-
Notifications
You must be signed in to change notification settings - Fork 1
/
CompressPsf.C
1399 lines (1242 loc) · 45.1 KB
/
CompressPsf.C
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
#include <algorithm>
#include "CompressPsf.h"
#include "strlib.h"
#include "Molecule.h"
#include "Parameters.h"
#include "SimParameters.h"
#include "InfoStream.h"
#include "UniqueSet.h"
#include "UniqueSetIter.h"
using namespace std;
/**
* Very tricky part:
*
* Generating the tuple type affects the Parameter object.
* Particularly, the "multiplicity" field in the TUPLE_array
* will be changed in the function assign_TUPLE_index.
* TUPLE=dihedral, improper.
*
* And when we read the compressed psf files, assign_TUPLE_index
* functions will not be called so that the value of "multiplicity"
* will not be updated. This results in the incorrect energy from bonded
* interaction.
*
* Therefore, we need to store its value in the compressed psf
* file. When it comes to read them, we need to update the Parameter object
* with these values.
*/
//global variables recording compressed psf file information
Molecule *g_mol = NULL;
Parameters *g_param = NULL;
SimParameters *g_simParam = NULL;
ConfigList *g_cfgList = NULL;
struct BasicAtomInfo
{
Index segNameIdx;
Index resNameIdx;
Index atomNameIdx;
Index atomTypeIdx;
Index chargeIdx;
Index massIdx;
Index atomSigIdx;
Index exclSigIdx;
int resID;
};
void OutputAtomRecord::flip(){
flipNum((char *)&sSet, sizeof(short), sizeof(sSet)/sizeof(short));
flipNum((char *)&iSet, sizeof(int), sizeof(iSet)/sizeof(int));
flipNum((char *)&fSet, sizeof(float), sizeof(fSet)/sizeof(float));
}
struct AtomSigInfo
{
vector<SigIndex> bondSigIndices;
vector<SigIndex> angleSigIndices;
vector<SigIndex> dihedralSigIndices;
vector<SigIndex> improperSigIndices;
vector<SigIndex> crosstermSigIndices;
AtomSigInfo()
{}
AtomSigInfo(const AtomSigInfo& sig)
{
bondSigIndices.clear();
for(int i=0; i<sig.bondSigIndices.size(); i++)
bondSigIndices.push_back(sig.bondSigIndices[i]);
angleSigIndices.clear();
for(int i=0; i<sig.angleSigIndices.size(); i++)
angleSigIndices.push_back(sig.angleSigIndices[i]);
dihedralSigIndices.clear();
for(int i=0; i<sig.dihedralSigIndices.size(); i++)
dihedralSigIndices.push_back(sig.dihedralSigIndices[i]);
improperSigIndices.clear();
for(int i=0; i<sig.improperSigIndices.size(); i++)
improperSigIndices.push_back(sig.improperSigIndices[i]);
crosstermSigIndices.clear();
for(int i=0; i<sig.crosstermSigIndices.size(); i++)
crosstermSigIndices.push_back(sig.crosstermSigIndices[i]);
}
~AtomSigInfo()
{
bondSigIndices.clear();
angleSigIndices.clear();
dihedralSigIndices.clear();
improperSigIndices.clear();
crosstermSigIndices.clear();
}
void sortTupleSigIndices()
{
sort(bondSigIndices.begin(), bondSigIndices.end());
sort(angleSigIndices.begin(), angleSigIndices.end());
sort(dihedralSigIndices.begin(), dihedralSigIndices.end());
sort(improperSigIndices.begin(), improperSigIndices.end());
sort(crosstermSigIndices.begin(), crosstermSigIndices.end());
}
inline CkHashCode hash() const {
// What's a good hash function for this? Lets make something up
// Concatenate all the index lists into a list of chars, then hash that
// string using Charm's string hash function
// To keep memory allocation cheap, we'll just use a 32-byte buffer
// and wrap around if we have more sigs
const int maxlen = 32;
unsigned char keydata[maxlen+1];
const int maxchar = 256;
int i,j;
for(j=0;j<=maxlen;j++) keydata[j] = 0;
j=0;
for(i=0; i<bondSigIndices.size(); i++,j++) {
keydata[j % maxlen] ^= (bondSigIndices[i] % maxchar);
}
for(i=0; i<angleSigIndices.size(); i++,j++) {
keydata[j % maxlen] ^= (angleSigIndices[i] % maxchar);
}
for(i=0; i<dihedralSigIndices.size(); i++,j++) {
keydata[j % maxlen] ^= (dihedralSigIndices[i] % maxchar);
}
for(i=0; i<improperSigIndices.size(); i++,j++) {
keydata[j % maxlen] ^= (improperSigIndices[i] % maxchar);
}
for(i=0; i<crosstermSigIndices.size(); i++,j++) {
keydata[j % maxlen] ^= (crosstermSigIndices[i] % maxchar);
}
// CmiPrintf("Computed hash string len %d,%d\n",j,maxlen);
if (j > maxlen) j = maxlen;
// for(i=0; i < j; i++) {
// if (keydata[i] == 0)
// keydata[i] = 255;
// CmiPrintf("key[%d]=%d %p\n",i,keydata[i],keydata);
// }
return CkHashFunction_default((const void*)keydata,(size_t)j);
}
};
int operator==(const AtomSigInfo &s1, const AtomSigInfo& s2)
{
if(s1.bondSigIndices.size() != s2.bondSigIndices.size())
return 0;
if(s1.angleSigIndices.size() != s2.angleSigIndices.size())
return 0;
if(s1.dihedralSigIndices.size() != s2.dihedralSigIndices.size())
return 0;
if(s1.improperSigIndices.size() != s2.improperSigIndices.size())
return 0;
if(s1.crosstermSigIndices.size() != s2.crosstermSigIndices.size())
return 0;
int equalCnt;
equalCnt=0;
int bondSigCnt = s1.bondSigIndices.size();
for(int i=0; i<bondSigCnt; i++)
equalCnt += (s1.bondSigIndices[i]==s2.bondSigIndices[i]);
if(equalCnt!=bondSigCnt)
return 0;
equalCnt=0;
int angleSigCnt = s1.angleSigIndices.size();
for(int i=0; i<angleSigCnt; i++)
equalCnt += (s1.angleSigIndices[i]==s2.angleSigIndices[i]);
if(equalCnt!=angleSigCnt)
return 0;
equalCnt=0;
int dihedralSigCnt = s1.dihedralSigIndices.size();
for(int i=0; i<dihedralSigCnt; i++)
equalCnt += (s1.dihedralSigIndices[i]==s2.dihedralSigIndices[i]);
if(equalCnt!=dihedralSigCnt)
return 0;
equalCnt=0;
int improperSigCnt = s1.improperSigIndices.size();
for(int i=0; i<improperSigCnt; i++)
equalCnt += (s1.improperSigIndices[i]==s2.improperSigIndices[i]);
if(equalCnt!=improperSigCnt)
return 0;
equalCnt=0;
int crosstermSigCnt = s1.crosstermSigIndices.size();
for(int i=0; i<crosstermSigCnt; i++)
equalCnt += (s1.crosstermSigIndices[i]==s2.crosstermSigIndices[i]);
if(equalCnt!=crosstermSigCnt)
return 0;
return 1;
}
struct ExclSigInfo
{
vector<int> fullExclOffset;
vector<int> modExclOffset;
ExclSigInfo()
{}
ExclSigInfo(const ExclSigInfo& sig)
{
fullExclOffset.clear();
for(int i=0; i<sig.fullExclOffset.size(); i++)
fullExclOffset.push_back(sig.fullExclOffset[i]);
modExclOffset.clear();
for(int i=0; i<sig.modExclOffset.size(); i++)
modExclOffset.push_back(sig.modExclOffset[i]);
}
~ExclSigInfo()
{
fullExclOffset.clear();
modExclOffset.clear();
}
void sortExclOffset()
{
sort(fullExclOffset.begin(), fullExclOffset.end());
sort(modExclOffset.begin(), modExclOffset.end());
}
int hash() const {
unsigned int code = 0x1234;
unsigned int codesz = 8 * sizeof(int);
const unsigned int numFoffset = fullExclOffset.size();
const unsigned int numMoffset = modExclOffset.size();
const unsigned int numOffsets = numFoffset + numMoffset;
// No excluded atoms? Just hash to 0
if (numOffsets == 0)
return 0;
unsigned int shift = codesz / numOffsets;
if (shift == 0) shift=1;
unsigned int i;
for(i=0; i < numFoffset; i++) {
code = circShift(code,shift);
code ^= fullExclOffset[i];
}
for(i=0; i < numMoffset; i++) {
code = circShift(code,shift);
code ^= modExclOffset[i];
}
return code;
}
};
int operator==(const ExclSigInfo &s1, const ExclSigInfo &s2)
{
if(s1.fullExclOffset.size()!=s2.fullExclOffset.size())
return 0;
if(s1.modExclOffset.size()!=s2.modExclOffset.size())
return 0;
for(int i=0; i<s1.fullExclOffset.size(); i++)
{
if(s1.fullExclOffset[i] != s2.fullExclOffset[i])
return 0;
}
for(int i=0; i<s1.modExclOffset.size(); i++)
{
if(s1.modExclOffset[i] != s2.modExclOffset[i])
return 0;
}
return 1;
}
class HashString : public string {
public:
int hash() const {
const char* d = this->c_str();
int ret=0;
for (int i=0;d[i]!=0;i++) {
int shift1=((5*i)%16)+0;
int shift2=((6*i)%16)+8;
ret+=((0xa5^d[i])<<shift2)+(d[i]<<shift1);
}
return ret;
}
};
class HashReal {
Real val;
public:
HashReal(Real v) : val(v) {}
operator Real & () { return val; }
operator const Real & () const { return val; }
int hash() const {
const char* d = (const char *)&val;
int ret=0;
for (int i=0;i < sizeof(Real);i++) {
int shift1=((5*i)%16)+0;
int shift2=((6*i)%16)+8;
ret+=((0xa5^d[i])<<shift2)+(d[i]<<shift1);
}
return ret;
};
};
HashPool<HashString> segNamePool;
HashPool<HashString> resNamePool;
HashPool<HashString> atomNamePool;
HashPool<HashString> atomTypePool;
HashPool<HashReal> chargePool;
HashPool<HashReal> massPool;
HashPool<AtomSigInfo> atomSigPool;
BasicAtomInfo *atomData;
//Recording cluster information after reading all bonds info
int *eachAtomClusterID = NULL;
vector<int> eachClusterSize;
vector<int> eachClusterID;
int g_numClusters = 0;
HashPool<TupleSignature> sigsOfBonds;
HashPool<TupleSignature> sigsOfAngles;
HashPool<TupleSignature> sigsOfDihedrals;
HashPool<TupleSignature> sigsOfImpropers;
HashPool<TupleSignature> sigsOfCrossterms;
AtomSigInfo *eachAtomSigs;
HashPool<ExclSigInfo> sigsOfExclusions;
ExclSigInfo *eachAtomExclSigs;
//Structures for extraBond options
vector<Bond> extraBonds;
vector<Angle> extraAngles;
vector<Dihedral> extraDihedrals;
vector<Improper> extraImpropers;
vector<BondValue> extraBondParams;
vector<AngleValue> extraAngleParams;
vector<DihedralValue> extraDihedralParams;
vector<ImproperValue> extraImproperParams;
int operator==(const BondValue &b1, const BondValue &b2)
{
return (b1.k==b2.k) && (b1.x0==b2.x0);
}
int operator==(const AngleValue &a1, const AngleValue &a2)
{
return (a1.k==a2.k) && (a1.k_ub==a2.k_ub) && (a1.r_ub==a2.r_ub) && (a1.theta0==a2.theta0);
}
int operator!=(const FourBodyConsts& f1, const FourBodyConsts& f2)
{
return (f1.delta!=f2.delta) || (f1.k!=f2.k) || (f1.n!=f2.n);
}
int operator==(const DihedralValue &d1, const DihedralValue &d2)
{
if(d1.multiplicity != d2.multiplicity)
return 0;
for(int i=0; i<MAX_MULTIPLICITY; i++)
{
if(d1.values[i] != d2.values[i])
return 0;
}
return 1;
}
int operator==(const ImproperValue &d1, const ImproperValue &d2)
{
if(d1.multiplicity != d2.multiplicity)
return 0;
for(int i=0; i<MAX_MULTIPLICITY; i++)
{
if(d1.values[i] != d2.values[i])
return 0;
}
return 1;
}
void loadMolInfo();
void integrateAllAtomSigs();
void outputCompressedFile(FILE *txtOfp, FILE *binOfp);
//reading extraBond's information
void getExtraBonds(StringList *file);
void buildAtomData();
void buildBondData();
void buildAngleData();
void buildDihedralData();
void buildImproperData();
void buildCrosstermData();
void buildExclusionData();
//Functions related with building exclusions
void buildExclusions();
void build12Excls(UniqueSet<Exclusion>&, vector<int> *);
void build13Excls(UniqueSet<Exclusion>&, vector<int> *);
void build14Excls(UniqueSet<Exclusion>&, vector<int> *, int);
//reverse the byte-order of every element starting at "elem"
void flipNum(char *elem, int elemSize, int numElems){
int mid = elemSize/2;
char *ptr = elem;
for(int i=0; i<numElems; i++) {
for(int j=0; j<mid; j++) {
char tmp = ptr[j];
ptr[j] = ptr[elemSize-1-j];
ptr[elemSize-1-j] = tmp;
}
ptr += elemSize;
}
}
void clearGlobalVectors()
{
segNamePool.clear();
resNamePool.clear();
atomNamePool.clear();
atomTypePool.clear();
chargePool.clear();
massPool.clear();
sigsOfBonds.clear();
sigsOfAngles.clear();
sigsOfDihedrals.clear();
sigsOfImpropers.clear();
sigsOfExclusions.clear();
eachClusterSize.clear();
}
void compress_molecule_info(Molecule *mol, char *psfFileName, Parameters *param, SimParameters *simParam, ConfigList* cfgList)
{
g_mol = mol;
g_param = param;
g_simParam = simParam; //used for building exclusions
g_cfgList = cfgList; //used for integrating extra bonds
//read psf files
//readPsfFile(psfFileName);
loadMolInfo();
integrateAllAtomSigs();
buildExclusions();
//buildParamData();
char *outFileName = new char[strlen(psfFileName)+20];
sprintf(outFileName, "%s.inter", psfFileName);
//the text file for signatures and other non-per-atom info
FILE *txtOfp = fopen(outFileName, "w");
sprintf(outFileName, "%s.inter.bin", psfFileName);
//the binary file for per-atom info
FILE *binOfp = fopen(outFileName, "wb");
delete [] outFileName;
//output compressed psf file
outputCompressedFile(txtOfp, binOfp);
fclose(txtOfp);
fclose(binOfp);
}
/** Before entering this function, all information about
* Molecule has been obtained. The
* Molecule::build_atom_status should also be called.
*/
void loadMolInfo()
{
char err_msg[512]; // Error message for NAMD_die
char buffer[512]; // Buffer for file reading
int i; // Loop counter
int NumTitle; // Number of Title lines in .psf file
int ret_code; // ret_code from NAMD_read_line calls
FILE *psf_file;
Parameters *params = g_param;
buildAtomData();
//read extra bonds/angles/dihedrals/impropers information first
//and then integrate them into the following reading procedures
/*if(g_simParam->extraBondsOn)
{
getExtraBonds(g_cfgList->find("extraBondsFile"));
}*/
//initialize eachAtomSigs
eachAtomSigs = new AtomSigInfo[g_mol->numAtoms];
buildBondData();
buildAngleData();
buildDihedralData();
buildImproperData();
buildCrosstermData();
// analyze the data and find the status of each atom
//build_atom_status();
}
void integrateAllAtomSigs()
{
printf("Bond sigs: %d\n", (int)sigsOfBonds.size());
printf("Angle sigs: %d\n", (int)sigsOfAngles.size());
printf("Dihedral sigs: %d\n", (int)sigsOfDihedrals.size());
printf("Improper sigs: %d\n", (int)sigsOfImpropers.size());
printf("Crossterm sigs: %d\n", (int)sigsOfCrossterms.size());
for(int i=0; i<g_mol->numAtoms; i++)
{
eachAtomSigs[i].sortTupleSigIndices();
int poolIndex = atomSigPool.lookupCstPool(eachAtomSigs[i]);
if(poolIndex==-1)
{
atomSigPool.push_back(eachAtomSigs[i]);
poolIndex = atomSigPool.size()-1;
}
atomData[i].atomSigIdx = poolIndex;
}
printf("Atom's sigs: %d\n", (int)atomSigPool.size());
delete[] eachAtomSigs;
}
/**
* Output the compressed psf files. The binary per-atom file
* contains two part. The first part is used for the parallel
* input containing info such as atom signature ids; the second
* part is used for the parallel output containing infor such as
* cluster ids and whether an atom is water or not.
*
* -Chao Mei
*
*/
void outputCompressedFile(FILE *txtOfp, FILE *binOfp)
{
#ifndef MEM_OPT_VERSION
fprintf(txtOfp, "FORMAT VERSION: %f\n", COMPRESSED_PSF_VER);
fprintf(txtOfp, "%d !NSEGMENTNAMES\n", segNamePool.size());
for(int i=0; i<segNamePool.size(); i++)
{
fprintf(txtOfp, "%s\n", segNamePool[i].c_str());
}
fprintf(txtOfp, "%d !NRESIDUENAMES\n", resNamePool.size());
for(int i=0; i<resNamePool.size(); i++)
{
fprintf(txtOfp, "%s\n", resNamePool[i].c_str());
}
fprintf(txtOfp, "%d !NATOMNAMES\n", atomNamePool.size());
for(int i=0; i<atomNamePool.size(); i++)
{
fprintf(txtOfp, "%s\n", atomNamePool[i].c_str());
}
fprintf(txtOfp, "%d !NATOMTYPES\n", atomTypePool.size());
for(int i=0; i<atomTypePool.size(); i++)
{
fprintf(txtOfp, "%s\n", atomTypePool[i].c_str());
}
fprintf(txtOfp, "%d !NCHARGES\n", chargePool.size());
for(int i=0; i<chargePool.size(); i++)
{
const Real charge = chargePool[i];
fprintf(txtOfp, "%f\n", charge);
}
fprintf(txtOfp, "%d !NMASSES\n", massPool.size());
for(int i=0; i<massPool.size(); i++)
{
const Real mass = massPool[i];
fprintf(txtOfp, "%f\n", mass);
}
fprintf(txtOfp, "%d !NATOMSIGS\n", atomSigPool.size());
for(int i=0; i<atomSigPool.size(); i++)
{
AtomSigInfo& oneAtomSig = atomSigPool[i];
int oneTypeCnt = oneAtomSig.bondSigIndices.size();
fprintf(txtOfp, "%d !%sSIGS\n", oneTypeCnt, "NBOND");
for(int j=0; j<oneTypeCnt; j++)
{
SigIndex idx = oneAtomSig.bondSigIndices[j];
TupleSignature& tSig = sigsOfBonds[idx];
tSig.output(txtOfp);
}
oneTypeCnt = oneAtomSig.angleSigIndices.size();
fprintf(txtOfp, "%d !%sSIGS\n", oneTypeCnt, "NTHETA");
for(int j=0; j<oneTypeCnt; j++)
{
SigIndex idx = oneAtomSig.angleSigIndices[j];
TupleSignature& tSig = sigsOfAngles[idx];
tSig.output(txtOfp);
}
oneTypeCnt = oneAtomSig.dihedralSigIndices.size();
fprintf(txtOfp, "%d !%sSIGS\n", oneTypeCnt, "NPHI");
for(int j=0; j<oneTypeCnt; j++)
{
SigIndex idx = oneAtomSig.dihedralSigIndices[j];
TupleSignature& tSig = sigsOfDihedrals[idx];
tSig.output(txtOfp);
}
oneTypeCnt = oneAtomSig.improperSigIndices.size();
fprintf(txtOfp, "%d !%sSIGS\n", oneTypeCnt, "NIMPHI");
for(int j=0; j<oneTypeCnt; j++)
{
SigIndex idx = oneAtomSig.improperSigIndices[j];
TupleSignature& tSig = sigsOfImpropers[idx];
tSig.output(txtOfp);
}
oneTypeCnt = oneAtomSig.crosstermSigIndices.size();
fprintf(txtOfp, "%d !%sSIGS\n", oneTypeCnt, "NCRTERM");
for(int j=0; j<oneTypeCnt; j++)
{
SigIndex idx = oneAtomSig.crosstermSigIndices[j];
TupleSignature& tSig = sigsOfCrossterms[idx];
tSig.output(txtOfp);
}
}
//2. Output exclusion signatures
int exclSigCnt = sigsOfExclusions.size();
fprintf(txtOfp, "%d !NEXCLSIGS\n", exclSigCnt);
for(int i=0; i<exclSigCnt; i++)
{
ExclSigInfo *sig = &sigsOfExclusions[i];
//first line is for full exclusions (1-2, 1-3) in the format of count offset1 offset2 offset3 ...
fprintf(txtOfp, "%d", sig->fullExclOffset.size());
for(int j=0; j<sig->fullExclOffset.size(); j++)
fprintf(txtOfp, " %d", sig->fullExclOffset[j]);
fprintf(txtOfp, "\n");
//second line is for modified exclusions (1-4)
fprintf(txtOfp, "%d", sig->modExclOffset.size());
for(int j=0; j<sig->modExclOffset.size(); j++)
fprintf(txtOfp, " %d", sig->modExclOffset[j]);
fprintf(txtOfp, "\n");
}
//3. Output the cluster information
fprintf(txtOfp, "%d !NCLUSTERS\n", g_numClusters);
//4. Output atom info
fprintf(txtOfp, "%d !NATOM\n", g_mol->numAtoms);
fprintf(txtOfp, "%d !NHYDROGENGROUP\n", g_mol->numHydrogenGroups);
fprintf(txtOfp, "%d !MAXHYDROGENGROUPSIZE\n", g_mol->maxHydrogenGroupSize);
fprintf(txtOfp, "%d !NMIGRATIONGROUP\n", g_mol->numMigrationGroups);
fprintf(txtOfp, "%d !MAXMIGRATIONGROUPSIZE\n", g_mol->maxMigrationGroupSize);
//5. Output rigid bond type
fprintf(txtOfp, "%d !RIGIDBONDTYPE\n", g_simParam->rigidBonds);
#if 0
const float *atomOccupancy = g_mol->getOccupancyData();
const float *atomBFactor = g_mol->getBFactorData();
fprintf(txtOfp, "%d !OCCUPANCYVALID\n", (atomOccupancy==NULL)?0:1);
fprintf(txtOfp, "%d !TEMPFACTORVALID\n", (atomBFactor==NULL)?0:1);
float *zeroFloats = NULL;
if(atomOccupancy==NULL || atomBFactor==NULL) {
zeroFloats = new float[g_mol->numAtoms];
memset(zeroFloats, 0, sizeof(float)*g_mol->numAtoms);
if(atomOccupancy==NULL) atomOccupancy = (const float *)zeroFloats;
if(atomBFactor==NULL) atomBFactor = (const float *)zeroFloats;
}
#endif
Atom *atoms = g_mol->getAtoms(); //need to output its partner and hydrogenList
HydrogenGroupID *hg = g_mol->hydrogenGroup.begin();
//First, output magic number
int magicNum = COMPRESSED_PSF_MAGICNUM;
fwrite(&magicNum, sizeof(int), 1, binOfp);
//Second, version number
float verNum = (float)COMPRESSED_PSF_VER;
fwrite(&verNum, sizeof(float), 1, binOfp);
//Third, the per-atom record size
int recSize = sizeof(OutputAtomRecord);
fwrite(&recSize, sizeof(int), 1, binOfp);
//Fourth, each atom info
OutputAtomRecord oneRec;
for(int i=0; i<g_mol->numAtoms; i++)
{
oneRec.sSet.segNameIdx = atomData[i].segNameIdx;
oneRec.sSet.resNameIdx = atomData[i].resNameIdx;
oneRec.sSet.atomNameIdx = atomData[i].atomNameIdx;
oneRec.sSet.atomTypeIdx = atomData[i].atomTypeIdx;
oneRec.sSet.chargeIdx = atomData[i].chargeIdx;
oneRec.sSet.massIdx = atomData[i].massIdx;
oneRec.iSet.atomSigIdx = atomData[i].atomSigIdx;
oneRec.iSet.exclSigIdx = atomData[i].exclSigIdx;
oneRec.sSet.vdw_type = atoms[i].vdw_type;
oneRec.iSet.resID = atomData[i].resID;
int hydIdx = atoms[i].hydrogenList;
oneRec.iSet.hydrogenList = hydIdx;
oneRec.iSet.atomsInGroup = hg[hydIdx].atomsInGroup;
oneRec.iSet.GPID = hg[hydIdx].GPID;
//oneRec.waterVal = hg[hydIdx].waterVal;
oneRec.iSet.atomsInMigrationGroup = hg[hydIdx].atomsInMigrationGroup;
oneRec.iSet.MPID = hg[hydIdx].MPID;
//oneRec.fSet.occupancy = atomOccupancy[i];
//oneRec.fSet.bfactor = atomBFactor[i];
oneRec.fSet.rigidBondLength = g_mol->rigid_bond_length(i);
fwrite(&oneRec, sizeof(OutputAtomRecord), 1, binOfp);
}
//if(zeroFloats) delete[] zeroFloats;
delete[] atomData;
//Output the info required for parallel output:
//1. Cluster ids of each atom
//Since the cluster info is only when doing output under
//wrapAll or wrapWater conditions, for the purpose of parallel
//output, it is reasonable to separate the cluster info from
//the above per-atom info. So whole the binary per-atom file
//contains two parts, one for parallel input to create patches
//and computes; the other for parallel output -Chao Mei
//Fifth: output the cluster id of each atoms
fwrite(eachAtomClusterID, sizeof(int), g_mol->numAtoms, binOfp);
//2. Whether each atom is water or not
char *isWater = new char[g_mol->numAtoms];
for(int i=0; i<g_mol->numAtoms; i++){
isWater[i] = g_mol->is_water(i);
}
fwrite(isWater, sizeof(char), g_mol->numAtoms, binOfp);
delete [] isWater;
delete[] atoms;
g_mol->hydrogenGroup.resize(0);
delete[] eachAtomClusterID;
//Output the parameter new values if extraBonds are present.
//The parameters are not needed since now extraBonds' parameters will be
//read again during running the simulation
//6. Output the "multiplicity" field TUPLE_array of the Parameter object
fprintf(txtOfp, "!DIHEDRALPARAMARRAY\n");
for(int i=0; i<g_param->NumDihedralParams; i++)
{
fprintf(txtOfp, "%d ", g_param->dihedral_array[i].multiplicity);
}
fprintf(txtOfp, "\n");
fprintf(txtOfp, "!IMPROPERPARAMARRAY\n");
for(int i=0; i<g_param->NumImproperParams; i++)
{
fprintf(txtOfp, "%d ", g_param->improper_array[i].multiplicity);
}
fprintf(txtOfp, "\n");
#endif
}
void buildAtomData()
{
#ifndef MEM_OPT_VERSION
int numAtoms = g_mol->numAtoms;
//1. parse atom data to build constant pool (atom name, mass, charge etc.)
atomData = new BasicAtomInfo[numAtoms];
Atom *atoms = g_mol->getAtoms();
AtomNameInfo *atomNames = g_mol->getAtomNames();
AtomSegResInfo *atomSegResids = g_mol->getAtomSegResInfo();
for(int atomID=0; atomID < numAtoms; atomID++)
{
//building constant pool
int poolIndex;
HashString fieldName;
fieldName.assign(atomSegResids[atomID].segname);
poolIndex = segNamePool.lookupCstPool(fieldName);
if(poolIndex==-1)
{
segNamePool.push_back(fieldName);
poolIndex = segNamePool.size()-1;
}
atomData[atomID].segNameIdx = poolIndex;
atomData[atomID].resID = atomSegResids[atomID].resid;
fieldName.assign(atomNames[atomID].resname);
poolIndex = resNamePool.lookupCstPool(fieldName);
if(poolIndex==-1)
{
resNamePool.push_back(fieldName);
poolIndex = resNamePool.size()-1;
}
atomData[atomID].resNameIdx = poolIndex;
fieldName.assign(atomNames[atomID].atomname);
poolIndex = atomNamePool.lookupCstPool(fieldName);
if(poolIndex==-1)
{
atomNamePool.push_back(fieldName);
poolIndex = atomNamePool.size()-1;
}
atomData[atomID].atomNameIdx = poolIndex;
fieldName.assign(atomNames[atomID].atomtype);
poolIndex = atomTypePool.lookupCstPool(fieldName);
if(poolIndex==-1)
{
atomTypePool.push_back(fieldName);
poolIndex = atomTypePool.size()-1;
}
atomData[atomID].atomTypeIdx = poolIndex;
poolIndex = chargePool.lookupCstPool(atoms[atomID].charge);
if(poolIndex==-1)
{
chargePool.push_back(atoms[atomID].charge);
poolIndex = chargePool.size()-1;
}
atomData[atomID].chargeIdx = poolIndex;
poolIndex = massPool.lookupCstPool(atoms[atomID].mass);
if(poolIndex==-1)
{
massPool.push_back(atoms[atomID].mass);
poolIndex = massPool.size()-1;
}
atomData[atomID].massIdx = poolIndex;
}
//Free those space to reduce transient memory usage
//delete [] atoms; (deleted until per-atom info is output)
delete [] atomNames;
delete [] atomSegResids;
#endif
}
void buildBondData()
{
#ifndef MEM_OPT_VERSION
Bond *bonds = g_mol->getAllBonds();
//then creating bond's tupleSignature
for(int i=0; i<g_mol->numBonds; i++)
{
Bond *b = bonds+i;
TupleSignature oneSig(1,BOND,b->bond_type);
oneSig.offset[0] = b->atom2 - b->atom1;
oneSig.isReal = (i<g_mol->numRealBonds);
int poolIndex = sigsOfBonds.lookupCstPool(oneSig);
int newSig=0;
if(poolIndex == -1)
{
sigsOfBonds.push_back(oneSig);
poolIndex = (SigIndex)sigsOfBonds.size()-1;
newSig=1;
}
if(!newSig)
{//check duplicate bonds in the form of (a, b) && (a, b);
int dupIdx = lookupCstPool(eachAtomSigs[b->atom1].bondSigIndices, (SigIndex)poolIndex);
if(dupIdx!=-1)
{
char err_msg[128];
sprintf(err_msg, "Duplicate bond %d-%d!", b->atom1+1, b->atom2+1);
NAMD_die(err_msg);
}
}
eachAtomSigs[b->atom1].bondSigIndices.push_back(poolIndex);
}
//check duplicate bonds in the form of (a, b) && (b, a)
for(int i=0; i<g_mol->numBonds; i++)
{
Bond *b=bonds+i;
int atom2 = b->atom2;
int thisOffset = atom2 - b->atom1;
for(int j=0; j<eachAtomSigs[atom2].bondSigIndices.size(); j++)
{
SigIndex atom2BondId = eachAtomSigs[atom2].bondSigIndices[j];
TupleSignature *secSig = &(sigsOfBonds[atom2BondId]);
if(thisOffset== -(secSig->offset[0]))
{
char err_msg[128];
sprintf(err_msg, "Duplicate bond %d-%d because two atoms are just reversed!", b->atom1+1, atom2+1);
NAMD_die(err_msg);
}
}
}
//building clusters for this simulation system in two steps
//1. create a list for each atom where each atom in the list is bonded with that atom
vector<int> *atomListOfBonded = new vector<int>[g_mol->numAtoms];
for(int i=0; i<g_mol->numRealBonds; i++)
{
Bond *b=bonds+i;
int atom1 = b->atom1;
int atom2 = b->atom2;
atomListOfBonded[atom1].push_back(atom2);
atomListOfBonded[atom2].push_back(atom1);
}
delete [] bonds;
//2. using breadth-first-search to build the clusters. Here, we avoid recursive call
// because the depth of calls may be of thousands which will blow up the stack, and
//recursive call is slower than the stack-based BFS.
//Considering such structure
//1->1245; 7->1243; 1243->1245
eachAtomClusterID = new int[g_mol->numAtoms];
for(int i=0; i<g_mol->numAtoms; i++)
eachAtomClusterID[i] = -1;
//It is guaranteed that the clusters found in this way use the
//smallest atom id of this cluster because each atom is at least
//connected to one of the atoms in its cluster (by atomListOfBonded
//constructed above).
//--Chao Mei
for(int i=0; i<g_mol->numAtoms; i++)
{
int curClusterID=eachAtomClusterID[i];
//if the atom's cluster id is not -1, the atom has been visited
if(curClusterID!=-1) continue;
curClusterID=i;
deque<int> toVisitAtoms;
eachAtomClusterID[i] = curClusterID;
toVisitAtoms.push_back(i);
while(!toVisitAtoms.empty())
{
int visAtomID = toVisitAtoms.front();
toVisitAtoms.pop_front();
for(int j=0; j<atomListOfBonded[visAtomID].size(); j++)
{
int otherAtom = atomListOfBonded[visAtomID][j];
if(eachAtomClusterID[otherAtom]!=curClusterID){
eachAtomClusterID[otherAtom]=curClusterID;
toVisitAtoms.push_back(otherAtom);
}
}
}
}
#if 0
//Now the clusterID of each atom should be usually in the non-decreasing
//order. In other words, the atom ids of a cluster are generally contiguous.
//If this is the case, the temporary memory usage of output IO during
//the simulation can be dramatically reduced. So g_isClusterContiguous
//is used to differentiate the two cases:
//1. if the cluster id of atoms is monotonically increasing
//(g_isClusterContiguous=1), the size of the cluster can be used as this
//this cluster's signature (represented by "eachAtomClusterID").
//The atom who is the first atom (in terms of atom id) in this cluster will
//store the cluster size as its signature. The remaining atoms in this
//cluster store -1.
//2. if the cluster id of atoms is not monotonically increasing, that is,
//the atom ids of a cluster are not contiguous (g_isClusterContiguous=0).
//Then we have to still record each atom's cluster id in variable
//"eachAtomClusterID", and we have to change each unique cluster id (valued
//at the scale of numAtoms) into indexes at the scale of numClusters. For
//example, the cluster ids of atoms up to this point may be (0...0, 24...24,
//789...789,...), after re-scaling, cluster ids should be (0...0,1...1,
//2...2,....) for atoms.
//We made an ASSUMPTION here: the cluster ids are in the non-decreasing
//order, and when a cluster discontiguity happens, the cluster id has
//appeared before! Such ASSUMPTION is to make implementation easy.
int curClusterID;
int prevClusterID = eachAtomClusterID[0];
int curClusterSize = 1;
//step1: segment all atoms according to each cluster
for(int i=1; i<g_mol->numAtoms; i++){
curClusterID = eachAtomClusterID[i];
if(curClusterID == prevClusterID){
curClusterSize++;
}else{
eachClusterSize.push_back(curClusterSize);
eachClusterID.push_back(prevClusterID);
curClusterSize=1;
}
prevClusterID = curClusterID;
}
//record the info of the last cluster segment
eachClusterSize.push_back(curClusterSize);
eachClusterID.push_back(prevClusterID);
//step2: detect contiguity of atoms in each cluster and re-scale
//cluster id.
g_isClusterContiguous = 1;
int *newClusterIDs = new int[eachClusterID.size()];
memset(newClusterIDs, 0, sizeof(int)*eachClusterID.size());