├── .gitattributes ├── .gitignore ├── Makefile ├── README.md ├── data ├── query ├── ssh └── ssh.cpp /.gitattributes: -------------------------------------------------------------------------------- 1 | # Auto detect text files and perform LF normalization 2 | * text=auto 3 | 4 | # Custom for Visual Studio 5 | *.cs diff=csharp 6 | 7 | # Standard to msysgit 8 | *.doc diff=astextplain 9 | *.DOC diff=astextplain 10 | *.docx diff=astextplain 11 | *.DOCX diff=astextplain 12 | *.dot diff=astextplain 13 | *.DOT diff=astextplain 14 | *.pdf diff=astextplain 15 | *.PDF diff=astextplain 16 | *.rtf diff=astextplain 17 | *.RTF diff=astextplain 18 | -------------------------------------------------------------------------------- /.gitignore: -------------------------------------------------------------------------------- 1 | # Windows image file caches 2 | Thumbs.db 3 | ehthumbs.db 4 | 5 | # Folder config file 6 | Desktop.ini 7 | 8 | # Recycle Bin used on file shares 9 | $RECYCLE.BIN/ 10 | 11 | # Windows Installer files 12 | *.cab 13 | *.msi 14 | *.msm 15 | *.msp 16 | 17 | # Windows shortcuts 18 | *.lnk 19 | 20 | # ========================= 21 | # Operating System Files 22 | # ========================= 23 | 24 | # OSX 25 | # ========================= 26 | 27 | .DS_Store 28 | .AppleDouble 29 | .LSOverride 30 | 31 | # Thumbnails 32 | ._* 33 | 34 | # Files that might appear in the root of a volume 35 | .DocumentRevisions-V100 36 | .fseventsd 37 | .Spotlight-V100 38 | .TemporaryItems 39 | .Trashes 40 | .VolumeIcon.icns 41 | 42 | # Directories potentially created on remote AFP share 43 | .AppleDB 44 | .AppleDesktop 45 | Network Trash Folder 46 | Temporary Items 47 | .apdisk 48 | -------------------------------------------------------------------------------- /Makefile: -------------------------------------------------------------------------------- 1 | OBJ = clean ssh 2 | 3 | CC = g++ 4 | 5 | all: $(OBJ) 6 | 7 | matmult: ssh.cpp 8 | $(CC) $(OPT) -o ssh ssh.c 9 | 10 | clean: 11 | /bin/rm -rf $(OBJ) core* 12 | 13 | veryclean: clean 14 | /bin/rn -f *~ -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | SSH: (Sketch, Shingle, & Hash) for Indexing Massive-Scale Time Series 2 | ============ 3 | 4 | SSH is a time series indexing scheme. It allows you to approximately indexing the time series data. For the details SSH scheme, please refer the paper bellow: 5 | 6 | Chen Luo and Anshumali Shrivastava "SSH (Sketch, Shingle, & Hash) for Indexing Massive-Scale Time Series" 7 | Published in Proceedings of Journal of Machine Learning Research 2017. 8 | 9 | Step by step Guide. 10 | ============ 11 | 1. Preliminaries: g++, linux. 12 | 13 | 2. Running the code 14 | 15 | -- cd \ 16 | 17 | -- make 18 | 19 | -- ./ssh \ \ \ \ \ \ \ \ 20 | 21 | 3. Running example for the given dataset 22 | 23 | -- cd ~/SSH 24 | 25 | -- make 26 | 27 | -- ./ssh data query 1024 2000 100 3 15 10 28 | 29 | If you have any questions or problems regarding the code, please feel free to contact Chen Luo (cl67@rice.edu) 30 | -------------------------------------------------------------------------------- /query: -------------------------------------------------------------------------------- 1 | 2,0.33863,0.35982,0.3308,0.29919,0.34894,0.36611,0.20004,-0.27379,-0.30087,-0.2465,-0.24004,-0.26625,-0.26348,-0.2464,-0.23763,-0.23,-0.22831,-0.23763,-0.25015,-0.26514,-0.28056,-0.28651,-0.2852,-0.28674,-0.28431,-0.26624,-0.23971,-0.21964,-0.21236,-0.21668,-0.22475,-0.22817,-0.22761,-0.22882,-0.22805,-0.22077,-0.21553,-0.21596,-0.216,-0.2162,-0.21827,-0.21461,-0.20082,-0.18619,-0.1732,-0.15419,-0.13368,-0.12435,-0.12545,-0.12511,-0.12498,-0.13732,-0.15478,-0.15779,-0.14906,-0.14925,-0.15906,-0.1618,-0.15772,-0.15771,-0.16071,-0.15937,-0.15315,-0.14571,-0.13845,-0.13461,-0.13587,-0.13431,-0.12178,-0.10589,-0.10085,-0.10226,-0.093864,-0.077089,-0.067358,-0.068102,-0.068112,-0.065059,-0.067439,-0.070637,-0.063781,-0.055836,-0.058328,-0.056207,-0.040222,-0.032904,-0.041539,-0.044368,-0.043572,-0.060152,-0.082067,-0.083866,-0.073165,-0.067767,-0.063677,-0.054411,-0.046483,-0.042992,-0.036072,-0.019167,0.0029932,0.019913,0.024684,0.016953,0.0047517,0.0015466,0.01279,0.029507,0.042883,0.05538,0.069446,0.079607,0.081847,0.081301,0.082924,0.082409,0.076309,0.069321,0.063079,0.054484,0.047868,0.048015,0.046997,0.039597,0.034908,0.035806,0.034981,0.036157,0.048064,0.063566,0.070128,0.069883,0.070343,0.071192,0.069433,0.062799,0.053399,0.047699,0.04663,0.048149,0.059542,0.082819,0.094353,0.07783,0.057097,0.055792,0.059077,0.050005,0.045291,0.06023,0.074115,0.063192,0.040112,0.036988,0.058839,0.07887,0.080979,0.079322,0.089739,0.1041,0.10768,0.10387,0.10344,0.10613,0.10523,0.10268,0.10577,0.11312,0.11774,0.11695,0.11225,0.10371,0.090334,0.076705,0.070511,0.07108,0.070308,0.068552,0.07179,0.072342,0.065874,0.069854,0.087314,0.092322,0.082655,0.087383,0.10375,0.10299,0.091423,0.090711,0.095089,0.09391,0.089556,0.08025,0.070191,0.07394,0.082164,0.073315,0.058263,0.056517,0.059416,0.058439,0.062817,0.070374,0.070975,0.070737,0.072055,0.063543,0.048377,0.039802,0.034351,0.026368,0.025639,0.034684,0.039624,0.034693,0.027144,0.023305,0.024175,0.023328,0.011306,-0.0059964,-0.0078904,0.0084364,0.017639,0.00071479,-0.030091,-0.047452,-0.039599,-0.019804,-0.0099067,-0.019052,-0.038124,-0.052706,-0.057569,-0.05619,-0.050801,-0.042806,-0.04053,-0.051394,-0.065817,-0.066588,-0.05393,-0.044574,-0.045477,-0.045573,-0.038328,-0.032716,-0.032932,-0.032053,-0.031247,-0.040256,-0.055735,-0.065413,-0.066952,-0.060676,-0.045746,-0.033154,-0.0394,-0.061183,-0.076856,-0.077994,-0.077941,-0.088203,-0.10439,-0.11877,-0.1281,-0.12654,-0.11592,-0.11375,-0.12737,-0.13942,-0.13769,-0.13377,-0.13619,-0.13354,-0.11868,-0.10407,-0.10072,-0.10172,-0.10021,-0.10328,-0.11131,-0.11193,-0.10282,-0.093957,-0.089511,-0.088381,-0.093659,-0.10107,-0.09917,-0.088703,-0.080373,-0.078206,-0.078373,-0.081312,-0.091013,-0.10505,-0.1168,-0.12308,-0.12826,-0.13607,-0.14353,-0.14917,-0.15592,-0.16268,-0.15983,-0.14536,-0.13491,-0.14073,-0.15629,-0.16913,-0.17702,-0.18274,-0.18422,-0.17746,-0.1625,-0.14885,-0.14807,-0.15588,-0.15949,-0.15879,-0.16034,-0.15889,-0.15047,-0.1482,-0.15643,-0.1592,-0.14818,-0.13635,-0.13796,-0.15187,-0.16715,-0.17109,-0.15995,-0.14488,-0.13698,-0.13601,-0.13619,-0.13525,-0.13649,-0.1427,-0.15052,-0.14981,-0.13686,-0.12419,-0.1268,-0.14187,-0.14905,-0.13937,-0.12474,-0.11187,-0.096922,-0.088029,-0.10048,-0.12331,-0.13036,-0.12259,-0.12106,-0.13,-0.13595,-0.13536,-0.13652,-0.138,-0.13259,-0.12537,-0.12456,-0.12431,-0.11738,-0.11167,-0.11279,-0.11118,-0.10139,-0.092665,-0.088642,-0.075565,-0.050132,-0.03314,-0.03286,-0.031391,-0.023193,-0.021654,-0.021975,-0.013066,-0.011082,-0.026212,-0.03431,-0.023549,-0.019412,-0.035407,-0.047535,-0.037738,-0.022897,-0.02393,-0.033235,-0.030519,-0.015268,-0.0017833,0.0089305,0.024361,0.036976,0.032552,0.016706,0.014238,0.034,0.054867,0.05829,0.055893,0.065983,0.080377,0.086529,0.095694,0.11655,0.13436,0.13864,0.14488,0.16428,0.18224,0.18253,0.16681,0.15462,0.16262,0.18365,0.2019,0.22086,0.25307,0.28714,0.30868,0.32834,0.34979,0.3642,0.37801,0.39659,0.40225,0.39235,0.39601,0.41241,0.41561,0.4153,0.43361,0.45859,0.47011,0.47854,0.49599,0.51163,0.52052,0.53365,0.55865,0.58645,0.60356,0.61062,0.61076,0.60569,0.59869,0.58889,0.56888,0.53423,0.50101,0.48473,0.47212,0.4427,0.40431,0.38626,0.39007,0.39305,0.39138,0.3917,0.38857,0.37359,0.35823,0.35731,0.36438,0.36921,0.36866,0.36967,0.37953,0.39346,0.40596,0.41907,0.43405,0.43811,0.42858,0.4268,0.43563,0.43398,0.4177,0.40228,0.38975,0.37444,0.36673,0.3687,0.36242,0.34463,0.33354,0.33492,0.32846,0.30841,0.29982,0.30881,0.30848,0.28698,0.26158,0.2338,0.19236,0.15532,0.14109,0.13649,0.12891,0.12221,0.1145,0.10075,0.08718,0.073925,0.056817,0.047079,0.047999,0.044874,0.036999,0.036687,0.035767,0.017915,-0.00548,-0.018082,-0.030322,-0.054568,-0.076267,-0.077047,-0.068086,-0.070425,-0.078394,-0.075164,-0.065819,-0.064688,-0.070578,-0.077365,-0.085189,-0.090591,-0.090615,-0.090433,-0.09082,-0.085847,-0.078428,-0.076907,-0.07967,-0.079497,-0.077217,-0.078705,-0.084712,-0.092499,-0.099467,-0.10382,-0.1022,-0.091257,-0.075124,-0.066296,-0.071097,-0.078253,-0.078731,-0.07896,-0.079804,-0.077768,-0.085773,-0.10999,-0.12491,-0.1175,-0.11365,-0.12287,-0.12243,-0.11306,-0.1208,-0.14332,-0.15083,-0.13736,-0.12491,-0.12376,-0.12444,-0.12369,-0.12652,-0.12544,-0.11588,-0.11396,-0.1226,-0.12258,-0.11406,-0.11303,-0.11252,-0.10387,-0.10314,-0.11177,-0.1129,-0.11352,-0.12659,-0.1397,-0.14509,-0.15605,-0.17041,-0.17302,-0.16981,-0.17055,-0.16769,-0.15952,-0.16031,-0.174,-0.18328,-0.17607,-0.16101,-0.15119,-0.14691,-0.14092,-0.13059,-0.11638,-0.10366,-0.10204,-0.10922,-0.1148,-0.12106,-0.13613,-0.14762,-0.13728,-0.12037,-0.12727,-0.15261,-0.15957,-0.13856,-0.12276,-0.13036,-0.14042,-0.13692,-0.13266,-0.13725,-0.13909,-0.13116,-0.12447,-0.13177,-0.14979,-0.16042,-0.15531,-0.14794,-0.14911,-0.14711,-0.1331,-0.12193,-0.12312,-0.12637,-0.1247,-0.11859,-0.11141,-0.11672,-0.13898,-0.14827,-0.12889,-0.11035,-0.10027,-0.07306,-0.045766,-0.052559,-0.068749,-0.067654,-0.073143,-0.096993,-0.11668,-0.13913,-0.18286,-0.21442,-0.21149,-0.18929,-0.14635,-0.057018,0.084746,0.22866,0.35057,0.46577,0.59213,0.74066,0.90515,1.1031,1.3273,1.5546,1.7911,2.0461,2.3471,2.6866,3.0449,3.3917,3.6844,3.9359,4.1638,4.3855,4.6021,4.819,5.0798,5.3813,5.69,5.9749,6.2317,6.4652,6.6197,6.6822,6.7202,6.7636,6.7586,6.6486,6.4725,6.2563,5.9955,5.7456,5.5287,5.308,5.0756,4.8688,4.7022,4.5627,4.4933,4.5208,4.5409,4.4377,4.1774,3.8249,3.4464,3.0526,2.6501,2.1828,1.5869,0.88692,0.1638,-0.5528,-1.318,-2.1115,-2.839,-3.4295,-3.8331,-4.047,-4.1236,-4.1687,-4.2486,-4.3445,-4.4391,-4.5358,-4.6263,-4.6569,-4.5902,-4.4612,-4.3145,-4.1586,-4.0012,-3.8567,-3.7329,-3.6217,-3.5008,-3.3746,-3.234,-3.0755,-2.9131,-2.7653,-2.6548,-2.5644,-2.4813,-2.4017,-2.3217,-2.2517,-2.1779,-2.0965,-2.0083,-1.9212,-1.8559,-1.8135,-1.7905,-1.7521,-1.6824,-1.6186,-1.5771,-1.5307,-1.444,-1.3309,-1.2164,-1.1089,-1.0282,-0.96217,-0.88921,-0.83076,-0.81,-0.7984,-0.76256,-0.7346,-0.73258,-0.72296,-0.69654,-0.66798,-0.64145,-0.61333,-0.59659,-0.59503,-0.58962,-0.58029,-0.57356,-0.56814,-0.56123,-0.55289,-0.548,-0.54734,-0.55046,-0.55003,-0.53855,-0.52392,-0.51554,-0.51339,-0.50626,-0.48797,-0.46704,-0.45141,-0.44241,-0.4358,-0.43172,-0.43449,-0.44154,-0.44612,-0.43938,-0.42369,-0.41184,-0.40559,-0.39676,-0.38262,-0.36984,-0.35621,-0.33515,-0.31688,-0.30985,-0.30429,-0.28892,-0.27037,-0.26179,-0.2618,-0.26105,-0.25351,-0.24346,-0.23922,-0.23469,-0.22297,-0.21562,-0.22012,-0.21668,-0.19498,-0.17902,-0.17452,-0.15897,-0.13735,-0.13253,-0.13628,-0.12988,-0.11846,-0.10466,-0.082433,-0.065828,-0.066076,-0.064958,-0.048884,-0.030984,-0.020557,-0.0082421,0.0075213,0.013936,0.0075413,0.0015454,0.006689,0.015204,0.014697,0.011957,0.02342,0.046163,0.061539,0.067092,0.077209,0.091924,0.097223,0.092174,0.090433,0.097309,0.10393,0.10839,0.11758,0.12606,0.12893,0.14336,0.17343,0.18482,0.16847,0.16334,0.17348,0.16488,0.15072,0.16577,0.19075,0.19643,0.19925,0.20775,0.20103,0.1804,0.16573,0.15847,0.15156,0.14835,0.15135,0.15733,0.16743,0.18189,0.19332,0.19659,0.19619,0.19629,0.18941,0.17266,0.16214,0.16799,0.17384,0.1599,0.13417,0.1169,0.10584,0.088151,0.072387,0.071501,0.071424,0.057538,0.046512,0.048286,0.041504,0.017841,-0.002152,-0.010881,-0.023156,-0.03928,-0.044706,-0.044098,-0.056878,-0.080871,-0.091218,-0.082304,-0.075053,-0.078623,-0.085528,-0.091851,-0.092524,-0.086199,-0.089152,-0.10216,-0.099026,-0.087259,-0.10167,-0.12373,-0.11504,-0.10123,-0.1148,-0.12774,-0.11695,-0.10194,-0.088897,-0.071788,-0.069196,-0.085681,-0.094374,-0.090175,-0.093519,-0.10547,-0.1123,-0.11364,-0.11324,-0.11167,-0.11446,-0.1217,-0.12778,-0.13503,-0.1413,-0.1333,-0.11689,-0.11333,-0.1143,-0.10488,-0.1022,-0.10702,-0.092833,-0.077319,-0.094798,-0.11327,-0.099458,-0.089985,-0.10057,-0.094315,-0.078837,-0.089579,-0.10277,-0.087333,-0.071714,-0.077829,-0.082789,-0.077418,-0.076627,-0.080779,-0.079592,-0.072572,-0.066821,-0.071256,-0.085744,-0.091729,-0.079234,-0.066736,-0.066413,-0.06675,-0.065912,-0.073601,-0.07871,-0.066665,-0.052481,-0.053043,-0.060854,-0.067001,-0.070321,-0.064639,-0.049866,-0.040871,-0.043804,-0.043416,-0.029612,-0.014196,-0.0072623,-0.00030575,0.0072916,-0.00015722,-0.013131,-0.00052793,0.030078,0.041763,0.035776,0.040129,0.049667,0.041747,0.025825,0.024369,0.029347,0.020217,0.0046,0.0010652,0.0018228,-0.002512,0.005039,0.027726,0.039536,0.036282,0.039621,0.048329,0.048421,0.048403,0.0567,0.064795,0.070973,0.079083,0.081327,0.073142,0.067068,0.070451,0.069431,0.051632,0.029513,0.021209,0.024233,0.029791,0.038147,0.040451,0.026271,0.013281,0.019079,0.023983,0.010283,-0.0042948,-0.0088738,-0.018788,-0.034317,-0.037276,-0.029592,-0.030391,-0.041038,-0.044115,-0.031998,-0.016405,-0.0096795,-0.012922,-0.021177,-0.029526,-0.033238,-0.032286,-0.037049,-0.053977,-0.068805,-0.064888,-0.054588,-0.063372,-0.086686,-0.10386,-0.1147,-0.12331,-0.12348,-0.12283,-0.1357,-0.14792,-0.142,-0.13474,-0.14307,-0.15782,-0.17067,-0.18048,-0.18055,-0.17251,-0.16939,-0.17047,-0.1729,-0.18646,-0.20256,-0.20336,-0.2036,-0.21971,-0.22712,-0.21135,-0.20527,-0.22366,-0.23869,-0.23909,-0.23877,-0.2382,-0.23021,-0.22413,-0.22817,-0.23149,-0.22358,-0.20882,-0.2009,-0.20479,-0.20838,-0.20216,-0.19389,-0.19619,-0.20446,-0.20291,-0.19035,-0.18147,-0.18496,-0.19246,-0.19799,-0.20617,-0.20963,-0.19738,-0.18295,-0.18566,-0.19365,-0.18788,-0.1797,-0.18113,-0.18045,-0.17041,-0.16063,-0.15959,-0.16655,-0.17693,-0.18198,-0.1803,-0.18365,-0.19222,-0.18983,-0.1729,-0.15888,-0.16016,-0.16799,-0.17115,-0.17054,-0.17088,-0.17018,-0.16403,-0.1554,-0.14893,-0.14357,-0.13565,-0.12271,-0.1083,-0.10126,-0.10561,-0.11305,-0.11435,-0.11279,-0.11315,-0.11141,-0.10282,-0.089426,-0.075753,-0.062,-0.049848,-0.044482,-0.044804,-0.044007,-0.042851,-0.046305,-0.046254,-0.033525,-0.020893,-0.02477,-0.03774,-0.044131,-0.043854,-0.044347,-0.045369,-0.043166,-0.036555,-0.025299,-0.010827,0.0035447,0.016008,0.028854,0.046687,0.067891,0.081789,0.080681,0.072327,0.07133,0.07981,0.088744,0.09384,0.099746,0.10868,0.11536,0.11719,0.11618,0.11086,0.10169,0.10112,0.11579,0.12771,0.11852,0.10007,0.093035,0.092967,0.083289,0.070107,0.072238,0.081824,0.076774,0.061218,0.048633,0.033358,0.020924,0.033666,0.056459,0.05645,0.047104,0.048772,0.042925,0.02688,0.026751,0.03601,0.034747,0.039451,0.055985,0.05483,0.036781,0.036141,0.050336,0.045568,0.022911,0.012799,0.017366,0.010045,-0.012856,-0.023407,-0.012452,-0.0071592,-0.023929,-0.042562,-0.040889,-0.025139,-0.016333,-0.021792,-0.030294,-0.033818,-0.040257,-0.052788,-0.05598,-0.042585,-0.031247,-0.038138,-0.052266,-0.056007,-0.053605,-0.056867,-0.05846,-0.051859,-0.053637,-0.0706,-0.078462,-0.06903,-0.071814,-0.096095,-0.11051,-0.10051,-0.089598,-0.09199,-0.092191,-0.085096,-0.090341,-0.1122,-0.12735,-0.12604,-0.12387,-0.12469,-0.12247,-0.127,-0.14192,-0.15082,-0.14823,-0.14284,-0.13409,-0.12435,-0.12946,-0.14537,-0.1475,-0.13468,-0.13073,-0.14039,-0.13979,-0.12233,-0.1132,-0.12096,-0.12363,-0.11492,-0.11432,-0.12434,-0.13519,-0.15387,-0.17646,-0.18387,-0.18181,-0.18825,-0.19329,-0.18524,-0.18217,-0.19184,-0.19628,-0.19313,-0.19782,-0.21062,-0.21652,-0.21418,-0.21701,-0.23013,-0.24306,-0.24212,-0.2317,-0.22791,-0.22995,-0.22509,-0.2165,-0.22137,-0.23825,-0.25309,-0.26497,-0.27285,-0.27258,-0.27299,-0.28104,-0.28506,-0.28257,-0.29142,-0.30667,-0.30536,-0.29615,-0.30109,-0.31077,-0.302,-0.28174,-0.27381,-0.28053,-0.28553,-0.27791,-0.26624,-0.26233,-0.25924,-0.24581,-0.22988,-0.23118,-0.24896,-0.26229,-0.26055,-0.24692,-0.23182,-0.22721,-0.2339,-0.23958,-0.23882,-0.24015,-0.24026,-0.23512,-0.2403,-0.25677,-0.26119,-0.25179,-0.25584,-0.27387,-0.27884,-0.27099,-0.27031,-0.28143,-0.29581,-0.30921,-0.32289,-0.33649,-0.34863,-0.35394,-0.35291,-0.35455,-0.35607,-0.34976,-0.34271,-0.34622,-0.3548,-0.36024,-0.37089,-0.38584,-0.38814,-0.37653,-0.37067,-0.38234,-0.39731,-0.39901,-0.38887,-0.37878,-0.37695,-0.37759,-0.37387,-0.36632,-0.35857,-0.35268,-0.34568,-0.33979,-0.34141,-0.34944,-0.35396,-0.34756,-0.33932,-0.34073,-0.34904,-0.35416,-0.35359,-0.3545,-0.35496,-0.34945,-0.3426,-0.33683,-0.32708,-0.31261,-0.30276,-0.29722,-0.28532,-0.27028,-0.26261,-0.26248,-0.26245,-0.26098,-0.26269,-0.26417,-0.25733,-0.24222,-0.23296,-0.24071,-0.25172,-0.24511,-0.22241,-0.2055,-0.20509,-0.20739,-0.2027,-0.20187,-0.21427,-0.22766,-0.2296,-0.22704,-0.2216,-0.20513,-0.18383,-0.16886,-0.15598,-0.1412,-0.13255,-0.12646,-0.11175,-0.10137,-0.11303,-0.13157,-0.12873,-0.10957,-0.10172,-0.10508,-0.095429,-0.072729,-0.062212,-0.067296,-0.065091,-0.051125,-0.043444,-0.048828,-0.055663,-0.052244,-0.042001,-0.040572,-0.049543,-0.046557,-0.024604,-0.008266,-0.0034542,0.0092967,0.023513,0.022572,0.025914,0.048207,0.060931,0.050505,0.046512,0.059418,0.064171,0.055842,0.0558,0.064525,0.059411,0.041197,0.037445,0.052848,0.062711,0.059365,0.057457,0.059049,0.058268,0.06129,0.069115,0.075211,0.083208,0.091983,0.092091,0.091447,0.10392,0.11672,0.11618,0.11893,0.13441,0.1443,0.13921,0.1347,0.14272,0.15638,0.17282,0.19506,0.2107,0.20482,0.19405,0.20661,0.22907,0.23323,0.23149,0.24422,0.26164,0.27425,0.28618,0.28752,0.27232,0.26437,0.27372,0.28062,0.27664,0.26954,0.26315,0.26234,0.27478,0.28817,0.28465,0.27448,0.27429,0.27932,0.2774,0.27323,0.27858,0.28675,0.28766,0.28758,0.29412,0.30014,0.29963,0.30043,0.30085,0.29302,0.28852,0.30062,0.31461,0.30606,0.28703,0.28839,0.3013,0.29498,0.27797,0.28334,0.29939,0.29173,0.27323,0.2724,0.28075,0.27805,0.26885,0.26501,0.26445,0.26702,0.27497,0.28078,0.27582,0.26618,0.2671,0.27505,0.27976,0.28885,0.30509,0.31033,0.29492,0.27858,0.27675,0.28258,0.29041,0.29768,0.3005,0.29976,0.30384,0.31318,0.32049,0.32412,0.32338,0.32031,0.32344,0.32521,0.32189 -------------------------------------------------------------------------------- /ssh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/rackingroll/ssh/e70e04a5a82f08db8f04dad237c5b75b154d8f75/ssh -------------------------------------------------------------------------------- /ssh.cpp: -------------------------------------------------------------------------------- 1 | /************************* DISCLAIMER **********************************/ 2 | /* 3 | This SSH code is copyright protected © 2017 by Chen Luo and Anshumali Shrivastava 4 | For the details about SSH, please refer the paper: 5 | Chen Luo and Anshumail Shrivastava, "SSH (Sketch, Shingle, & Hash) for Indexing Massive-Scale Time Series" NIPS Time Series Workshop. 6 | */ 7 | /***********************************************************************/ 8 | 9 | #include 10 | #include 11 | #include 12 | #include 13 | #include 14 | #include 15 | #include 16 | 17 | #define INF 1e20 //Pseudo Infitinte number for this code 18 | #define dist(x,y) ((x-y)*(x-y)) 19 | 20 | using namespace std; 21 | 22 | typedef struct Set 23 | { 24 | int* weights; // list of the number of the patterns happened. 25 | int** patterns; // list of the patterns. 26 | int length; // the length of the Set 27 | 28 | } Set; 29 | 30 | typedef struct HashPair 31 | { 32 | int k; // index of the sample 33 | int y; // value that between [0,x_k] 34 | } HashPair; 35 | 36 | 37 | typedef struct HashTable 38 | { 39 | std::list ** HashBuckets; // Hash Buckets in the hashtables 40 | int length; // number of buckets in the hash table 41 | } HashTable; 42 | 43 | 44 | /// If serious error happens, terminate the program. 45 | void error(int id) 46 | { 47 | if(id==1) 48 | printf("ERROR : Memory can't be allocated!!!\n\n"); 49 | else if ( id == 2 ) 50 | printf("ERROR : File not Found!!!\n\n"); 51 | else if ( id == 3 ) 52 | printf("ERROR : Can't create Output File!!!\n\n"); 53 | else if ( id == 4 ) 54 | { 55 | printf("ERROR: Invalid Number of Arguments!!!\n"); 56 | printf("Command Usage: UCR_ED.exe data_file query_file m \n"); 57 | printf("For example : UCR_ED.exe data.txt query.txt 128 \n"); 58 | } 59 | exit(1); 60 | } 61 | 62 | // Generate a random number based on the mean and variance 63 | double randn (double mu, double sigma) 64 | { 65 | double U1, U2, W, mult; 66 | static double X1, X2; 67 | static int call = 0; 68 | 69 | if (call == 1) 70 | { 71 | call = !call; 72 | return (mu + sigma * (double) X2); 73 | } 74 | 75 | do 76 | { 77 | U1 = -1 + ((double) rand () / RAND_MAX) * 2; 78 | U2 = -1 + ((double) rand () / RAND_MAX) * 2; 79 | W = pow (U1, 2) + pow (U2, 2); 80 | } 81 | while (W >= 1 || W == 0); 82 | 83 | mult = sqrt ((-2 * log (W)) / W); 84 | X1 = U1 * mult; 85 | X2 = U2 * mult; 86 | 87 | call = !call; 88 | 89 | return (mu + sigma * (double) X1); 90 | } 91 | 92 | // generate and iid random filter r 93 | void iid (int length, double * r) 94 | { 95 | for (int i=0; i=0) return 1; 110 | else return -1; 111 | 112 | } 113 | 114 | // Do n-gram on Db 115 | void ngram (int * Db, int k, int ** set, int m) 116 | { 117 | 118 | for (int i=0;ipatterns[i],d,k)) 146 | { 147 | return i; 148 | } 149 | } 150 | return -1; 151 | } 152 | 153 | //Check whether an item d is in the set list 154 | int inList (int * set, int d, int s) 155 | { 156 | for (int i=0;ipatterns = (int **) malloc(s*sizeof(int*)); 169 | for (int t=0;tpatterns[t] = (int*)malloc((k)*sizeof(int)) ; 172 | } 173 | set->weights = (int *) malloc(s*sizeof(int)) ; 174 | set->length = 0; 175 | 176 | int index = 0; 177 | 178 | for (int i=0;iweights[index] ++ ; 184 | } 185 | else 186 | { 187 | for (int j=0;jpatterns[set->length][j] = Ds[i][j] ; 190 | } 191 | 192 | set->weights[set->length] = 1; 193 | set->length ++ ; 194 | } 195 | } 196 | return set; 197 | 198 | } 199 | 200 | // Print set 201 | void print_set(Set *set) 202 | { 203 | for (int i=0;ilength;i++) 204 | { 205 | printf("%d ", set->weights[i]); 206 | } 207 | printf("\n"); 208 | } 209 | 210 | // Calculate the Jaccard Similarity between two set1 and set2 211 | double jsim (int ** set1, int ** set2, int k, int s) // k is the object length, s is the set size 212 | { 213 | 214 | Set * s1 = set_create(set1, k, s) ; 215 | Set * s2 = set_create(set2, k, s) ; 216 | 217 | double * mins = (double*)malloc((s1->length + s2->length)*sizeof(double)) ; 218 | double * maxs = (double*)malloc((s1->length + s2->length)*sizeof(double)) ; 219 | 220 | Set* set = (Set*)malloc(sizeof(Set)) ; 221 | set->patterns = (int **) malloc((s1->length + s2->length)*sizeof(int*)); 222 | 223 | for (int t=0;t<(s1->length + s2->length);t++) 224 | { 225 | set->patterns[t] = (int*)malloc((k)*sizeof(int)) ; 226 | } 227 | set->weights = (int *) malloc((s1->length + s2->length)*sizeof(int)) ; 228 | set->length = 0; 229 | 230 | // Add s2 into set list 231 | for (int i=0;ilength;i++) 232 | { 233 | for (int j=0;jpatterns[i][j] = s2->patterns[i][j] ; 236 | } 237 | // Give them min value. 238 | maxs[i] = s2->weights[i] ; 239 | set->length ++ ; 240 | } 241 | 242 | // Add s1 into set list 243 | int index = 0; 244 | for (int i=0;ilength;i++) 245 | { 246 | index = inSet(set, s1->patterns[i], k, (s1->length + s2->length)); 247 | if ( index!= -1) 248 | { 249 | if (s1->weights[index] < maxs[index]) mins[index]=s1->weights[index] ; 250 | else 251 | { 252 | mins[index] = maxs[index] ; 253 | maxs[index] = s1->weights[index]; 254 | } 255 | } 256 | else 257 | { 258 | for (int j=0;jpatterns[set->length][j] = s1->patterns[i][j] ; 261 | } 262 | 263 | maxs[set->length] = s1->weights[i]; 264 | set->length ++ ; 265 | } 266 | } 267 | 268 | double max = 0.0 ; 269 | double min = 0.0 ; 270 | 271 | for (int i=0;ilength;i++) 272 | { 273 | max += maxs[i] ; 274 | min += mins[i] ; 275 | } 276 | 277 | return min/max; 278 | } 279 | 280 | // generate a random number from a uniform distribution 281 | double uniform(double a, double b) 282 | { 283 | return rand() / (RAND_MAX + 1.0) * (b - a) + a; 284 | } 285 | 286 | // Generate a random number from a gaussion distribution 287 | double gauss(double mu,double sigma) 288 | { 289 | double x1, x2, w, y1, y2; 290 | 291 | do { 292 | x1 = 2.0 * uniform(0,1) - 1.0; 293 | x2 = 2.0 * uniform(0,1) - 1.0; 294 | w = x1 * x1 + x2 * x2; 295 | } while ( w >= 1.0 ); 296 | 297 | w = sqrt( (-2.0 * log( w ) ) / w ); 298 | y1 = x1 * w; 299 | y2 = x2 * w; 300 | return mu+sigma*y1; 301 | } 302 | 303 | // Generate a random number from a gamma distribution 304 | double rgamma(double a,double b) 305 | { 306 | double d,c,x,v,u; 307 | if(a>=1) 308 | { 309 | d = a-1./3.; 310 | c = 1./sqrt(9.*d); 311 | while(1) 312 | { 313 | do 314 | { 315 | x=gauss(0,1.0); 316 | v=1.+c*x; 317 | } 318 | while(v<=0.); 319 | 320 | v = v * v * v; 321 | u = uniform(0,1); 322 | if( u < 1.0-0.0331*(x*x)*(x*x) ) 323 | { 324 | return d*v*b; 325 | } 326 | if( log(u) < 0.5*x*x+d*(1.0-v+log(v)) ) 327 | { 328 | return d*v*b; 329 | } 330 | } 331 | } else 332 | { 333 | x = rgamma(a+1,b); 334 | x = x * pow(uniform(0,1), 1.0/a); 335 | return x; 336 | } 337 | } 338 | 339 | // Calculate argmin 340 | int argmin (double * a, int n) 341 | { 342 | int k = 0; 343 | double min = a[0]; 344 | for (int i=0;i a[i]) k = i; 347 | } 348 | return k; 349 | } 350 | 351 | /* 352 | The function CWS is Consistent weighted sampling, for the details of this function, please refer the following paper: 353 | Ioffe, Sergey. "Improved consistent sampling, weighted minhash and l1 sketching." Data Mining (ICDM), 2010 IEEE 10th International Conference on. IEEE, 2010. 354 | */ 355 | int CWS (Set * set) 356 | { 357 | 358 | double *r = (double*)malloc(set->length*sizeof(double)) ; 359 | double *c = (double*)malloc(set->length*sizeof(double)) ; 360 | double *b = (double*)malloc(set->length*sizeof(double)) ; 361 | 362 | double *t = (double*)malloc(set->length*sizeof(double)) ; 363 | double *y = (double*)malloc(set->length*sizeof(double)) ; 364 | double *a = (double*)malloc(set->length*sizeof(double)) ; 365 | 366 | HashPair * pair = (HashPair*)malloc(sizeof(HashPair)) ; 367 | 368 | 369 | //print_set(set); 370 | 371 | int i; 372 | for (int i=0;ilength;i++) 373 | { 374 | r[i] = rgamma(2.0,1.0); 375 | c[i] = rgamma(2.0,1.0); 376 | b[i] = uniform(0.0,1); 377 | } 378 | i = 0; 379 | while(set->weights[i] > 0 && ilength) 380 | { 381 | double s = (double) set->weights[i]; 382 | t[i] = (log(s)/r[i]) + b[i] ; 383 | 384 | y[i] = exp(r[i]*(t[i]-b[i])); 385 | a[i] = c[i]/(y[i]*exp(r[i])) ; 386 | 387 | i++; 388 | } 389 | 390 | pair->k = argmin(a, set->length); 391 | pair->y = (int) y[pair->k]; 392 | 393 | return pair->k; 394 | } 395 | 396 | /// Calculate Dynamic Time Wrapping distance 397 | /// A,B: data and query, respectively 398 | /// cb : cummulative bound used for early abandoning 399 | /// r : size of Sakoe-Chiba warpping band 400 | double dtw(double* A, double* B, double *cb, int m, int r, double bsf = INF) 401 | { 402 | 403 | double *cost; 404 | double *cost_prev; 405 | double *cost_tmp; 406 | int i,j,k; 407 | double x,y,z,min_cost; 408 | 409 | /// Instead of using matrix of size O(m^2) or O(mr), we will reuse two array of size O(r). 410 | cost = (double*)malloc(sizeof(double)*(2*r+1)); 411 | for(k=0; k<2*r+1; k++) cost[k]=INF; 412 | 413 | cost_prev = (double*)malloc(sizeof(double)*(2*r+1)); 414 | for(k=0; k<2*r+1; k++) cost_prev[k]=INF; 415 | 416 | for (i=0; i2*r)) x = INF; 434 | else x = cost_prev[k+1]; 435 | if ((i-1<0)||(j-1<0)) z = INF; 436 | else z = cost_prev[k]; 437 | 438 | /// Classic DTW calculation 439 | cost[k] = min( min( x, y) , z) + dist(A[i],B[j]); 440 | 441 | /// Find minimum cost in row for early abandoning (possibly to use column instead of row). 442 | if (cost[k] < min_cost) 443 | { min_cost = cost[k]; 444 | } 445 | } 446 | 447 | /// We can abandon early if the current cummulative distace with lower bound together are larger than bsf 448 | if (i+r < m-1 && min_cost + cb[i+r+1] >= bsf) 449 | { free(cost); 450 | free(cost_prev); 451 | return min_cost + cb[i+r+1]; 452 | } 453 | 454 | /// Move current array to previous array. 455 | cost_tmp = cost; 456 | cost = cost_prev; 457 | cost_prev = cost_tmp; 458 | } 459 | k--; 460 | 461 | /// the DTW distance is in the last cell in the matrix of size O(m^2) or at the middle of our array. 462 | double final_dtw = cost_prev[k]; 463 | free(cost); 464 | free(cost_prev); 465 | return final_dtw; 466 | } 467 | 468 | int main(int argc , char *argv[]) 469 | { 470 | FILE *fp_data; // the input file pointer 471 | FILE *fp_query; 472 | FILE *fp_result; 473 | int m; // the length of the time series. 1639 in current data set 474 | int n; // the number of time series 475 | int w; 476 | int delta; 477 | int k; 478 | int scband = 10; 479 | int hash_num = 20; 480 | 481 | char d[20000]; 482 | 483 | if (argc<=2) 484 | error(4); 485 | 486 | fp_data = fopen(argv[1],"r"); 487 | if( fp_data == NULL ) 488 | exit(2); 489 | 490 | fp_query = fopen(argv[2],"r"); 491 | if( fp_query == NULL ) 492 | exit(2); 493 | 494 | m = atol(argv[3]); 495 | n = atol(argv[4]); 496 | w = atol(argv[5]); 497 | delta = atol(argv[6]); 498 | k = atol(argv[7]); 499 | scband = atol(argv[8]); 500 | 501 | int i = 0 ; 502 | int j = 0 ; 503 | 504 | 505 | // Extract the time series data set 506 | double ** D = (double**)malloc(n*sizeof(double*)) ; // Initialize Datas 507 | for (int k=0;k **)malloc(bucket_num*sizeof(list*)); 591 | for (j=0;j[n](); 594 | } 595 | } 596 | 597 | // Insert indexes into Hash tables 598 | for (i=2;ipush_back(i); 604 | } 605 | } 606 | 607 | printf("Weighted Minwise Hashing Finished, Hashed into %d Hashtables!\n", hash_num); 608 | 609 | // Here we do time series query. 610 | 611 | printf("Start Query ......\n"); 612 | // Extract the query time series 613 | double * Q = (double*)malloc(m*sizeof(double)) ; // Initialize Datas 614 | j = 0; 615 | while(fscanf(fp_query,"%[^\n]\n",d) != EOF && j< n) 616 | { 617 | 618 | char *p = strtok(d,","); 619 | i = 0; 620 | p = strtok (NULL, ","); 621 | while (p != NULL && i::iterator it; 664 | for(it = tables[j].HashBuckets[index]->begin();it!=tables[j].HashBuckets[index]->end();it++) 665 | { 666 | if (inList(search_s, *it, search_n) == -1) 667 | { 668 | search_s[search_n++] = (*it); 669 | } 670 | } 671 | } 672 | 673 | // Search time series using Branch and Bound (lb_keogh) 674 | //int scband = 10; 675 | 676 | double best = INF ; 677 | int best_index = 0; 678 | for (int i=0;i0?(j-scband):0; t< j+scband;t++) 689 | { 690 | if (U[j] < Q[t]) U[j] = Q[t]; 691 | if (L[j] > Q[t]) L[j] = Q[t]; 692 | } 693 | double c = D[search_s[i]][j] ; 694 | if (c > U[j]) lb_keogh += (c-U[j])*(c-U[j]) ; 695 | else if( c < L[j]) lb_keogh += (c-L[j])*(c-L[j]) ; 696 | } 697 | lb_keogh = sqrt(lb_keogh); 698 | 699 | double dist = 0.0; 700 | if (lb_keogh < best) 701 | { 702 | dist = dtw(D[search_s[i]], Q, Q, m, scband); 703 | if (dist < best) 704 | { 705 | best_index = search_s[i]; 706 | best = dist; 707 | } 708 | } 709 | } 710 | clock_t end_time = clock(); 711 | double time_spent = (double)(end_time - begin_time) / CLOCKS_PER_SEC; 712 | 713 | printf("Query Finished!\n"); 714 | printf("The nereast neighbor time series is: %d, Query time: %f seconds\n", best_index, time_spent); 715 | 716 | fclose(fp_data); 717 | fclose(fp_query); 718 | 719 | return 0; 720 | } 721 | --------------------------------------------------------------------------------