-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathNetflixDataNG.cpp
executable file
·1128 lines (989 loc) · 33.1 KB
/
NetflixDataNG.cpp
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 "NetflixDataNG.h"
#include "MD5.h"
namespace utils {
// Constructor and Destructor
NetflixDataNG::NetflixDataNG() {
debug = false;
test = false;
// Populate path variables
basePathToUse = beastBasePath;
dataPath = basePathToUse + pathToData;
qualDataFile = basePathToUse + pathToQualifyingDataFile;
// Populate the User id converter
populateUserIdConverter(basePathToUse + pathToConversionFile);
// Now need to populate the array of calendar information
// First do 1998 itself. Populate with 0s up until November
for (int month = 1; month <= 10; month++) {
daysAfter30Sept1998[0][month - 1] = 0;
}
// November (October has 31)
daysAfter30Sept1998[0][10] = 31;
// December (November has 30)
daysAfter30Sept1998[0][11] = 61;
// Now iterate over years and populate forward depending on number of days in PREVIOUS month
for (int year = 1999; year <= 2005; year++) {
// Jan (Dec has 31, from previous year)
daysAfter30Sept1998[year - 1998][0] =
daysAfter30Sept1998[year - 1999][11] + 31;
// Feb (Jan has 31)
daysAfter30Sept1998[year - 1998][1] =
daysAfter30Sept1998[year - 1998][0] + 31;
// March (February has 28, or 29 in 2000 and 2004)
if ((year % 4) == 0) {
daysAfter30Sept1998[year - 1998][2] = daysAfter30Sept1998[year
- 1998][1] + 29;
} else {
daysAfter30Sept1998[year - 1998][2] = daysAfter30Sept1998[year
- 1998][1] + 28;
}
// April (March has 31)
daysAfter30Sept1998[year - 1998][3] =
daysAfter30Sept1998[year - 1998][2] + 31;
// May (April has 30)
daysAfter30Sept1998[year - 1998][4] =
daysAfter30Sept1998[year - 1998][3] + 30;
// June (May has 31)
daysAfter30Sept1998[year - 1998][5] =
daysAfter30Sept1998[year - 1998][4] + 31;
// July (June has 30)
daysAfter30Sept1998[year - 1998][6] =
daysAfter30Sept1998[year - 1998][5] + 30;
// August (July has 31)
daysAfter30Sept1998[year - 1998][7] =
daysAfter30Sept1998[year - 1998][6] + 31;
// September (August has 31)
daysAfter30Sept1998[year - 1998][8] =
daysAfter30Sept1998[year - 1998][7] + 31;
// October (September has 30)
daysAfter30Sept1998[year - 1998][9] =
daysAfter30Sept1998[year - 1998][8] + 30;
// November (October has 31)
daysAfter30Sept1998[year - 1998][10] =
daysAfter30Sept1998[year - 1998][9] + 31;
// December (November has 30)
daysAfter30Sept1998[year - 1998][11] =
daysAfter30Sept1998[year - 1998][10] + 30;
}
}
NetflixDataNG::~NetflixDataNG() {
}
// File management public helper functions
void NetflixDataNG::deleteFileAsynch(string fileName) {
stringstream s("");
s << "rm " << fileName << " > /dev/null 2>&1 &"; // This extra bit silences it if the delete fails (after all, who cares?)
string command = s.str();
if (system(&command[0]) != 0) {
// Don't care if file isn't deleted
}
}
void NetflixDataNG::zipFileAsynch(string fileName, string newFileName) {
stringstream s("");
s << "gzip -f -q " << fileName << " " << newFileName << " &";
string command = s.str();
if (system(&command[0]) != 0) {
cerr << "Couldn't zip file " << fileName << "\n";
}
}
void NetflixDataNG::zipFileSynch(string fileName, string newFileName) {
stringstream s("");
s << "gzip -f -q " << fileName << " " << newFileName;
string command = s.str();
if (system(&command[0]) != 0) {
cerr << "Couldn't synchronously zip file " << fileName << "\n";
}
}
void NetflixDataNG::unzipFileSynch(string fileName, string newFileName) {
stringstream s("");
s << "gunzip " << fileName << " " << newFileName;
string command = s.str();
if (system(&command[0]) != 0) {
cerr << "Couldn't synchronously unzip file " << fileName << "\n";
}
}
void NetflixDataNG::makeEntryAsynch(string submissionFileName) {
cerr << "Submitting latest prediction file\n";
// Zip the file
string zippedFileName = submissionFileName + ".gz";
zipFileSynch(submissionFileName, zippedFileName);
// Calculate the md5sum
MD5 myWrapper;
string hash = myWrapper.getHashFromFile(zippedFileName);
// Command:
// -s for silent, i.e. no progress or error messages.
// -F for each input
// URL to submit to
// Send output html to null to avoid displaying.
stringstream s("");
s << "curl " << "-s " << "-F 'teamName=TeamOz' "
<< "-F 'teamPassword=Selmer3' " << "-F 'MD5Hash=" << hash << "' "
<< "-F 'file=@" << zippedFileName << "' "
<< "http://www.netflixprize.com/submissions " << ">> null &";
string command = s.str();
if (system(&command[0]) != 0) {
cerr << "Couldn't asynchronously submit file " << zippedFileName
<< ", hash " << hash << "\n";
}
}
// Data read functions
void NetflixDataNG::getDataMeanAndVariance(double mAndV[]) {
try {
// Open VitalStatistics file
stringstream s("");
s << basePathToUse;
s << pathToVitalStatistics;
string path = s.str();
individualReader.open(&path[0]);
// Read line which is "mean, variance"
string line(100, '\0');
individualReader.getline(&line[0], 99);
char * pEnd;
mAndV[0] = strtod(&line[0], &pEnd);
pEnd = &pEnd[1];
mAndV[1] = strtod(pEnd, NULL);
} catch (...) {
cerr << "Could not get the Data mean and variance";
}
individualReader.close();
}
//void NetflixDataNG::allViewingsUsers(
// movie** &moviesByUser,
// rate** &ratesByUser,
// unsigned short* &noViewingsPerUser)
//{
// if(!test)
// {
// // Define variables for reading from files
// string path;
// ifstream reader;
// string line(100, '\0');
// char*pEnd;
//
// // Define variables for storing read ratings, movies and recording number of viewings etc.
// movie mov;
// rate rat;
// vector<rate> rates;
// vector<movie> moviesVec;
// int views;
//
// // Parallelise making user-specific stuff private to each thread.
//#pragma omp parallel for private(pEnd, mov, rat, path, views, reader) firstprivate(rates, moviesVec, line) schedule(guided,5)
//
// for (user u=0; u<U; u++)
// {
// if(u%48000==0)
// cout<<u<<"\n";
//
// rates.clear();
// moviesVec.clear();
//
// // Open file and get to ratings data (past header line)
// path = getTrainingDataPath(u, true);
// reader.open(&path[0], ios_base::in);
// reader.getline(&line[0], 99);
//
// while (true)
// {
// // Read lines until EOF
// reader.getline(&line[0], 99);
// if (line[0] == '\0')
// break;
//
// // Obtain movie and rating
// getQuartetFromUserData(mov, rat, line, pEnd);
//
// // Add these to the vector for this user.
// moviesVec.push_back(mov);
// rates.push_back(rat);
// }
//
// // Close reader (ASAP)
// reader.close();
//
// // Update data arrays with the number of viewings by this user
// views = moviesVec.size();
// noViewingsPerUser[u] = (short) views;
// moviesByUser[u] = new movie[views];
// ratesByUser[u] = new rate[views];
//
// // Update data arrays with the movies and ratings
// for (int index=0; index<views; index++)
// {
// moviesByUser[u][index] = moviesVec[index];
// ratesByUser[u][index] = rates[index];
// }
// }
//#pragma omp barrier
// }
// else
// {
// // If in test mode we populate the arrays with 3 random movies and ratings
// srand(time(0));
// rand();
// int views=3;
//
// // Iterate over users
// for (user u=0; u<U; u++)
// {
// noViewingsPerUser[u] = views;
// moviesByUser[u] = new movie[views];
// ratesByUser[u] = new rate[views];
//
// // First rating simply random
// moviesByUser[u][0] = (rand() % M);
// ratesByUser[u][0] = 1 + (rand() % 5);
//
// // Second rating only OK if movie isn't the same
// do
// {
// moviesByUser[u][1] = (rand() % M);
// }
// while(moviesByUser[u][1] == moviesByUser[u][0]);
// ratesByUser[u][1] = 1 + (rand() % 5);
//
// // Third rating only OK if movie isn't the same as first 2
// do
// {
// moviesByUser[u][2] = (rand() % M);
// }
// while(moviesByUser[u][2]==moviesByUser[u][1] || moviesByUser[u][2]==moviesByUser[u][0]);
// ratesByUser[u][2] = 1 + (rand() % 5);
// }
// }
//}
void NetflixDataNG::allViewingsMovies(user** &usersByMovie,
unsigned long* &noViewingsPerMovie, rate** &ratesByMovie,
rate** &ratesByUser, movie** &moviesByUser,
unsigned short* &noViewingsPerUser,
unsigned short ** &whatIndexAMovieIsToAUser) {
if (!test) {
// Variables for each rating
user use;
rate rat;
// Variables for each movie
ifstream reader;
string path;
long views;
char*pEnd;
#pragma omp parallel for private(use,rat,path,views, pEnd) private(reader) schedule(guided,5)
for (movie m = 0; m < M; m++) {
pEnd = 0;
// Make vectors to store the ratings and user data
vector<rate> rates;
vector<user> usersVec;
// IO stuff
string line(100, '\0');
path = getTrainingDataPath(m, false);
reader.open(&path[0], ios_base::in);
reader.getline(&line[0], 99); // Line of rubbish
// Until EOF, read in line and get user and rating from it.
while (true) {
reader.getline(&line[0], 99);
if (line[0] == '\0') {
break;
}
getQuartetFromMovieData(use, rat, line, pEnd);
usersVec.push_back(use);
rates.push_back(rat);
}
// Close reader ASAP
reader.close();
// Use number of viewings to make arrays and populate noViewingsPerMovie
views = (long) usersVec.size();
noViewingsPerMovie[m] = views;
usersByMovie[m] = new user[views];
ratesByMovie[m] = new rate[views];
whatIndexAMovieIsToAUser[m] = new unsigned short[views];
// Populate usersByMovie & ratesByMovie
for (int index = 0; index < views; index++) {
usersByMovie[m][index] = usersVec[index];
ratesByMovie[m][index] = rates[index];
}
}
#pragma omp barrier
// This is new stuff for getting user-perspective ratings from movie-perspective ones.
// Make vector arrays for storing the ratings user-side.
#define USERSATONCE 1000
vector<rate> urates[USERSATONCE];
vector<movie> umovies[USERSATONCE];
#pragma omp parallel for firstprivate(urates, umovies)
for (uint i = 0; i <= U / USERSATONCE; i++) {
// Limit to number of users dealt with by one iteration is either USERSATONCE or
// on last iteration 189 ( = U - i*USERSATONCE)
uint jLim = min((uint) USERSATONCE, U - i * USERSATONCE);
// Make sure vectors are empty
for (uint jj = 0; jj < jLim; jj++) {
urates[jj].clear();
umovies[jj].clear();
}
// Go through each movie and index & add rating and movie to the vectors
for (movie m = 0; m < M; m++) {
for (uint index = 0; index < noViewingsPerMovie[m]; index++) {
user u = usersByMovie[m][index];
if (u / USERSATONCE == i) {
urates[u % USERSATONCE].push_back(
ratesByMovie[m][index]);
umovies[u % USERSATONCE].push_back(m);
whatIndexAMovieIsToAUser[m][index] = urates[u
% USERSATONCE].size() - 1;
}
}
}
// Go through each user in this range
for (uint jj = 0; jj < jLim; jj++) {
// Get user number & number of ratings
uint views = urates[jj].size();
user u = i * USERSATONCE + jj;
// Fill in user info
noViewingsPerUser[u] = views;
moviesByUser[u] = new movie[views];
ratesByUser[u] = new rate[views];
for (uint index = 0; index < views; index++) {
moviesByUser[u][index] = umovies[jj][index];
ratesByUser[u][index] = urates[jj][index];
}
}
}
#pragma omp barrier
} else {
// If in test mode we populate the user arrays with 3 random movies and ratings
srand(time(0));
rand();
int views = 3;
// Iterate over users
for (user u = 0; u < U; u++) {
noViewingsPerUser[u] = views;
moviesByUser[u] = new movie[views];
ratesByUser[u] = new rate[views];
// First rating simply random
moviesByUser[u][0] = (rand() % M);
ratesByUser[u][0] = 1 + (rand() % 5);
// Second rating only OK if movie isn't the same
do {
moviesByUser[u][1] = (rand() % M);
} while (moviesByUser[u][1] == moviesByUser[u][0]);
ratesByUser[u][1] = 1 + (rand() % 5);
// Third rating only OK if movie isn't the same as first 2
do {
moviesByUser[u][2] = (rand() % M);
} while (moviesByUser[u][2] == moviesByUser[u][1]
|| moviesByUser[u][2] == moviesByUser[u][0]);
ratesByUser[u][2] = 1 + (rand() % 5);
}
// Vectors for each movie (rate, user, index)
vector<rate> rates[M];
vector<user> usersVec[M];
vector<rate> indexes[M];
for (user u = 0; u < U; u++) {
for (int index = 0; index < noViewingsPerUser[u]; index++) {
rates[moviesByUser[u][index]].push_back(ratesByUser[u][index]);
usersVec[moviesByUser[u][index]].push_back(u);
indexes[moviesByUser[u][index]].push_back(index);
}
}
int length;
for (movie m = 0; m < M; m++) {
// Use length to make arrays
length = usersVec[m].size();
noViewingsPerMovie[m] = length;
usersByMovie[m] = new user[length];
ratesByMovie[m] = new rate[length];
whatIndexAMovieIsToAUser[m] = new short unsigned int[length];
// Populate arrays with user, rate, index
for (int i = 0; i < length; i++) {
usersByMovie[m][i] = usersVec[m][i];
ratesByMovie[m][i] = rates[m][i];
whatIndexAMovieIsToAUser[m][i] = indexes[m][i];
}
}
}
}
void NetflixDataNG::allViewingsMovies(user** &usersByMovie,
ulong* &noViewingsPerMovie, rate** &ratesByMovie, uint** &datesByMovie,
rate** &ratesByUser, movie** &moviesByUser, ushort* &noViewingsPerUser,
uint** &datesByUser, unsigned short ** &whatIndexAMovieIsToAUser) {
if (!test) {
// Variables for each rating
user use;
rate rat;
uint date;
// Variables for each movie
ifstream reader;
string path;
long views;
char*pEnd;
#pragma omp parallel for private(use,rat, date, path,views, pEnd) private(reader) schedule(guided,5)
for (movie m = 0; m < M; m++) {
pEnd = 0;
// Make vectors to store the ratings and user data
vector<rate> rates;
vector<user> usersVec;
vector < uint > dateVec;
// IO stuff
string line(100, '\0');
path = getTrainingDataPath(m, false);
reader.open(&path[0], ios_base::in);
reader.getline(&line[0], 99); // Line of rubbish
// Until EOF, read in line and get user and rating from it.
while (true) {
reader.getline(&line[0], 99);
if (line[0] == '\0') {
break;
}
getInfoFromMovieData(use, rat, date, line, pEnd);
usersVec.push_back(use);
rates.push_back(rat);
dateVec.push_back(date);
}
// Close reader ASAP
reader.close();
// Use number of viewings to make arrays and populate noViewingsPerMovie
views = (long) usersVec.size();
noViewingsPerMovie[m] = views;
usersByMovie[m] = new user[views];
ratesByMovie[m] = new rate[views];
datesByMovie[m] = new uint[views];
whatIndexAMovieIsToAUser[m] = new ushort[views];
// Populate usersByMovie & ratesByMovie
for (int index = 0; index < views; index++) {
usersByMovie[m][index] = usersVec[index];
ratesByMovie[m][index] = rates[index];
datesByMovie[m][index] = dateVec[index];
}
}
#pragma omp barrier
// This is new stuff for getting user-perspective ratings from movie-perspective ones.
// Make vector arrays for storing the ratings user-side.
#define USERSATONCE 1000
vector<rate> urates[USERSATONCE];
vector<movie> umovies[USERSATONCE];
vector<int> udates[USERSATONCE];
#pragma omp parallel for firstprivate(urates, umovies, udates)
for (uint i = 0; i <= U / USERSATONCE; i++) {
// Limit to number of users dealt with by one iteration is either USERSATONCE or
// on last iteration 189 ( = U - i*USERSATONCE)
uint jLim = min((uint) USERSATONCE, U - i * USERSATONCE);
// Make sure vectors are empty
for (uint jj = 0; jj < jLim; jj++) {
urates[jj].clear();
umovies[jj].clear();
udates[jj].clear();
}
// Go through each movie and index & add rating and movie to the vectors
for (movie m = 0; m < M; m++) {
for (uint index = 0; index < noViewingsPerMovie[m]; index++) {
user u = usersByMovie[m][index];
if (u / USERSATONCE == i) {
urates[u % USERSATONCE].push_back(
ratesByMovie[m][index]);
umovies[u % USERSATONCE].push_back(m);
udates[u % USERSATONCE].push_back(
datesByMovie[m][index]);
whatIndexAMovieIsToAUser[m][index] = urates[u
% USERSATONCE].size() - 1;
}
}
}
// Go through each user in this range
for (uint jj = 0; jj < jLim; jj++) {
// Get user number & number of ratings
uint views = urates[jj].size();
user u = i * USERSATONCE + jj;
// Fill in user info
noViewingsPerUser[u] = views;
moviesByUser[u] = new movie[views];
ratesByUser[u] = new rate[views];
datesByUser[u] = new uint[views];
for (uint index = 0; index < views; index++) {
moviesByUser[u][index] = umovies[jj][index];
ratesByUser[u][index] = urates[jj][index];
datesByUser[u][index] = udates[jj][index];
}
}
}
#pragma omp barrier
} else {
// If in test mode we populate the user arrays with 3 random movies and ratings
srand(time(0));
rand();
int views = 3;
// Iterate over users
for (user u = 0; u < U; u++) {
noViewingsPerUser[u] = views;
moviesByUser[u] = new movie[views];
ratesByUser[u] = new rate[views];
datesByUser[u] = new uint[views];
// First rating simply random
moviesByUser[u][0] = (rand() % M);
ratesByUser[u][0] = 1 + (rand() % 5);
datesByUser[u][0] = 1 + (rand() % 2600);
// Second rating only OK if movie isn't the same
do {
moviesByUser[u][1] = (rand() % M);
} while (moviesByUser[u][1] == moviesByUser[u][0]);
ratesByUser[u][1] = 1 + (rand() % 5);
datesByUser[u][1] = 1 + (rand() % 2600);
// Third rating only OK if movie isn't the same as first 2
do {
moviesByUser[u][2] = (rand() % M);
} while (moviesByUser[u][2] == moviesByUser[u][1]
|| moviesByUser[u][2] == moviesByUser[u][0]);
ratesByUser[u][2] = 1 + (rand() % 5);
datesByUser[u][2] = 1 + (rand() % 2600);
}
// Vectors for each movie (rate, user, index)
vector<rate> rates[M];
vector<user> usersVec[M];
vector<rate> indexes[M];
vector<int> dates[M];
for (user u = 0; u < U; u++) {
for (uint index = 0; index < noViewingsPerUser[u]; index++) {
dates[moviesByUser[u][index]].push_back(datesByUser[u][index]);
rates[moviesByUser[u][index]].push_back(ratesByUser[u][index]);
usersVec[moviesByUser[u][index]].push_back(u);
indexes[moviesByUser[u][index]].push_back(index);
}
}
int length;
for (movie m = 0; m < M; m++) {
// Use length to make arrays
length = usersVec[m].size();
noViewingsPerMovie[m] = length;
usersByMovie[m] = new user[length];
ratesByMovie[m] = new rate[length];
datesByMovie[m] = new uint[length];
whatIndexAMovieIsToAUser[m] = new ushort[length];
// Populate arrays with user, rate, index
for (int i = 0; i < length; i++) {
usersByMovie[m][i] = usersVec[m][i];
ratesByMovie[m][i] = rates[m][i];
datesByMovie[m][i] = dates[m][i];
whatIndexAMovieIsToAUser[m][i] = indexes[m][i];
}
}
}
}
void NetflixDataNG::readBigRatings(unsigned short* &noViewingsPerUser,
q** &bigRatingsByUser) {
if (!test) {
ifstream f;
f.open(HcTrainingFile);
string readLine(400000, '\0');
for (user u = 0; u < U; u++) {
f.getline(&readLine[0], 399999); // f << "!" << u << "!" << noViewingsPerUser[u] <<"\n";
char* pos = &readLine[0];
pos = &pos[1];
int views = strtoul(pos, &pos, 10); // Actually this is the user number
pos = &pos[1];
views = strtoul(pos, &pos, 10);
noViewingsPerUser[u] = views;
bigRatingsByUser[u] = new q[views];
f.getline(&readLine[0], 3999999); //Comma-separated viewings
pos = &readLine[0];
for (int i = 0; i < views; i++) {
bigRatingsByUser[u][i] = strtoull(pos, &pos, 10);
pos = &pos[1];
}
}
f.close();
} else {
srand(time(0));
rand();
for (user u = 0; u < U; u++) {
noViewingsPerUser[u] = 3;
bigRatingsByUser[u] = new q[3];
movie m0, m1, m2;
m0 = (rand() % M);
do {
m1 = (rand() % M);
} while (m1 == m0);
do {
m2 = (rand() % M);
} while (m2 == m0 || m2 == m1);
bigRatingsByUser[u][0] = GetBigRating(1 + (rand() % 5), m0,
rand() % 5, rand() % 50, rand() % 7, 0, rand() % 101,
rand() % 101, rand() % 101, rand() % 101, rand() % 101);
bigRatingsByUser[u][1] = GetBigRating(1 + (rand() % 5), m1,
rand() % 5, rand() % 50, rand() % 7, 0, rand() % 101,
rand() % 101, rand() % 101, rand() % 101, rand() % 101);
bigRatingsByUser[u][2] = GetBigRating(1 + (rand() % 5), m2,
rand() % 5, rand() % 50, rand() % 7, 0, rand() % 101,
rand() % 101, rand() % 101, rand() % 101, rand() % 101);
}
}
}
void NetflixDataNG::readBigRatings(user** &usersByMovie,
ulong* &noViewingsPerMovie, rate** &ratesByMovie, q** &qsByUser,
ushort* &noViewingsPerUser, ushort ** &whatIndexAMovieIsToAUser) {
readBigRatings(noViewingsPerUser, qsByUser);
// This is new stuff for getting movie-perspective ratings from use-perspective ones.
// Make vector arrays for storing the ratings movie-side.
#define MOVIESATONCE 500
vector<user> musers[MOVIESATONCE];
vector<rate> mrates[MOVIESATONCE];
vector < ushort > mindices[MOVIESATONCE];
#pragma omp parallel for firstprivate(musers, mrates, mindices)
for (uint i = 0; i <= M / MOVIESATONCE; i++) {
// Limit to number of movies dealt with by one iteration is either MOVIESATONCE or
// on last iteration 270 ( = M - i*MOVIESATONCE)
uint jLim = min((uint) MOVIESATONCE, M - i * MOVIESATONCE);
// Make sure vectors are empty
for (uint jj = 0; jj < jLim; jj++) {
musers[jj].clear();
mrates[jj].clear();
mindices[jj].clear();
}
// Go through each user and index & add rating and user to the vectors
for (user u = 0; u < U; u++) {
for (uint index = 0; index < noViewingsPerUser[u]; index++) {
movie m = qsByUser[u][index] % MOV;
if (m / MOVIESATONCE == i) {
musers[m % MOVIESATONCE].push_back(u);
mrates[m % MOVIESATONCE].push_back(
qsByUser[u][index] % RATE);
mindices[m % MOVIESATONCE].push_back(index);
}
}
}
// Go through each movie in this range
for (uint jj = 0; jj < jLim; jj++) {
// Get movie number & number of ratings
uint views = mrates[jj].size();
movie m = i * MOVIESATONCE + jj;
// Fill in user info
noViewingsPerMovie[m] = views;
usersByMovie[m] = new user[views];
ratesByMovie[m] = new rate[views];
whatIndexAMovieIsToAUser[m] = new ushort[views];
for (uint index = 0; index < views; index++) {
usersByMovie[m][index] = musers[jj][index];
ratesByMovie[m][index] = mrates[jj][index];
whatIndexAMovieIsToAUser[m][index] = mindices[jj][index];
}
}
}
#pragma omp barrier
}
// Data utility functions
string NetflixDataNG::getBaseSaveFileName() {
stringstream s("");
s << basePathToUse;
s << pathToResults;
return s.str();
}
void NetflixDataNG::writeQualifyingFile(string qualifyingFileName,
string fullStatsFileName, model::DotProductGaussianAddNG* node) {
try {
// Open qualifying datafile and prepare both output file writers
individualReader.open(&qualDataFile[0]);
ofstream qualWriter;
ofstream fullWriter;
qualWriter.open(&qualifyingFileName[0]);
fullWriter.open(&fullStatsFileName[0]);
cerr << "\nQualifying data file: " << qualDataFile << "\n";
cerr << "Qualifying file: " << qualifyingFileName << "\n";
cerr << "Full Stats file: " << fullStatsFileName << "\n";
// Start reading, make variables for the prediction
char line[100];
individualReader.getline(line, 99);
unsigned int position;
char* pEnd;
bool movieNoLine = false;
movie movieNumber = 0;
user userNumber = 0;
double rating = 0.0;
double stats[] = { 0.0, 0.0 };
int lObservations = 0;
double lTotalMean = 0.0;
double lTotalMeanSquared = 0.0;
double lTotalBoundedMean = 0.0;
double lTotalBoundedMeanSquared = 0.0;
double lTotalPredictionVariance = 0.0;
do {
pEnd = NULL;
// Read line & check if it's a line with a movie number on it
for (position = 0; position < 100; position++) {
if (line[position] == '\0') {
movieNoLine = false;
break;
} else if (line[position] == ':') {
//If it is, record the movie number, and output the line to both output files
pEnd = &line[position];
movieNumber = convertNxMovieToHcMovie(
strtoul(line, &pEnd, 10));
movieNoLine = true;
qualWriter << line << "\n";
fullWriter << line << "\n";
break;
}
}
if (!movieNoLine) {
// On normal lines, get user and predict stats as appropriate
userNumber = convertNxUserToHcUser(strtoul(line, &pEnd, 10));
node->predictUnknownIndex(stats, userNumber, movieNumber);
// Output rating (in right range) and stats
rating = stats[0];
if (rating > 5.0) {
rating = 5.0;
} else if (rating < 1.0) {
rating = 1.0;
}
qualWriter << rating << '\n';
fullWriter << stats[0] << ',' << stats[1] << '\n';
lObservations++;
lTotalMean += stats[0];
lTotalMeanSquared += stats[0] * stats[0];
lTotalBoundedMean += rating;
lTotalBoundedMeanSquared += rating * rating;
lTotalPredictionVariance += stats[1] - stats[0] * stats[0];
}
individualReader.getline(&line[0], 99);
} while (line[0] != '\0');
// At EOF close all streams
qualWriter.close();
fullWriter.close();
individualReader.close();
// Output statistics
double lObsDouble = (double) lObservations;
lTotalMean /= lObsDouble;
lTotalMeanSquared /= lObsDouble;
lTotalBoundedMean /= lObsDouble;
lTotalBoundedMeanSquared /= lObsDouble;
lTotalPredictionVariance /= lObsDouble;
cerr << "\n\n===Prediction Statistics===\n";
cerr << "Observations = " << lObservations << "\n";
cerr << "E(predicted) = " << lTotalMean << "\n";
cerr << "Var(predicted) = "
<< (lTotalMeanSquared - lTotalMean * lTotalMean) << "\n";
cerr << "E(Var(predicted)) = " << lTotalPredictionVariance << "\n";
cerr << "E(|predicted|) = " << lTotalBoundedMean << "\n";
cerr << "Var(|predicted|) = "
<< (lTotalBoundedMeanSquared
- lTotalBoundedMean * lTotalBoundedMean) << "\n";
cerr << "===========================\n";
} catch (...) {
cerr << ("Couldn't read the prediction data-file");
}
}
void NetflixDataNG::getAllBigPredictionData(vector<user> &users,
vector<q> &preds) {
ifstream f;
f.open(HcPredictionFile);
char line[100];
f.getline(line, 99);
char* pEnd;
do {
users.push_back(strtoul(line, &pEnd, 10));
pEnd = &pEnd[1];
preds.push_back(strtoull(pEnd, &pEnd, 10));
f.getline(line, 99);
} while (line[0] != '\0' && line[1] != '\0');
f.close();
}
void NetflixDataNG::getAllPredictionData(vector<user> &users,
vector<movie> &movies, vector<uint> &dates) {
// Open qualifying datafile
individualReader.open(&qualDataFile[0]);
// Start reading, make variables for the prediction
char line[100];
individualReader.getline(line, 99);
unsigned int position;
char* pEnd;
bool movieNoLine = false;
movie movieNumber = 0;
user userNumber = 0;
do {
pEnd = NULL;
// Read line & check if it's a line with a movie number on it
for (position = 0; position < 100; position++) {
if (line[position] == '\0') {
movieNoLine = false;
break;
} else if (line[position] == ':') {
// If it is, record the movie number, and output the line to both output files
pEnd = &line[position];
movieNumber = convertNxMovieToHcMovie(strtoul(line, &pEnd, 10));
movieNoLine = true;
break;
}
}
if (!movieNoLine) {
// On normal lines, get user and predict stats as appropriate
userNumber = convertNxUserToHcUser(strtoul(line, &pEnd, 10));
pEnd = &pEnd[1];
int year = strtoul(pEnd, &pEnd, 10);
pEnd = &pEnd[1];
int month = strtoul(pEnd, &pEnd, 10);
pEnd = &pEnd[1];
int day = strtoul(pEnd, &pEnd, 10);
int date = daysAfter30Sept1998[year - 1998][month - 1] + day;
users.push_back(userNumber);
movies.push_back(movieNumber);
dates.push_back(date);
}
individualReader.getline(&line[0], 99);
} while (line[0] != '\0');
individualReader.close();
}
q NetflixDataNG::GetBigRating(rate rating, movie m, uint userEra, uint movieEra,
uint weekday, uint precisionNumber, uint useHistory1, uint useHistory3,
uint useHistory5, uint useHistory10, uint useHistory30) {
q bigRatingSoFar = 0;
q baseValue = 1;
updateBigRating(bigRatingSoFar, baseValue, rating, RATE);
updateBigRating(bigRatingSoFar, baseValue, m, MOV);
updateBigRating(bigRatingSoFar, baseValue, userEra, ERAU);
updateBigRating(bigRatingSoFar, baseValue, movieEra, ERAM);
updateBigRating(bigRatingSoFar, baseValue, weekday, WKDY);
updateBigRating(bigRatingSoFar, baseValue, precisionNumber, PREC);
updateBigRating(bigRatingSoFar, baseValue, useHistory1, HIST1);
updateBigRating(bigRatingSoFar, baseValue, useHistory3, HIST3);
updateBigRating(bigRatingSoFar, baseValue, useHistory5, HIST5);
updateBigRating(bigRatingSoFar, baseValue, useHistory10, HIST10);
updateBigRating(bigRatingSoFar, baseValue, useHistory30, HIST30);
return bigRatingSoFar;
}
void NetflixDataNG::updateBigRating(q ¤tBigRatingValue,
q ¤tBaseValue, q newValue, q newBaseValue) {
for (q i = 0; i < newBaseValue; i++) {
if ((currentBigRatingValue % newBaseValue) == newValue) {
break;
}
currentBigRatingValue += currentBaseValue;
}
if ((currentBigRatingValue % newBaseValue) != newValue) {
// We went round the whole loop without success
throw "BUGGER, check the bases are all prime";
}
currentBaseValue *= newBaseValue;
}
// Setting internal debug/test variables
void NetflixDataNG::setDebug() {
debug = true;
}
void NetflixDataNG::setTest() {
test = true;
}
// Helper functions for IO
void NetflixDataNG::getQuartetFromUserData(movie& val, rate& r,
string& fromThis, char* pEnd) {
val = convertNxMovieToHcMovie(strtoul(&fromThis[0], &pEnd, 10));
pEnd = &pEnd[1];
r = (rate) strtoul(pEnd, &pEnd, 10);
}
void NetflixDataNG::getQuartetFromMovieData(user& val, rate& r,
string& fromThis, char* pEnd) {