// include the following to set defaults
#ifndef USE_PMRT
# include "stir/recon_buildblock/ForwardProjectorByBinUsingRayTracing.h"
# include "stir/recon_buildblock/BackProjectorByBinUsingInterpolation.h"
#else
# include "stir/recon_buildblock/ForwardProjectorByBinUsingProjMatrixByBin.h"
# include "stir/recon_buildblock/ProjMatrixByBinUsingRayTracing.h"
#endif
#include "stir/recon_buildblock/BackProjectorByBinUsingProjMatrixByBin.h"
#include "stir/recon_buildblock/ProjectorByBinPairUsingSeparateProjectors.h"
#include "stir/recon_buildblock/find_basic_vs_nums_in_subsets.h"
#include "stir/recon_buildblock/BinNormalisationWithCalibration.h"
#include "stir/ProjDataInMemory.h"
#include "stir/Viewgram.h"
#include "stir/recon_array_functions.h"
#include "stir/is_null_ptr.h"
#include <iostream>
#include <algorithm>
#include <functional>
#include <sstream>
#ifdef STIR_MPI
# include "stir/recon_buildblock/distributed_functions.h"
#endif
#include "stir/CPUTimer.h"
#include "stir/info.h"
#include <boost/format.hpp>
using std::vector;
using std::pair;
using std::ends;
using std::max;
START_NAMESPACE_STIR
const int rim_truncation_sino = 0; // TODO get rid of this
template <typename TargetT>
const char* const PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::registered_name
= "PoissonLogLikelihoodWithLinearModelForMeanAndProjData";
template <typename TargetT>
void
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::set_defaults()
{
base_type::set_defaults();
this->input_filename = "";
this->max_segment_num_to_process = -1;
this->max_timing_pos_num_to_process = 0;
// KT 20/06/2001 disabled
// num_views_to_add=1;
this->proj_data_sptr.reset(); // MJ added
this->zero_seg0_end_planes = 0;
this->use_tofsens = false;
this->additive_projection_data_filename = "0";
this->additive_proj_data_sptr.reset();
// set default for projector_pair_ptr
#ifndef USE_PMRT
shared_ptr<ForwardProjectorByBin> forward_projector_ptr(new ForwardProjectorByBinUsingRayTracing());
shared_ptr<BackProjectorByBin> back_projector_ptr(new BackProjectorByBinUsingInterpolation());
#else
shared_ptr<ProjMatrixByBinUsingRayTracing> PM(new ProjMatrixByBinUsingRayTracing());
// PM->set_num_tangential_LORs(5);
shared_ptr<ForwardProjectorByBin> forward_projector_ptr(new ForwardProjectorByBinUsingProjMatrixByBin(PM));
shared_ptr<BackProjectorByBin> back_projector_ptr(new BackProjectorByBinUsingProjMatrixByBin(PM));
#endif
this->projector_pair_ptr.reset(new ProjectorByBinPairUsingSeparateProjectors(forward_projector_ptr, back_projector_ptr));
this->normalisation_sptr.reset(new TrivialBinNormalisation);
this->frame_num = 1;
this->frame_definition_filename = "";
// make a single frame starting from 0 to 1.
vector<pair<double, double>> frame_times(1, pair<double, double>(0, 1));
this->frame_defs = TimeFrameDefinitions(frame_times);
this->target_parameter_parser.set_defaults();
#ifdef STIR_MPI
// distributed stuff
this->distributed_cache_enabled = false;
this->distributed_tests_enabled = false;
this->message_timings_enabled = false;
this->message_timings_threshold = 0.1;
this->rpc_timings_enabled = false;
#endif
}
template <typename TargetT>
void
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::initialise_keymap()
{
base_type::initialise_keymap();
this->parser.add_start_key("PoissonLogLikelihoodWithLinearModelForMeanAndProjData Parameters");
this->parser.add_stop_key("End PoissonLogLikelihoodWithLinearModelForMeanAndProjData Parameters");
this->parser.add_key("use time-of-flight sensitivities", &this->use_tofsens);
this->parser.add_key("input file", &this->input_filename);
// KT 20/06/2001 disabled
// parser.add_key("mash x views", &num_views_to_add);
this->parser.add_key("maximum absolute segment number to process", &this->max_segment_num_to_process);
this->parser.add_key("zero end planes of segment 0", &this->zero_seg0_end_planes);
this->target_parameter_parser.add_to_keymap(this->parser);
this->parser.add_parsing_key("Projector pair type", &this->projector_pair_ptr);
this->parser.add_key("additive sinogram", &this->additive_projection_data_filename);
// normalisation (and attenuation correction)
this->parser.add_key("time frame definition filename", &this->frame_definition_filename);
this->parser.add_key("time frame number", &this->frame_num);
this->parser.add_parsing_key("Bin Normalisation type", &this->normalisation_sptr);
#ifdef STIR_MPI
// distributed stuff
this->parser.add_key("enable distributed caching", &distributed_cache_enabled);
this->parser.add_key("enable distributed tests", &distributed_tests_enabled);
this->parser.add_key("enable message timings", &message_timings_enabled);
this->parser.add_key("message timings threshold", &message_timings_threshold);
this->parser.add_key("enable rpc timings", &rpc_timings_enabled);
#endif
}
template <typename TargetT>
bool
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::post_processing()
{
if (base_type::post_processing() == true)
return true;
// KT 20/06/2001 disabled as not functional yet
#if 0
if (num_views_to_add!=1 && (num_views_to_add<=0 || num_views_to_add%2 != 0))
{ warning("The 'mash x views' key has an invalid value (must be 1 or even number)"); return true; }
#endif
if (this->input_filename.length() > 0)
{
this->proj_data_sptr = ProjData::read_from_file(input_filename);
if (is_null_ptr(this->proj_data_sptr))
{
error("Failed to read input file %s", input_filename.c_str());
return true;
}
}
target_parameter_parser.check_values();
if (this->additive_projection_data_filename != "0")
{
info(boost::format("Reading additive projdata data %1%") % this->additive_projection_data_filename);
this->additive_proj_data_sptr = ProjData::read_from_file(this->additive_projection_data_filename);
};
// read time frame def
if (this->frame_definition_filename.size() != 0)
this->frame_defs = TimeFrameDefinitions(this->frame_definition_filename);
else
{
// make a single frame starting from 0 to 1.
vector<pair<double, double>> frame_times(1, pair<double, double>(0, 1));
this->frame_defs = TimeFrameDefinitions(frame_times);
}
#ifndef STIR_MPI
# if 0
//check caching enabled value
if (this->distributed_cache_enabled==true)
{
warning("STIR must be compiled with MPI-compiler to use distributed caching.\n\tDistributed Caching support will be disabled!");
this->distributed_cache_enabled=false;
}
//check tests enabled value
if (this->distributed_tests_enabled==true || rpc_timings_enabled==true || message_timings_enabled==true)
{
warning("STIR must be compiled with MPI-compiler and debug symbols to use distributed testing.\n\tDistributed tests will not be performed!");
this->distributed_tests_enabled=false;
}
# endif
#else
// check caching enabled value
if (this->distributed_cache_enabled == true)
info("Will use distributed caching!");
else
info("Distributed caching is disabled. Will use standard distributed version without forced caching!");
# ifndef NDEBUG
// check tests enabled value
if (this->distributed_tests_enabled == true)
{
warning("\nWill perform distributed tests! Beware that this decreases the performance");
distributed::test = true;
}
# else
// check tests enabled value
if (this->distributed_tests_enabled == true)
{
warning("\nDistributed tests only abvailable in debug mode!");
distributed::test = false;
}
# endif
// check timing values
if (this->message_timings_enabled == true)
{
info("Will print timings of MPI-Messages! This is used to find bottlenecks!");
distributed::test_send_receive_times = true;
}
// set timing threshold
distributed::min_threshold = this->message_timings_threshold;
if (this->rpc_timings_enabled == true)
{
info("Will print run-times of processing RPC_process_related_viewgrams_gradient for every slave! This will give an idea of "
"the parallelization effect!");
distributed::rpc_time = true;
}
#endif
// this->already_setup = false;
return false;
}
template <typename TargetT>
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::PoissonLogLikelihoodWithLinearModelForMeanAndProjData()
{
this->set_defaults();
}
template <typename TargetT>
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::~PoissonLogLikelihoodWithLinearModelForMeanAndProjData()
{
end_distributable_computation();
}
template <typename TargetT>
TargetT*
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::construct_target_ptr() const
{
return target_parameter_parser.create(this->get_input_data());
}
/***************************************************************
get_ functions
***************************************************************/
template <typename TargetT>
const ProjData&
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::get_proj_data() const
{
return *this->proj_data_sptr;
}
template <typename TargetT>
const shared_ptr<ProjData>&
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::get_proj_data_sptr() const
{
return this->proj_data_sptr;
}
template <typename TargetT>
const int
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::get_max_segment_num_to_process() const
{
return this->max_segment_num_to_process;
}
template <typename TargetT>
const int
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::get_max_timing_pos_num_to_process() const
{
return this->max_timing_pos_num_to_process;
}
template <typename TargetT>
const bool
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::get_zero_seg0_end_planes() const
{
return this->zero_seg0_end_planes;
}
template <typename TargetT>
const ProjData&
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::get_additive_proj_data() const
{
return *this->additive_proj_data_sptr;
}
template <typename TargetT>
const shared_ptr<ProjData>&
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::get_additive_proj_data_sptr() const
{
return this->additive_proj_data_sptr;
}
template <typename TargetT>
const ProjectorByBinPair&
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::get_projector_pair() const
{
return *this->projector_pair_ptr;
}
template <typename TargetT>
const shared_ptr<ProjectorByBinPair>&
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::get_projector_pair_sptr() const
{
return this->projector_pair_ptr;
}
template <typename TargetT>
const int
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::get_time_frame_num() const
{
return this->frame_num;
}
template <typename TargetT>
const TimeFrameDefinitions&
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::get_time_frame_definitions() const
{
return this->frame_defs;
}
template <typename TargetT>
const BinNormalisation&
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::get_normalisation() const
{
return *this->normalisation_sptr;
}
template <typename TargetT>
const shared_ptr<BinNormalisation>&
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::get_normalisation_sptr() const
{
return this->normalisation_sptr;
}
/***************************************************************
set_ functions
***************************************************************/
template <typename TargetT>
int
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::set_num_subsets(const int new_num_subsets)
{
this->already_set_up = this->already_set_up && (this->num_subsets == new_num_subsets);
this->num_subsets = std::max(new_num_subsets, 1);
return this->num_subsets;
}
template <typename TargetT>
void
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::set_proj_data_sptr(const shared_ptr<ProjData>& arg)
{
this->already_set_up = false;
this->proj_data_sptr = arg;
}
template <typename TargetT>
void
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::set_max_segment_num_to_process(const int arg)
{
this->already_set_up = this->already_set_up && (this->max_segment_num_to_process == arg);
this->max_segment_num_to_process = arg;
}
template <typename TargetT>
void
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::set_max_timing_pos_num_to_process(const int arg)
{
this->already_set_up = this->already_set_up && (this->max_timing_pos_num_to_process == arg);
this->max_timing_pos_num_to_process = arg;
}
template <typename TargetT>
void
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::set_zero_seg0_end_planes(const bool arg)
{
this->already_set_up = this->already_set_up && (this->zero_seg0_end_planes == arg);
this->zero_seg0_end_planes = arg;
}
template <typename TargetT>
void
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::set_additive_proj_data_sptr(const shared_ptr<ExamData>& arg)
{
this->already_set_up = false;
this->additive_proj_data_sptr = dynamic_pointer_cast<ProjData>(arg);
}
template <typename TargetT>
void
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::set_projector_pair_sptr(const shared_ptr<ProjectorByBinPair>& arg)
{
this->already_set_up = false;
this->projector_pair_ptr = arg;
}
template <typename TargetT>
void
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::set_frame_num(const int arg)
{
this->already_set_up = this->already_set_up && (this->frame_num == arg);
this->frame_num = arg;
}
template <typename TargetT>
void
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::set_frame_definitions(const TimeFrameDefinitions& arg)
{
this->already_set_up = this->already_set_up && (this->frame_defs == arg);
this->frame_defs = arg;
}
template <typename TargetT>
void
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::set_normalisation_sptr(const shared_ptr<BinNormalisation>& arg)
{
this->already_set_up = false;
this->normalisation_sptr = arg;
}
template <typename TargetT>
void
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::set_input_data(const shared_ptr<ExamData>& arg)
{
this->already_set_up = false;
this->proj_data_sptr = dynamic_pointer_cast<ProjData>(arg);
}
template <typename TargetT>
const ProjData&
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::get_input_data() const
{
return *this->proj_data_sptr;
}
/***************************************************************
subset balancing
***************************************************************/
template <typename TargetT>
bool
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::actual_subsets_are_approximately_balanced(
std::string& warning_message) const
{
assert(this->num_subsets > 0);
const DataSymmetriesForViewSegmentNumbers& symmetries
= *this->projector_pair_ptr->get_back_projector_sptr()->get_symmetries_used();
Array<1, int> num_vs_in_subset(this->num_subsets);
num_vs_in_subset.fill(0);
for (int subset_num = 0; subset_num < this->num_subsets; ++subset_num)
{
for (int segment_num = -this->max_segment_num_to_process; segment_num <= this->max_segment_num_to_process; ++segment_num)
for (int view_num = this->proj_data_sptr->get_min_view_num() + subset_num;
view_num <= this->proj_data_sptr->get_max_view_num();
view_num += this->num_subsets)
{
const ViewSegmentNumbers view_segment_num(view_num, segment_num);
if (!symmetries.is_basic(view_segment_num))
continue;
num_vs_in_subset[subset_num] += symmetries.num_related_view_segment_numbers(view_segment_num);
}
}
for (int subset_num = 1; subset_num < this->num_subsets; ++subset_num)
{
if (num_vs_in_subset[subset_num] != num_vs_in_subset[0])
{
std::stringstream str(warning_message);
str << "Number of subsets is such that subsets will be very unbalanced.\n"
<< "Number of viewgrams in each subset would be:\n"
<< num_vs_in_subset
<< "\nEither reduce the number of symmetries used by the projector, or\n"
"change the number of subsets. It usually should be a divisor of\n"
<< this->proj_data_sptr->get_num_views() << "/4 (or if that's not an integer, a divisor of "
<< this->proj_data_sptr->get_num_views() << "/2 or " << this->proj_data_sptr->get_num_views() << ").\n";
warning_message = str.str();
return false;
}
}
return true;
}
template <typename TargetT>
bool
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::sensitivity_uses_same_projector() const
{
return !this->proj_data_sptr->get_proj_data_info_sptr()->is_tof_data() || this->use_tofsens;
}
/***************************************************************
set_up()
***************************************************************/
template <typename TargetT>
void
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::ensure_norm_is_set_up(bool for_original_data) const
{
for_original_data = for_original_data || this->sensitivity_uses_same_projector();
if (for_original_data)
{
if (!this->norm_already_setup || !this->latest_setup_norm_was_with_orig_data)
{
if (this->normalisation_sptr->set_up(proj_data_sptr->get_exam_info_sptr(),
this->proj_data_sptr->get_proj_data_info_sptr())
== Succeeded::no)
error("Set_up of norm with original data failed.");
}
}
else
{
if (!this->norm_already_setup || this->latest_setup_norm_was_with_orig_data)
{
if (this->normalisation_sptr->set_up(proj_data_sptr->get_exam_info_sptr(), this->sens_proj_data_info_sptr)
== Succeeded::no)
error("Set_up of norm with non-TOF data failed.\n"
"If your norm is TOF, set \"use time-of-flight sensitivities\" to true.");
}
}
this->norm_already_setup = true;
this->latest_setup_norm_was_with_orig_data = for_original_data;
}
template <typename TargetT>
void
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::ensure_norm_is_set_up_for_sensitivity() const
{
this->ensure_norm_is_set_up(false);
}
template <typename TargetT>
Succeeded
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::set_up_before_sensitivity(
shared_ptr<const TargetT> const& target_sptr)
{
if (is_null_ptr(this->proj_data_sptr))
error("you need to set the input data before calling set_up");
if (this->max_segment_num_to_process == -1)
this->max_segment_num_to_process = this->proj_data_sptr->get_max_segment_num();
if (this->max_segment_num_to_process > this->proj_data_sptr->get_max_segment_num())
{
error("max_segment_num_to_process (%d) is too large", this->max_segment_num_to_process);
return Succeeded::no;
}
this->max_timing_pos_num_to_process = this->proj_data_sptr->get_max_tof_pos_num();
shared_ptr<ProjDataInfo> proj_data_info_sptr(this->proj_data_sptr->get_proj_data_info_sptr()->clone());
#if 0
// KT 4/3/2017 disabled this. It isn't necessary and resolves modyfing the projectors in unexpected ways.
proj_data_info_sptr->
reduce_segment_range(-this->max_segment_num_to_process,
+this->max_segment_num_to_process);
#endif
if (is_null_ptr(this->projector_pair_ptr))
{
error("You need to specify a projector pair");
return Succeeded::no;
}
#ifdef STIR_MPI
// set up distributed caching object
if (distributed_cache_enabled)
{
this->caching_info_ptr = new DistributedCachingInformation(distributed::num_processors);
}
else
caching_info_ptr = NULL;
#else
// non parallel version
caching_info_ptr = NULL;
#endif
this->projector_pair_ptr->set_up(proj_data_info_sptr, target_sptr);
// TODO check compatibility between symmetries for forward and backprojector
this->symmetries_sptr.reset(this->projector_pair_ptr->get_back_projector_sptr()->get_symmetries_used()->clone());
// we postpone calling setup_distributable_computation until we know which projectors we will use
this->distributable_computation_already_setup = false;
// similar for norm
this->norm_already_setup = false;
if (this->get_recompute_sensitivity())
{
if (is_null_ptr(this->normalisation_sptr))
{
error("Invalid normalisation object");
return Succeeded::no;
}
if (!this->use_tofsens && proj_data_info_sptr->is_tof_data() && normalisation_sptr->is_TOF_only_norm())
{
info("Detected TOF normalisation data, so using time-of-flight sensitivities");
this->use_tofsens = true;
}
if (this->sensitivity_uses_same_projector())
{
this->sens_backprojector_sptr = projector_pair_ptr->get_back_projector_sptr();
this->sens_symmetries_sptr = this->symmetries_sptr;
this->sens_proj_data_info_sptr = proj_data_info_sptr;
}
else
{
// sets non-tof backprojector for sensitivity calculation (clone of the back_projector + set projdatainfo to non-tof)
auto pdi_non_tof_sptr = proj_data_info_sptr->create_non_tof_clone();
this->sens_proj_data_info_sptr = pdi_non_tof_sptr;
this->sens_backprojector_sptr.reset(projector_pair_ptr->get_back_projector_sptr()->clone());
if (auto sens_bp_pm_sptr
= std::dynamic_pointer_cast<BackProjectorByBinUsingProjMatrixByBin>(this->sens_backprojector_sptr))
{
// There is no point caching the projection matrix as we will use it only once
// Furthermore, disabling the cache will mean less memory used
// (and we don't have to release it)
sens_bp_pm_sptr->get_proj_matrix_sptr()->enable_cache(false);
}
this->sens_backprojector_sptr->set_up(pdi_non_tof_sptr, target_sptr);
this->sens_symmetries_sptr.reset(this->sens_backprojector_sptr->get_symmetries_used()->clone());
}
}
if (frame_num <= 0)
{
error("frame_num should be >= 1");
return Succeeded::no;
}
if (static_cast<unsigned>(frame_num) > frame_defs.get_num_frames())
{
error("frame_num is %d, but should be less than the number of frames %d.", frame_num, frame_defs.get_num_frames());
return Succeeded::no;
}
return Succeeded::yes;
}
/***************************************************************
functions that compute the value/gradient of the objective function etc
***************************************************************/
template <typename TargetT>
void
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::actual_compute_subset_gradient_without_penalty(
TargetT& gradient, const TargetT& current_estimate, const int subset_num, const bool add_sensitivity)
{
assert(subset_num >= 0);
assert(subset_num < this->num_subsets);
if (!this->distributable_computation_already_setup || !this->latest_setup_distributable_computation_was_with_orig_projectors)
{
// set TOF projectors to be used for the calculations
setup_distributable_computation(this->projector_pair_ptr,
this->proj_data_sptr->get_exam_info_sptr(),
this->proj_data_sptr->get_proj_data_info_sptr(),
std::shared_ptr<TargetT>(gradient.clone()),
zero_seg0_end_planes,
distributed_cache_enabled);
this->distributable_computation_already_setup = true;
this->latest_setup_distributable_computation_was_with_orig_projectors = true;
}
if (!this->distributable_computation_already_setup)
error("PoissonLogLikelihoodWithLinearModelForMeanAndProjData internal error: setup_distributable_computation not called "
"(gradient calculation)");
if (!add_sensitivity)
this->ensure_norm_is_set_up();
distributable_compute_gradient(this->projector_pair_ptr->get_forward_projector_sptr(),
this->projector_pair_ptr->get_back_projector_sptr(),
this->symmetries_sptr,
gradient,
current_estimate,
this->proj_data_sptr,
subset_num,
this->num_subsets,
-this->max_segment_num_to_process,
this->max_segment_num_to_process,
this->zero_seg0_end_planes != 0,
NULL,
this->additive_proj_data_sptr,
this->normalisation_sptr,
caching_info_ptr,
-this->max_timing_pos_num_to_process,
this->max_timing_pos_num_to_process,
add_sensitivity);
}
template <typename TargetT>
double
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::actual_compute_objective_function_without_penalty(
const TargetT& current_estimate, const int subset_num)
{
if (this->distributable_computation_already_setup || !this->latest_setup_distributable_computation_was_with_orig_projectors)
{
// set TOF projectors to be used for the calculations
setup_distributable_computation(this->projector_pair_ptr,
this->proj_data_sptr->get_exam_info_sptr(),
this->proj_data_sptr->get_proj_data_info_sptr(),
std::shared_ptr<TargetT>(current_estimate.clone()),
zero_seg0_end_planes,
distributed_cache_enabled);
this->distributable_computation_already_setup = true;
this->latest_setup_distributable_computation_was_with_orig_projectors = true;
}
if (!this->distributable_computation_already_setup)
error("PoissonLogLikelihoodWithLinearModelForMeanAndProjData internal error: setup_distributable_computation not called "
"(function calculation)");
this->ensure_norm_is_set_up();
double accum = 0.;
distributable_accumulate_loglikelihood(this->projector_pair_ptr->get_forward_projector_sptr(),
this->projector_pair_ptr->get_back_projector_sptr(),
this->symmetries_sptr,
current_estimate,
this->proj_data_sptr,
subset_num,
this->get_num_subsets(),
-this->max_segment_num_to_process,
this->max_segment_num_to_process,
this->zero_seg0_end_planes != 0,
&accum,
this->additive_proj_data_sptr,
this->normalisation_sptr,
this->get_time_frame_definitions().get_start_time(this->get_time_frame_num()),
this->get_time_frame_definitions().get_end_time(this->get_time_frame_num()),
this->caching_info_ptr,
-this->max_timing_pos_num_to_process,
this->max_timing_pos_num_to_process);
return accum;
}
#if 0
template<typename TargetT>
float
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::
sum_projection_data() const
{
float counts=0.0F;
for (int segment_num = -max_segment_num_to_process; segment_num <= max_segment_num_to_process; ++segment_num)
{
for (int timing_pos_num = -max_timing_pos_num_to_process; timing_pos_num <= max_timing_pos_num_to_process; ++timing_pos_num)
{
for (int view_num = proj_data_sptr->get_min_view_num();
view_num <= proj_data_sptr->get_max_view_num();
++view_num)
{
Viewgram<float> viewgram=proj_data_sptr->get_viewgram(view_num,segment_num,false,timing_pos_num);
//first adjust data
// KT 05/07/2000 made parameters.zero_seg0_end_planes int
if(segment_num==0 && zero_seg0_end_planes!=0)
{
viewgram[viewgram.get_min_axial_pos_num()].fill(0);
viewgram[viewgram.get_max_axial_pos_num()].fill(0);
}
truncate_rim(viewgram,rim_truncation_sino);
//now take totals
counts+=viewgram.sum();
}
}
}
return counts;
}
#endif
template <typename TargetT>
void
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::add_subset_sensitivity(TargetT& sensitivity,
const int subset_num) const
{
const int min_segment_num = -this->max_segment_num_to_process;
const int max_segment_num = this->max_segment_num_to_process;
shared_ptr<TargetT> sensitivity_this_subset_sptr(sensitivity.clone());
// have to create a ProjData object filled with 1 here because otherwise zero_seg0_endplanes will not be effective
auto sens_proj_data_sptr
= std::make_shared<ProjDataInMemory>(this->proj_data_sptr->get_exam_info_sptr(), this->sens_proj_data_info_sptr);
sens_proj_data_sptr->fill(1.0F);
if (this->sensitivity_uses_same_projector()
&& (!this->distributable_computation_already_setup
|| !this->latest_setup_distributable_computation_was_with_orig_projectors))
{
// set TOF projectors to be used for the calculations
setup_distributable_computation(this->projector_pair_ptr,
this->proj_data_sptr->get_exam_info_sptr(),
sens_proj_data_sptr->get_proj_data_info_sptr(),
std::shared_ptr<TargetT>(sensitivity.clone()),
zero_seg0_end_planes,
distributed_cache_enabled);
this->distributable_computation_already_setup = true;
this->latest_setup_distributable_computation_was_with_orig_projectors = true;
}
else if (!this->sensitivity_uses_same_projector()
&& (!this->distributable_computation_already_setup
|| this->latest_setup_distributable_computation_was_with_orig_projectors))
{
// set non-TOF projector to be used for the calculations
shared_ptr<ForwardProjectorByBin> dummy_sptr;
auto sens_projector_pair_sptr
= std::make_shared<ProjectorByBinPairUsingSeparateProjectors>(dummy_sptr, this->sens_backprojector_sptr);
setup_distributable_computation(sens_projector_pair_sptr,
this->proj_data_sptr->get_exam_info_sptr(),
sens_proj_data_sptr->get_proj_data_info_sptr(),
std::shared_ptr<TargetT>(sensitivity.clone()),
zero_seg0_end_planes,
distributed_cache_enabled);
this->distributable_computation_already_setup = true;
this->latest_setup_distributable_computation_was_with_orig_projectors = false;
}
if (!this->distributable_computation_already_setup)
error("PoissonLogLikelihoodWithLinearModelForMeanAndProjData internal error: setup_distributable_computation not called "
"(sensitivity calculation)");
this->ensure_norm_is_set_up_for_sensitivity();
distributable_sensitivity_computation(this->sens_backprojector_sptr,
this->sens_symmetries_sptr,
*sensitivity_this_subset_sptr,
sensitivity,
sens_proj_data_sptr,
subset_num,
this->num_subsets,
min_segment_num,
max_segment_num,
this->zero_seg0_end_planes != 0,
NULL,
this->additive_proj_data_sptr,
this->normalisation_sptr,
this->get_time_frame_definitions().get_start_time(this->get_time_frame_num()),
this->get_time_frame_definitions().get_end_time(this->get_time_frame_num()),
this->caching_info_ptr,
use_tofsens ? -this->max_timing_pos_num_to_process : 0,
use_tofsens ? this->max_timing_pos_num_to_process : 0);
std::transform(sensitivity.begin_all(),
sensitivity.end_all(),
sensitivity_this_subset_sptr->begin_all(),
sensitivity.begin_all(),
std::plus<typename TargetT::full_value_type>());
}
template <typename TargetT>
std::unique_ptr<ExamInfo>
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<TargetT>::get_exam_info_uptr_for_target() const
{
auto exam_info_uptr = this->get_exam_info_uptr_for_target();
if (auto norm_ptr = dynamic_cast<BinNormalisationWithCalibration const* const>(get_normalisation_sptr().get()))
{
exam_info_uptr->set_calibration_factor(norm_ptr->get_calibration_factor());
// somehow tell the image that it's calibrated
}
else
{
exam_info_uptr->set_calibration_factor(-1.F);
// somehow tell the image that it's not calibrated
}
return exam_info_uptr;
}
static std::vector<ViewgramIndices>
find_basic_viewgram_indices_in_subset(const ProjDataInfo& proj_data_info,
const DataSymmetriesForViewSegmentNumbers& symmetries,
const int min_segment_num,
const int max_segment_num,
const int subset_num,
const int num_subsets)
{
const std::vector<ViewSegmentNumbers> vs_nums_to_process = detail::find_basic_vs_nums_in_subset(
proj_data_info, symmetries, min_segment_num, max_segment_num, subset_num, num_subsets);
std::vector<ViewgramIndices> vg_idx_to_process;
for (auto vs_num : vs_nums_to_process)
{
for (int k = proj_data_info.get_min_tof_pos_num(); k <= proj_data_info.get_max_tof_pos_num(); ++k)
{
ViewgramIndices viewgram_idx = vs_num;
viewgram_idx.timing_pos_num() = k;
vg_idx_to_process.push_back(viewgram_idx);
}
}
return vg_idx_to_process;
}
template <typename TargetT>
Succeeded
PoissonLogLikelihoodWithLinearModelForMeanAndProjData<
TargetT>::actual_add_multiplication_with_approximate_sub_Hessian_without_penalty(TargetT& output,
const TargetT& input,
const int subset_num) const
{
{
std::string explanation;
if (!input.has_same_characteristics(this->get_sensitivity(), explanation))
{
error("PoissonLogLikelihoodWithLinearModelForMeanAndProjData:\n"
"sensitivity and input for add_multiplication_with_approximate_Hessian_without_penalty\n"
"should have the same characteristics.\n%s",
explanation.c_str());
return Succeeded::no;
}
}
this->ensure_norm_is_set_up();
shared_ptr<DataSymmetriesForViewSegmentNumbers> symmetries_sptr(this->get_projector_pair().get_symmetries_used()->clone());
this->get_projector_pair().get_forward_projector_sptr()->set_input(input);
this->get_projector_pair().get_back_projector_sptr()->start_accumulating_in_new_target();
const std::vector<ViewgramIndices> vg_idx_to_process
= find_basic_viewgram_indices_in_subset(*this->get_proj_data().get_proj_data_info_sptr(),
*symmetries_sptr,
-this->get_max_segment_num_to_process(),
this->get_max_segment_num_to_process(),
subset_num,
this->get_num_subsets());
info("Forward projecting input image.", 2);
#ifdef STIR_OPENMP
# pragma omp parallel for schedule(dynamic)
#endif
// note: older versions of openmp need an int as loop
for (int i = 0; i < static_cast<int>(vg_idx_to_process.size()); ++i)
{
const auto viewgram_idx = vg_idx_to_process[i];
{
#ifdef STIR_OPENMP
const int thread_num = omp_get_thread_num();
const int num_threads = omp_get_num_threads();
#else
const int thread_num = 0;
const int num_threads = 1;
#endif
info(boost::format("Thread %d/%d calculating segment_num: %d, view_num: %d, TOF: %d") % thread_num % num_threads
% viewgram_idx.segment_num() % viewgram_idx.view_num() % viewgram_idx.timing_pos_num(),
2);
}
// first compute data-term: y*norm^2
RelatedViewgrams<float> viewgrams = this->get_proj_data().get_related_viewgrams(viewgram_idx, symmetries_sptr);