forked from YujieChi/gPET
-
Notifications
You must be signed in to change notification settings - Fork 0
/
detector.cu
executable file
·410 lines (347 loc) · 13 KB
/
detector.cu
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
#ifndef __DETECT__CU__
#define __DETECT__CU__
#include <algorithm>
#include "gPET.h"
#include "externCUDA.h"
#include <thrust/device_ptr.h>
#include <thrust/sort.h>
compare_parn compare1;
compare_siten compare2;
compare_t compare3;
using namespace std;
struct object_t InitializeObject()
{
struct object_t q;
q.panel = 0;
q.lengthx=0.0f;
q.lengthy=0.0f;
q.lengthz=0.0f;
q.MODx=0.0f;
q.MODy=0.0f;
q.MODz=0.0f;
q.Mspacex=0.0f;
q.Mspacey=0.0f;
q.Mspacez=0.0f;
q.LSOx=0.0f;
q.LSOy=0.0f;
q.LSOz=0.0f;
q.spacex=0.0f;
q.spacey=0.0f;
q.spacez=0.0f;
q.offsetx=0.0f;
q.offsety=0.0f;
q.offsetz=0.0f;
q.directionx=0.0f;
q.directiony=0.0f;
q.directionz=0.0f;
q.UniXx=0.0f;
q.UniXy=0.0f;
q.UniXz=0.0f;
q.UniYx=0.0f;
q.UniYy=0.0f;
q.UniYz=0.0f;
q.UniZx=0.0f;
q.UniZy=0.0f;
q.UniZz=0.0f;
return q;
}
void read_file_ro(struct object_t** objectArray, struct object_v** objectMaterial, int* total_Panels, char fname[100])
/********************************************************************************
c* read geometry files using rotational definition *
c* Input: *
c* fname: input geometry file *
c* Output: *
objectArray: buildup geometry *
c* total_Panels: total panel numbers *
/*******************************************************************************/
{
printf("loading PET detector geometry parameters ... %s\n",fname);
FILE* fp=fopen(fname,"r");
char buffer[256];
int count = 0;
fgets(buffer, 256, fp);
fscanf(fp, "%d \n", &count);
*total_Panels = count;
cout << "total panels "<<*total_Panels << endl;
float rot[3];
fgets(buffer, 256, fp);
fscanf(fp, "%f %f %f\n", &rot[0], &rot[1], &rot[2]);
cout << "panel rotational axis "<<rot[0] <<" "<<rot[1] <<" "<<rot[2]<< endl;
float rotAng;
fgets(buffer, 256, fp);
fscanf(fp, "%f\n", &rotAng);
cout << "panel rotational angle "<<rotAng<< endl;
// read the file for the second time, to load all the parameters
struct object_t* temp;
temp = (object_t*)malloc(*total_Panels*sizeof(object_t));
struct object_v* temp1;
temp1 = (object_v*)malloc(2*sizeof(object_v));
for (int i = 0; i < *total_Panels; i++)
{
temp[i] = InitializeObject();
}
int mat = 0, pane = 0;
float den = 0.0f, lenx=0.0f, leny=0.0f, lenz=0.0f, Mx=0.0f,My=0.0f,Mz=0.0f, Msx=0.0f,Msy=0.0f,Msz=0.0f;
float Lx=0.0f,Ly=0.0f,Lz=0.0f,sx=0.0f, sy=0.0f, sz=0.0f, ox=0.0f, oy=0.0f, oz=0.0f, dx=0.0f,dy=0.0f,dz=0.0f;
float UXx=0.0f, UXy=0.0f, UXz=0.0f, UYx=0.0f, UYy=0.0f, UYz=0.0f, UZx=0.0f, UZy=0.0f, UZz=0.0f;
// only two materials (0:LSO, 1: air)
fgets(buffer, 256, fp);
for (int i = 0; i < 2; i++)
{
fscanf(fp, "%d %f \n", &mat, &den);
temp1[i].material=mat;
temp1[i].density=den;
//printf("mat=%d, den=%f\n", mat, den);
}
fgets(buffer, 256, fp);
// read in parameter for each Panel
for (int i = 0; i < 1; i++)
{
//printf("i=%d\n", i);
fgets(buffer, 256, fp);
fscanf(fp, "%d \n", &pane);
temp[i].panel = pane;
//printf("starting panel=%d\n", pane);
fgets(buffer, 256, fp);
fscanf(fp, "%f %f %f\n", &lenx, &leny, &lenz);
temp[i].lengthx = lenx;
temp[i].lengthy = leny;
temp[i].lengthz = lenz;
//printf("lengthx=%f, lengthy=%f, lengthz=%f\n", lenx, leny, lenz);
fgets(buffer, 256, fp);
fscanf(fp, "%f %f %f\n", &Mx, &My, &Mz);
temp[i].MODx = Mx;
temp[i].MODy = My;
temp[i].MODz = Mz;
//printf("MODx=%f, MODy=%f, MODz=%f\n", Mx, My, Mz);
fgets(buffer, 256, fp);
fscanf(fp, "%f %f %f\n", &Msx, &Msy, &Msz);
temp[i].Mspacex = Msx;
temp[i].Mspacey = Msy;
temp[i].Mspacez = Msz;
//printf("Mspacex=%f, Mspacey=%f, Mspacez=%f\n", Msx, Msy, Msz);
fgets(buffer, 256, fp);
fscanf(fp, "%f %f %f\n", &Lx, &Ly, &Lz);
temp[i].LSOx = Lx;
temp[i].LSOy = Ly;
temp[i].LSOz = Lz;
//printf("LSOx=%f, LSOy=%f, LSOz=%f\n", Lx, Ly, Lz);
fgets(buffer, 256, fp);
fscanf(fp, "%f %f %f\n", &sx, &sy, &sz);
temp[i].spacex = sx;
temp[i].spacey = sy;
temp[i].spacez = sz;
//printf("spacex=%f, spacey=%f, spacez=%f\n", sx, sy, sz);
fgets(buffer, 256, fp);
fscanf(fp, "%f %f %f\n", &dx, &dy, &dz);
temp[i].directionx = dx;
temp[i].directiony = dy;
temp[i].directionz = dz;
//printf("directionx=%f, directiony=%f, directionz=%f\n", dx, dy, dz);
fgets(buffer, 256, fp);
fscanf(fp, "%f %f %f\n", &ox, &oy, &oz);
temp[i].offsetx = ox;
temp[i].offsety = oy;
temp[i].offsetz = oz;
//printf("offsetx=%f, offsety=%f, offsetz=%f\n", ox, oy, oz);
fgets(buffer, 256, fp);
fscanf(fp, "%f %f %f\n", &UXx, &UXy, &UXz);
temp[i].UniXx = UXx;
temp[i].UniXy = UXy;
temp[i].UniXz = UXz;
//printf("UniXx=%f, UniXy=%f, UniXz=%f\n", UXx, UXy, UXz);
fgets(buffer, 256, fp);
fscanf(fp, "%f %f %f\n", &UYx, &UYy, &UYz);
temp[i].UniYx = UYx;
temp[i].UniYy = UYy;
temp[i].UniYz = UYz;
//printf("UniYx=%f, UniYy=%f, UniYz=%f\n", UYx, UYy, UYz);
fgets(buffer, 256, fp);
fscanf(fp, "%f %f %f\n", &UZx, &UZy, &UZz);
temp[i].UniZx = UZx;
temp[i].UniZy = UZy;
temp[i].UniZz = UZz;
//printf("UniZx=%f, UniZy=%f, UniZz=%f\n", UZx, UZy, UZz);
fgets(buffer, 256, fp);
}
fclose(fp);
float vec[3];
float ang;
for (int i = 1; i < *total_Panels; i++)
{
if (*total_Panels<=1)
break;
//printf("i=%d\n", i);
temp[i].panel = i;
//printf("panel=%d\n", temp[i].panel);
temp[i].lengthx = temp[0].lengthx;
temp[i].lengthy = temp[0].lengthy;
temp[i].lengthz = temp[0].lengthz;
//printf("lengthx=%f, lengthy=%f, lengthz=%f\n", temp[i].lengthx, temp[i].lengthy, temp[i].lengthz);
temp[i].MODx = temp[0].MODx;
temp[i].MODy = temp[0].MODy;
temp[i].MODz = temp[0].MODz;
temp[i].Mspacex = temp[0].Mspacex;
temp[i].Mspacey = temp[0].Mspacey;
temp[i].Mspacez = temp[0].Mspacez;
temp[i].LSOx = temp[0].LSOx;
temp[i].LSOy = temp[0].LSOy;
temp[i].LSOz = temp[0].LSOz;
temp[i].spacex = temp[0].spacex;
temp[i].spacey = temp[0].spacey;
temp[i].spacez = temp[0].spacez;
temp[i].directionx = temp[0].directionx;
temp[i].directiony = temp[0].directiony;
temp[i].directionz = temp[0].directionz;
ang=rotAng*PI/180.0f*i;
vec[0] = temp[0].offsetx;
vec[1] = temp[0].offsety;
vec[2] = temp[0].offsetz;
temp[i].offsetx=(1-cosf(ang))*(vec[0]*rot[0])*rot[0]+cosf(ang)*vec[0]+sinf(ang)*(rot[1]*vec[2]-rot[2]*vec[1]);
temp[i].offsety=(1-cosf(ang))*(vec[1]*rot[1])*rot[1]+cosf(ang)*vec[1]+sinf(ang)*(rot[2]*vec[0]-rot[0]*vec[2]);
temp[i].offsetz=(1-cosf(ang))*(vec[2]*rot[2])*rot[2]+cosf(ang)*vec[2]+sinf(ang)*(rot[0]*vec[1]-rot[1]*vec[0]);
//printf("offsetx=%f, offsety=%f, offsetz=%f\n", temp[i].offsetx, temp[i].offsety, temp[i].offsetz);
vec[0] = temp[0].UniXx;
vec[1] = temp[0].UniXy;
vec[2] = temp[0].UniXz;
temp[i].UniXx = (1-cosf(ang))*(vec[0]*rot[0])*rot[0]+cosf(ang)*vec[0]+sinf(ang)*(rot[1]*vec[2]-rot[2]*vec[1]);
temp[i].UniXy = (1-cosf(ang))*(vec[1]*rot[1])*rot[1]+cosf(ang)*vec[1]+sinf(ang)*(rot[2]*vec[0]-rot[0]*vec[2]);
temp[i].UniXz = (1-cosf(ang))*(vec[2]*rot[2])*rot[2]+cosf(ang)*vec[2]+sinf(ang)*(rot[0]*vec[1]-rot[1]*vec[0]);
//printf("UniXx=%f, UniXy=%f, UniXz=%f\n", temp[i].UniXx, temp[i].UniXy, temp[i].UniXz);
vec[0] = temp[0].UniYx;
vec[1] = temp[0].UniYy;
vec[2] = temp[0].UniYz;
temp[i].UniYx = (1-cosf(ang))*(vec[0]*rot[0])*rot[0]+cosf(ang)*vec[0]+sinf(ang)*(rot[1]*vec[2]-rot[2]*vec[1]);
temp[i].UniYy = (1-cosf(ang))*(vec[1]*rot[1])*rot[1]+cosf(ang)*vec[1]+sinf(ang)*(rot[2]*vec[0]-rot[0]*vec[2]);
temp[i].UniYz = (1-cosf(ang))*(vec[2]*rot[2])*rot[2]+cosf(ang)*vec[2]+sinf(ang)*(rot[0]*vec[1]-rot[1]*vec[0]);
//printf("UniYx=%f, UniYy=%f, UniYz=%f\n", temp[i].UniYx, temp[i].UniYy, temp[i].UniYz);
vec[0] = temp[0].UniZx;
vec[1] = temp[0].UniZy;
vec[2] = temp[0].UniZz;
temp[i].UniZx = (1-cosf(ang))*(vec[0]*rot[0])*rot[0]+cosf(ang)*vec[0]+sinf(ang)*(rot[1]*vec[2]-rot[2]*vec[1]);
temp[i].UniZy = (1-cosf(ang))*(vec[1]*rot[1])*rot[1]+cosf(ang)*vec[1]+sinf(ang)*(rot[2]*vec[0]-rot[0]*vec[2]);
temp[i].UniZz = (1-cosf(ang))*(vec[2]*rot[2])*rot[2]+cosf(ang)*vec[2]+sinf(ang)*(rot[0]*vec[1]-rot[1]*vec[0]);
//printf("UniZx=%f, UniZy=%f, UniZz=%f\n", temp[i].UniZx, temp[i].UniZy, temp[i].UniZz);
}
*objectArray = temp;
*objectMaterial = temp1;
//printf("\n");
printf("\n");
}
int outevents(int* num_d, Event* totalevents_d, const char *outputfilename)
{
//copy data from device to host
//renewed at 1025, do not have to be the memory on GPU
int num;
if(cudaMemcpy(&num, num_d, sizeof(int), cudaMemcpyDeviceToHost)!=cudaSuccess)
num=*num_d;
cout<<"num is "<<num<<endl;
if(num <1) return 0;
Event* tempData_h =(struct Event*) malloc( sizeof(Event)*num);
if(cudaMemcpy(tempData_h, totalevents_d, sizeof(Event)*num, cudaMemcpyDeviceToHost)!=cudaSuccess)
memcpy(tempData_h, totalevents_d, sizeof(Event)*num);
// write results to file
ofstream out(outputfilename,ios::app|ios::binary);
out.write((char*) tempData_h,sizeof(Event)*num);
out.close();
cout<<"data has been written to "<<outputfilename<<"\n";
// free space
free(tempData_h);
return 1;
}
void quicksort(Event* hits,int start, int stop, int sorttype)
{
//CPU sort function for ordering events in cpu memory
//sorttype 1 for ordering by particle #,2 for site number, 3 for flight time
switch(sorttype)
{
case 1:
{ sort(hits+start,hits+stop,compare1);
break;
}
case 2:
{ sort(hits+start,hits+stop,compare2);
break;
}
case 3:
{ sort(hits+start,hits+stop,compare3);
break;
}
}
}
void quicksort_d(Event* events_d,int start, int stop, int sorttype)
{
//GPU version for ordering the events in gpu memory,
//more suitable for large scale sorting
//sorttype 1 for ordering by particle #,2 for site number, 3 for flight time
printf("gpu sort starts!!\n");
thrust::device_ptr<Event> hits=thrust::device_pointer_cast(events_d);
switch(sorttype)
{
case 1:
{ thrust::sort(hits+start,hits+stop,compare1);
break;
}
case 2:
{ thrust::sort(hits+start,hits+stop,compare2);
break;
}
case 3:
{ thrust::sort(hits+start,hits+stop,compare3);
break;
}
}//*/
printf("gpu sort finishs!!\n");//*/
}
void quicksort_h(Event* events_d,int start, int stop, int sorttype)
{
//CPU version for ordering the events in gpu memory,
//more suitable for large scale sorting
//sorttype 1 for ordering by particle #,2 for site number, 3 for flight time
printf("cpu sort starts!!\n");
int counts = stop-start;
Event* events=(Event*) malloc(sizeof(Event)*counts);
cudaMemcpy(events,events_d+start,sizeof(Event)*counts,cudaMemcpyDeviceToHost);
quicksort(events,0,counts,sorttype);
cudaMemcpy(events_d+start,events,sizeof(Event)*counts,cudaMemcpyHostToDevice);
free(events);
printf("cpu sort finishs!!\n");//*/
}
void orderevents(int* counts,Event* events_d)
{
Event* events=(Event*) malloc(sizeof(Event)*counts[0]);
cudaMemcpy(events,events_d,sizeof(Event)*counts[0],cudaMemcpyDeviceToHost);
quicksort(events,0,counts[0],2);
int start=0;
for(int i=1; i<counts[0];)
{
while(events[i].siten==events[start].siten&&(i<counts[0]))
i++;
if(i>start+1) quicksort(events,start,i,3);
start=i;
i++;
}
cudaMemcpy(events_d,events,sizeof(Event)*counts[0],cudaMemcpyHostToDevice);
free(events);//*/
}
void outputData(void *src, const int size, const char *outputfilename, const char *mode)
// output data to file
{
// copy data from GPU to CPU
void *tempData_h = malloc( size );
if(cudaMemcpy( tempData_h, src, size, cudaMemcpyDeviceToHost)!= cudaSuccess)
memcpy(tempData_h, src, size);
// write results to file
FILE *fp;
fp = fopen(outputfilename, mode);
if( fp == NULL )
{
cout << "Can not open file to write results.";
exit(1);
}
fwrite (tempData_h, size, 1 , fp );
fclose(fp);
// free space
free(tempData_h);
}
#endif