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
|
include "../lib/obsfile.h"
define LOGPTR 32 # log2(maxpts) (4e9)
# PH_1C4R2ISORT -- Vector quicksort on the first and second indices arrays,
# where the second index is used to resolve ambiguities in the first index.
# An additional 4 input arrays are sorted as well.
procedure ph_1c4r2isort (label, d1, d2, d3, d4, findex, sindex, npix)
char label[DEF_LENLABEL,ARB] # the char array
real d1[ARB] # the first input data array
real d2[ARB] # the second input data array
real d3[ARB] # the third input data array
real d4[ARB] # the fourth input data array
int findex[ARB] # first index array which is sorted on
int sindex[ARB] # second index array which is sorted on
int npix # number of pixels
real tempr
int fpivot, spivot, tempi
int i, j, k, p, lv[LOGPTR], uv[LOGPTR]
char tempc[DEF_LENLABEL]
int ph_2icompare()
define swapi {tempi=$1;$1=$2;$2=tempi}
define swapr {tempr=$1;$1=$2;$2=tempr}
define swapc {call strcpy ($1, tempc, DEF_LENLABEL);call strcpy ($2, $1, DEF_LENLABEL);call strcpy (tempc, $2, DEF_LENLABEL)}
begin
lv[1] = 1
uv[1] = npix
p = 1
while (p > 0) {
if (lv[p] >= uv[p]) # only one elem in this subset
p = p - 1 # pop stack
else {
# Dummy do loop to trigger the Fortran optimizer.
do p = p, ARB {
i = lv[p] - 1
j = uv[p]
# Select as the pivot the element at the center of the
# array, to avoid quadratic behavior on an already sorted
# array.
k = (lv[p] + uv[p]) / 2
swapr (d1[j], d1[k])
swapr (d2[j], d2[k])
swapr (d3[j], d3[k])
swapr (d4[j], d4[k])
swapc (label[1,j], label[1,k])
swapi (findex[j], findex[k])
swapi (sindex[j], sindex[k])
fpivot = findex[j] # pivot line
spivot = sindex[j]
while (i < j) {
for (i=i+1; ph_2icompare (findex[i], sindex[i], fpivot,
spivot) < 0; i=i+1)
;
for (j=j-1; j > i; j=j-1)
if (ph_2icompare (findex[j], sindex[j], fpivot,
spivot) <= 0)
break
if (i < j) { # switch elements
swapr (d1[i], d1[j])
swapr (d2[i], d2[j])
swapr (d3[i], d3[j])
swapr (d4[i], d4[j])
swapc (label[1,i], label[1,j])
swapi (sindex[i], sindex[j])
swapi (findex[i], findex[j]) # interchange elements
}
}
j = uv[p] # move pivot to position i
swapr (d1[i], d1[j])
swapr (d2[i], d2[j])
swapr (d3[i], d3[j])
swapr (d4[i], d4[j])
swapc (label[1,i], label[1,j])
swapi (sindex[i], sindex[j])
swapi (findex[i], findex[j]) # interchange elements
if (i-lv[p] < uv[p] - i) { # stack so shorter done first
lv[p+1] = lv[p]
uv[p+1] = i - 1
lv[p] = i + 1
} else {
lv[p+1] = i + 1
uv[p+1] = uv[p]
uv[p] = i - 1
}
break
}
p = p + 1 # push onto stack
}
}
end
# PH_4R2ISORT -- Vector quicksort on the first and second indices arrays,
# where the second index is used to resolve ambiguities in the first index.
# An additional 4 input arrays are sorted as well.
procedure ph_4r2isort (d1, d2, d3, d4, findex, sindex, npix)
real d1[ARB] # the first input data array
real d2[ARB] # the second input data array
real d3[ARB] # the third input data array
real d4[ARB] # the fourth input data array
int findex[ARB] # first index array which is sorted on
int sindex[ARB] # second index array which is sorted on
int npix # number of pixels
int fpivot, spivot, tempi
int i, j, k, p, lv[LOGPTR], uv[LOGPTR]
real tempr
int ph_2icompare()
define swapi {tempi=$1;$1=$2;$2=tempi}
define swapr {tempr=$1;$1=$2;$2=tempr}
begin
lv[1] = 1
uv[1] = npix
p = 1
while (p > 0) {
if (lv[p] >= uv[p]) # only one elem in this subset
p = p - 1 # pop stack
else {
# Dummy do loop to trigger the Fortran optimizer.
do p = p, ARB {
i = lv[p] - 1
j = uv[p]
# Select as the pivot the element at the center of the
# array, to avoid quadratic behavior on an already sorted
# array.
k = (lv[p] + uv[p]) / 2
swapr (d1[j], d1[k])
swapr (d2[j], d2[k])
swapr (d3[j], d3[k])
swapr (d4[j], d4[k])
swapi (findex[j], findex[k])
swapi (sindex[j], sindex[k])
fpivot = findex[j] # pivot line
spivot = sindex[j]
while (i < j) {
for (i=i+1; ph_2icompare (findex[i], sindex[i], fpivot,
spivot) < 0; i=i+1)
;
for (j=j-1; j > i; j=j-1)
if (ph_2icompare (findex[j], sindex[j], fpivot,
spivot) <= 0)
break
if (i < j) { # switch elements
swapr (d1[i], d1[j])
swapr (d2[i], d2[j])
swapr (d3[i], d3[j])
swapr (d4[i], d4[j])
swapi (sindex[i], sindex[j])
swapi (findex[i], findex[j]) # interchange elements
}
}
j = uv[p] # move pivot to position i
swapr (d1[i], d1[j])
swapr (d2[i], d2[j])
swapr (d3[i], d3[j])
swapr (d4[i], d4[j])
swapi (sindex[i], sindex[j])
swapi (findex[i], findex[j]) # interchange elements
if (i-lv[p] < uv[p] - i) { # stack so shorter done first
lv[p+1] = lv[p]
uv[p+1] = i - 1
lv[p] = i + 1
} else {
lv[p+1] = i + 1
uv[p+1] = uv[p]
uv[p] = i - 1
}
break
}
p = p + 1 # push onto stack
}
}
end
# PH_3RIRSORT -- Vector quicksort on the index and the rindex array,
# where the rindex array is used to resolve ambiguities in the index array.
# The three real arrays are sorted as well.
procedure ph_3rirsort (d1, d2, d3, index, rindex, npix)
real d1[ARB] # the first input data array
real d2[ARB] # the second input data array
real d3[ARB] # the third input data array
int index[ARB] # first index array which is sorted on
real rindex[ARB] # second index array which is sorted on
int npix # number of pixels
int fpivot, tempi
int i, j, k, p, lv[LOGPTR], uv[LOGPTR]
real rpivot, tempr
int ph_ircompare()
define swapi {tempi=$1;$1=$2;$2=tempi}
define swapr {tempr=$1;$1=$2;$2=tempr}
begin
lv[1] = 1
uv[1] = npix
p = 1
while (p > 0) {
if (lv[p] >= uv[p]) # only one elem in this subset
p = p - 1 # pop stack
else {
# Dummy do loop to trigger the Fortran optimizer.
do p = p, ARB {
i = lv[p] - 1
j = uv[p]
# Select as the pivot the element at the center of the
# array, to avoid quadratic behavior on an already sorted
# array.
k = (lv[p] + uv[p]) / 2
swapr (d1[j], d1[k])
swapr (d2[j], d2[k])
swapr (d3[j], d3[k])
swapi (index[j], index[k])
swapr (rindex[j], rindex[k])
fpivot = index[j] # pivot line
rpivot = rindex[j]
while (i < j) {
for (i=i+1; ph_ircompare (index[i], rindex[i], fpivot,
rpivot) < 0; i=i+1)
;
for (j=j-1; j > i; j=j-1)
if (ph_ircompare (index[j], rindex[j], fpivot,
rpivot) <= 0)
break
if (i < j) { # switch elements
swapr (d1[i], d1[j])
swapr (d2[i], d2[j])
swapr (d3[i], d3[j])
swapr (rindex[i], rindex[j])
swapi (index[i], index[j]) # interchange elements
}
}
j = uv[p] # move pivot to position i
swapr (d1[i], d1[j])
swapr (d2[i], d2[j])
swapr (d3[i], d3[j])
swapr (rindex[i], rindex[j])
swapi (index[i], index[j]) # interchange elements
if (i-lv[p] < uv[p] - i) { # stack so shorter done first
lv[p+1] = lv[p]
uv[p+1] = i - 1
lv[p] = i + 1
} else {
lv[p+1] = i + 1
uv[p+1] = uv[p]
uv[p] = i - 1
}
break
}
p = p + 1 # push onto stack
}
}
end
# PH_QSORT -- Vector quicksort. In this version the index array is
# sorted.
procedure ph_qsort (data, index, npix, offset)
real data[ARB] # data array
int index[ARB] # index array
int npix # number of pixels
int offset # the index offset
int i, j, lv[LOGPTR], p, uv[LOGPTR], temp
real pivot
begin
# Initialize the indices for an inplace sort.
do i = 1, npix
index[i] = i
p = 1
lv[1] = 1
uv[1] = npix
while (p > 0) {
# If only one elem in subset pop stack otherwise pivot line.
if (lv[p] >= uv[p])
p = p - 1
else {
i = lv[p] - 1
j = uv[p]
pivot = data[index[j]]
while (i < j) {
for (i=i+1; data[index[i]] < pivot; i=i+1)
;
for (j=j-1; j > i; j=j-1)
if (data[index[j]] <= pivot)
break
if (i < j) { # out of order pair
temp = index[j] # interchange elements
index[j] = index[i]
index[i] = temp
}
}
j = uv[p] # move pivot to position i
temp = index[j] # interchange elements
index[j] = index[i]
index[i] = temp
if (i-lv[p] < uv[p] - i) { # stack so shorter done first
lv[p+1] = lv[p]
uv[p+1] = i - 1
lv[p] = i + 1
} else {
lv[p+1] = i + 1
uv[p+1] = uv[p]
uv[p] = i - 1
}
p = p + 1 # push onto stack
}
}
do i = 1, npix
index[i] = index[i] + offset - 1
end
# PH_2ICOMPARE -- Comparison routine for PH_4R2ISORT.
int procedure ph_2icompare (findex, sindex, fpivot, spivot)
int findex # the first index value
int sindex # the second index value
int fpivot # the first pivot value
int spivot # the second pivot value
begin
if (findex < fpivot)
return (-1)
else if (findex > fpivot)
return (1)
else if (sindex < spivot)
return (-1)
else if (sindex > spivot)
return (1)
else
return (0)
end
# PH_IRCOMPARE -- Comparison routine for PH_3RIRSORT.
int procedure ph_ircompare (findex, sindex, fpivot, spivot)
int findex # the first index value
real sindex # the second index value
int fpivot # the first pivot value
real spivot # the second pivot value
begin
if (findex < fpivot)
return (-1)
else if (findex > fpivot)
return (1)
else if (sindex < spivot)
return (-1)
else if (sindex > spivot)
return (1)
else
return (0)
end
# PH_5R3ISORT -- Vector quicksort on the first and second indices arrays,
# where the second index is used to resolve ambiguities in the first index.
# An additional 5 input arrays are sorted as well.
procedure ph_5r3isort (findex, sindex, i1, d1, d2, d3, d4, d5, naperts, npix)
int findex[ARB] # first index array which is sorted on
int sindex[ARB] # second index array which is sorted on
int i1[ARB] # the 3 integer array
real d1[ARB] # the first input data array
real d2[ARB] # the second input data array
real d3[naperts,ARB] # the third input data array
real d4[naperts,ARB] # the fourth input data array
real d5[naperts,ARB] # the fifth input data array
int naperts # number of apertures
int npix # number of pixels
int fpivot, spivot, tempi
int i, j, k, l,p, lv[LOGPTR], uv[LOGPTR]
real tempr
int ph_2aicompare()
define swapi {tempi=$1;$1=$2;$2=tempi}
define swapr {tempr=$1;$1=$2;$2=tempr}
begin
lv[1] = 1
uv[1] = npix
p = 1
while (p > 0) {
if (lv[p] >= uv[p]) # only one elem in this subset
p = p - 1 # pop stack
else {
# Dummy do loop to trigger the Fortran optimizer.
do p = p, ARB {
i = lv[p] - 1
j = uv[p]
# Select as the pivot the element at the center of the
# array, to avoid quadratic behavior on an already sorted
# array.
k = (lv[p] + uv[p]) / 2
swapi (findex[j], findex[k])
swapi (sindex[j], sindex[k])
swapi (i1[j], i1[k])
swapr (d1[j], d1[k])
swapr (d2[j], d2[k])
do l = 1, naperts {
swapr (d3[l,j], d3[l,k])
swapr (d4[l,j], d4[l,k])
swapr (d5[l,j], d5[l,k])
}
fpivot = findex[j] # pivot line
spivot = sindex[j]
while (i < j) {
for (i=i+1; ph_2aicompare (findex[i], sindex[i], fpivot,
spivot) < 0; i=i+1)
;
for (j=j-1; j > i; j=j-1)
if (ph_2aicompare (findex[j], sindex[j], fpivot,
spivot) <= 0)
break
if (i < j) { # switch elements
swapi (sindex[i], sindex[j])
swapi (findex[i], findex[j]) # interchange elements
swapi (i1[i], i1[j])
swapr (d1[i], d1[j])
swapr (d2[i], d2[j])
do l = 1, naperts {
swapr (d3[l,i], d3[l,j])
swapr (d4[l,i], d5[l,j])
swapr (d5[l,i], d5[l,j])
}
}
}
j = uv[p] # move pivot to position i
swapi (sindex[i], sindex[j])
swapi (findex[i], findex[j]) # interchange elements
swapi (i1[i], i1[j])
swapr (d1[i], d1[j])
swapr (d2[i], d2[j])
do l = 1, naperts {
swapr (d3[l,i], d3[l,j])
swapr (d4[l,i], d4[l,j])
swapr (d5[l,i], d5[l,j])
}
if (i-lv[p] < uv[p] - i) { # stack so shorter done first
lv[p+1] = lv[p]
uv[p+1] = i - 1
lv[p] = i + 1
} else {
lv[p+1] = i + 1
uv[p+1] = uv[p]
uv[p] = i - 1
}
break
}
p = p + 1 # push onto stack
}
}
end
# PH_2AICOMPARE -- Comparison routine for PH_5R3ISORT.
int procedure ph_2aicompare (findex, sindex, fpivot, spivot)
int findex # the first index value
int sindex # the second index value
int fpivot # the first pivot value
int spivot # the second pivot value
begin
if (findex < fpivot)
return (-1)
else if (findex > fpivot)
return (1)
else if (sindex < spivot)
return (-1)
else if (sindex > spivot)
return (1)
else
return (0)
end
|