├── .gitignore ├── README.md ├── data ├── sample_data01.csv ├── sample_data02.csv ├── sample_data03.csv ├── sample_data04.csv ├── sample_data05.csv ├── sample_data06.csv ├── sample_data07.csv ├── sample_data08.csv ├── sample_data09.csv └── sample_data10.csv └── notebook └── clustering_waveform_using_kshape.ipynb /.gitignore: -------------------------------------------------------------------------------- 1 | # Byte-compiled / optimized / DLL files 2 | __pycache__/ 3 | *.py[cod] 4 | *$py.class 5 | .DS_Store 6 | 7 | # C extensions 8 | *.so 9 | 10 | # Distribution / packaging 11 | .Python 12 | build/ 13 | develop-eggs/ 14 | dist/ 15 | downloads/ 16 | eggs/ 17 | .eggs/ 18 | lib/ 19 | lib64/ 20 | parts/ 21 | sdist/ 22 | var/ 23 | wheels/ 24 | pip-wheel-metadata/ 25 | share/python-wheels/ 26 | *.egg-info/ 27 | .installed.cfg 28 | *.egg 29 | MANIFEST 30 | 31 | # PyInstaller 32 | # Usually these files are written by a python script from a template 33 | # before PyInstaller builds the exe, so as to inject date/other infos into it. 34 | *.manifest 35 | *.spec 36 | 37 | # Installer logs 38 | pip-log.txt 39 | pip-delete-this-directory.txt 40 | 41 | # Unit test / coverage reports 42 | htmlcov/ 43 | .tox/ 44 | .nox/ 45 | .coverage 46 | .coverage.* 47 | .cache 48 | nosetests.xml 49 | coverage.xml 50 | *.cover 51 | *.py,cover 52 | .hypothesis/ 53 | .pytest_cache/ 54 | 55 | # Translations 56 | *.mo 57 | *.pot 58 | 59 | # Django stuff: 60 | *.log 61 | local_settings.py 62 | db.sqlite3 63 | db.sqlite3-journal 64 | 65 | # Flask stuff: 66 | instance/ 67 | .webassets-cache 68 | 69 | # Scrapy stuff: 70 | .scrapy 71 | 72 | # Sphinx documentation 73 | docs/_build/ 74 | 75 | # PyBuilder 76 | target/ 77 | 78 | # Jupyter Notebook 79 | .ipynb_checkpoints 80 | 81 | # IPython 82 | profile_default/ 83 | ipython_config.py 84 | 85 | # pyenv 86 | .python-version 87 | 88 | # pipenv 89 | # According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control. 90 | # However, in case of collaboration, if having platform-specific dependencies or dependencies 91 | # having no cross-platform support, pipenv may install dependencies that don't work, or not 92 | # install all needed dependencies. 93 | #Pipfile.lock 94 | 95 | # PEP 582; used by e.g. github.com/David-OConnor/pyflow 96 | __pypackages__/ 97 | 98 | # Celery stuff 99 | celerybeat-schedule 100 | celerybeat.pid 101 | 102 | # SageMath parsed files 103 | *.sage.py 104 | 105 | # Environments 106 | .env 107 | .venv 108 | env/ 109 | venv/ 110 | ENV/ 111 | env.bak/ 112 | venv.bak/ 113 | 114 | # Spyder project settings 115 | .spyderproject 116 | .spyproject 117 | 118 | # Rope project settings 119 | .ropeproject 120 | 121 | # mkdocs documentation 122 | /site 123 | 124 | # mypy 125 | .mypy_cache/ 126 | .dmypy.json 127 | dmypy.json 128 | 129 | # Pyre type checker 130 | .pyre/ 131 | 132 | # dataset 133 | data/*/*.csv 134 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # sample codes using tslearn. 2 | - Purpose: clustering for waveform data or time series data.
3 | - tslearn is one of the Machine Learning libraries based on python.
4 | - tslearn: https://github.com/rtavenar/tslearn 5 | 6 | # description 7 | - Waveform clustering is performed on the sample data using the KShape algorithm. 8 | - The number of clusters must be given as an argument to the algorithm. 9 | - In this case, we set `n_clusters=2` since we know that there are two classes after checking the data in advance. 10 | - There are several ways to check the number of clusters, but in this case the elbow method was used. 11 | - Other possible methods are 12 | - BIC・AIC 13 | - GAP method 14 | - Silhouette method 15 | - Elbow method 16 | 17 | # directory 18 | - `data/`: sample dataset for clustering waveform 19 | - `notebook/`: jupyter notebook implementing tslearn sample code -------------------------------------------------------------------------------- /data/sample_data01.csv: -------------------------------------------------------------------------------- 1 | time,data 2 | 2018/01/01 13:00:00,0.99364 3 | 2018/01/01 13:00:01,2.2738 4 | 2018/01/01 13:00:02,2.9139 5 | 2018/01/01 13:00:03,2.2738 6 | 2018/01/01 13:00:04,1.6337 7 | 2018/01/01 13:00:05,0.14021 8 | 2018/01/01 13:00:06,-0.71322 9 | 2018/01/01 13:00:07,-1.1399 10 | 2018/01/01 13:00:08,-0.92658 11 | 2018/01/01 13:00:09,-0.49987 12 | 2018/01/01 13:00:10,-0.92658 13 | 2018/01/01 13:00:11,-1.5667 14 | 2018/01/01 13:00:12,-1.78 15 | 2018/01/01 13:00:13,-1.9934 16 | 2018/01/01 13:00:14,-2.2067 17 | 2018/01/01 13:00:15,-1.78 18 | 2018/01/01 13:00:16,-1.3533 19 | 2018/01/01 13:00:17,-0.71322 20 | 2018/01/01 13:00:18,-0.49987 21 | 2018/01/01 13:00:19,-0.28651 22 | 2018/01/01 13:00:20,0.35356 23 | 2018/01/01 13:00:21,0.35356 24 | 2018/01/01 13:00:22,0.78028 25 | 2018/01/01 13:00:23,0.78028 26 | 2018/01/01 13:00:24,0.78028 27 | 2018/01/01 13:00:25,0.56692 28 | 2018/01/01 13:00:26,0.56692 29 | 2018/01/01 13:00:27,0.35356 30 | 2018/01/01 13:00:28,0.56692 31 | 2018/01/01 13:00:29,0.56692 32 | 2018/01/01 13:00:30,0.56692 33 | 2018/01/01 13:00:31,0.35356 34 | 2018/01/01 13:00:32,0.35356 35 | 2018/01/01 13:00:33,0.14021 36 | 2018/01/01 13:00:34,0.14021 37 | 2018/01/01 13:00:35,0.35356 38 | 2018/01/01 13:00:36,0.14021 39 | 2018/01/01 13:00:37,0.35356 40 | 2018/01/01 13:00:38,0.56692 41 | 2018/01/01 13:00:39,0.78028 42 | 2018/01/01 13:00:40,0.78028 43 | 2018/01/01 13:00:41,1.207 44 | 2018/01/01 13:00:42,1.4204 45 | 2018/01/01 13:00:43,1.4204 46 | 2018/01/01 13:00:44,0.35356 47 | 2018/01/01 13:00:45,-0.28651 48 | 2018/01/01 13:00:46,-0.71322 49 | 2018/01/01 13:00:47,-0.92658 50 | 2018/01/01 13:00:48,-0.92658 51 | 2018/01/01 13:00:49,-0.71322 52 | 2018/01/01 13:00:50,-0.71322 53 | 2018/01/01 13:00:51,-1.1399 54 | 2018/01/01 13:00:52,-1.5667 55 | 2018/01/01 13:00:53,-1.3533 56 | 2018/01/01 13:00:54,-1.1399 57 | 2018/01/01 13:00:55,-0.49987 58 | 2018/01/01 13:00:56,0.14021 59 | 2018/01/01 13:00:57,0.14021 60 | 2018/01/01 13:00:58,0.14021 61 | 2018/01/01 13:00:59,0.14021 62 | 2018/01/01 13:01:00,0.14021 63 | 2018/01/01 13:01:01,0.35356 64 | 2018/01/01 13:01:02,0.35356 65 | 2018/01/01 13:01:03,0.14021 66 | 2018/01/01 13:01:04,-0.073151 67 | 2018/01/01 13:01:05,-0.073151 68 | 2018/01/01 13:01:06,-0.073151 69 | 2018/01/01 13:01:07,0.14021 70 | 2018/01/01 13:01:08,0.14021 71 | 2018/01/01 13:01:09,-0.073151 -------------------------------------------------------------------------------- /data/sample_data02.csv: -------------------------------------------------------------------------------- 1 | time,data 2 | 2018/01/01 14:00:00,-0.060775 3 | 2018/01/01 14:00:01,-0.34439 4 | 2018/01/01 14:00:02,1.9246 5 | 2018/01/01 14:00:03,3.6263 6 | 2018/01/01 14:00:04,1.9246 7 | 2018/01/01 14:00:05,0.50646 8 | 2018/01/01 14:00:06,-0.62801 9 | 2018/01/01 14:00:07,-1.1953 10 | 2018/01/01 14:00:08,-0.91163 11 | 2018/01/01 14:00:09,0.22284 12 | 2018/01/01 14:00:10,1.3573 13 | 2018/01/01 14:00:11,0.50646 14 | 2018/01/01 14:00:12,-1.1953 15 | 2018/01/01 14:00:13,-2.3297 16 | 2018/01/01 14:00:14,-2.0461 17 | 2018/01/01 14:00:15,-0.91163 18 | 2018/01/01 14:00:16,-0.34439 19 | 2018/01/01 14:00:17,-0.34439 20 | 2018/01/01 14:00:18,-0.91163 21 | 2018/01/01 14:00:19,-0.91163 22 | 2018/01/01 14:00:20,-0.34439 23 | 2018/01/01 14:00:21,0.22284 24 | 2018/01/01 14:00:22,0.22284 25 | 2018/01/01 14:00:23,0.50646 26 | 2018/01/01 14:00:24,0.79008 27 | 2018/01/01 14:00:25,0.79008 28 | 2018/01/01 14:00:26,0.79008 29 | 2018/01/01 14:00:27,0.79008 30 | 2018/01/01 14:00:28,0.79008 31 | 2018/01/01 14:00:29,0.79008 32 | 2018/01/01 14:00:30,1.0737 33 | 2018/01/01 14:00:31,0.79008 34 | 2018/01/01 14:00:32,0.50646 35 | 2018/01/01 14:00:33,0.22284 36 | 2018/01/01 14:00:34,0.22284 37 | 2018/01/01 14:00:35,0.22284 38 | 2018/01/01 14:00:36,0.22284 39 | 2018/01/01 14:00:37,-0.060775 40 | 2018/01/01 14:00:38,0.79008 41 | 2018/01/01 14:00:39,1.0737 42 | 2018/01/01 14:00:40,0.79008 43 | 2018/01/01 14:00:41,0.50646 44 | 2018/01/01 14:00:42,0.22284 45 | 2018/01/01 14:00:43,0.22284 46 | 2018/01/01 14:00:44,-0.34439 47 | 2018/01/01 14:00:45,-1.1953 48 | 2018/01/01 14:00:46,-2.0461 49 | 2018/01/01 14:00:47,-1.4789 50 | 2018/01/01 14:00:48,-0.34439 51 | 2018/01/01 14:00:49,0.50646 52 | 2018/01/01 14:00:50,0.22284 53 | 2018/01/01 14:00:51,-1.1953 54 | 2018/01/01 14:00:52,-2.0461 55 | 2018/01/01 14:00:53,-1.7625 56 | 2018/01/01 14:00:54,-1.1953 57 | 2018/01/01 14:00:55,-0.34439 58 | 2018/01/01 14:00:56,0.50646 59 | 2018/01/01 14:00:57,0.79008 60 | 2018/01/01 14:00:58,0.50646 61 | 2018/01/01 14:00:59,0.22284 62 | 2018/01/01 14:01:00,-0.060775 63 | 2018/01/01 14:01:01,-0.060775 64 | 2018/01/01 14:01:02,-0.060775 65 | 2018/01/01 14:01:03,-0.34439 66 | 2018/01/01 14:01:04,-0.34439 67 | 2018/01/01 14:01:05,-0.060775 68 | 2018/01/01 14:01:06,0.22284 69 | 2018/01/01 14:01:07,0.22284 70 | 2018/01/01 14:01:08,-0.060775 71 | 2018/01/01 14:01:09,-0.34439 72 | 2018/01/01 14:01:10,-0.34439 -------------------------------------------------------------------------------- /data/sample_data03.csv: -------------------------------------------------------------------------------- 1 | time,data 2 | 2018/01/01 15:00:00,0.61165 3 | 2018/01/01 15:00:01,1.3174 4 | 2018/01/01 15:00:02,2.0231 5 | 2018/01/01 15:00:03,2.2584 6 | 2018/01/01 15:00:04,2.0231 7 | 2018/01/01 15:00:05,1.7879 8 | 2018/01/01 15:00:06,0.3764 9 | 2018/01/01 15:00:07,-0.79984 10 | 2018/01/01 15:00:08,-1.5056 11 | 2018/01/01 15:00:09,-1.9761 12 | 2018/01/01 15:00:10,-1.2703 13 | 2018/01/01 15:00:11,-0.79984 14 | 2018/01/01 15:00:12,-1.2703 15 | 2018/01/01 15:00:13,-1.5056 16 | 2018/01/01 15:00:14,-1.9761 17 | 2018/01/01 15:00:15,-1.5056 18 | 2018/01/01 15:00:16,-1.0351 19 | 2018/01/01 15:00:17,-0.79984 20 | 2018/01/01 15:00:18,-0.32935 21 | 2018/01/01 15:00:19,-0.32935 22 | 2018/01/01 15:00:20,0.14115 23 | 2018/01/01 15:00:21,0.3764 24 | 2018/01/01 15:00:22,0.61165 25 | 2018/01/01 15:00:23,0.84689 26 | 2018/01/01 15:00:24,0.61165 27 | 2018/01/01 15:00:25,0.61165 28 | 2018/01/01 15:00:26,0.3764 29 | 2018/01/01 15:00:27,0.3764 30 | 2018/01/01 15:00:28,0.14115 31 | 2018/01/01 15:00:29,0.14115 32 | 2018/01/01 15:00:30,0.14115 33 | 2018/01/01 15:00:31,0.14115 34 | 2018/01/01 15:00:32,0.14115 35 | 2018/01/01 15:00:33,0.3764 36 | 2018/01/01 15:00:34,0.14115 37 | 2018/01/01 15:00:35,0.14115 38 | 2018/01/01 15:00:36,0.14115 39 | 2018/01/01 15:00:37,0.61165 40 | 2018/01/01 15:00:38,1.0821 41 | 2018/01/01 15:00:39,1.3174 42 | 2018/01/01 15:00:40,0.3764 43 | 2018/01/01 15:00:41,0.3764 44 | 2018/01/01 15:00:42,1.5526 45 | 2018/01/01 15:00:43,2.0231 46 | 2018/01/01 15:00:44,1.5526 47 | 2018/01/01 15:00:45,0.14115 48 | 2018/01/01 15:00:46,-0.5646 49 | 2018/01/01 15:00:47,-1.2703 50 | 2018/01/01 15:00:48,-1.5056 51 | 2018/01/01 15:00:49,-0.79984 52 | 2018/01/01 15:00:50,-0.094099 53 | 2018/01/01 15:00:51,-0.094099 54 | 2018/01/01 15:00:52,-0.79984 55 | 2018/01/01 15:00:53,-1.5056 56 | 2018/01/01 15:00:54,-1.7408 57 | 2018/01/01 15:00:55,-1.0351 58 | 2018/01/01 15:00:56,-0.094099 59 | 2018/01/01 15:00:57,0.61165 60 | 2018/01/01 15:00:58,0.61165 61 | 2018/01/01 15:00:59,0.3764 62 | 2018/01/01 15:01:00,0.3764 63 | 2018/01/01 15:01:01,0.14115 64 | 2018/01/01 15:01:02,0.14115 65 | 2018/01/01 15:01:03,-0.094099 66 | 2018/01/01 15:01:04,-0.32935 67 | 2018/01/01 15:01:05,-0.5646 68 | 2018/01/01 15:01:06,-0.5646 69 | 2018/01/01 15:01:07,-0.5646 70 | 2018/01/01 15:01:08,-0.32935 71 | 2018/01/01 15:01:09,-0.094099 -------------------------------------------------------------------------------- /data/sample_data04.csv: -------------------------------------------------------------------------------- 1 | time,data 2 | 2018/01/01 16:00:00,0.6445 3 | 2018/01/01 16:00:01,1.2038 4 | 2018/01/01 16:00:02,1.9495 5 | 2018/01/01 16:00:03,2.1359 6 | 2018/01/01 16:00:04,2.1359 7 | 2018/01/01 16:00:05,1.3902 8 | 2018/01/01 16:00:06,0.45807 9 | 2018/01/01 16:00:07,-0.66048 10 | 2018/01/01 16:00:08,-1.2198 11 | 2018/01/01 16:00:09,-1.4062 12 | 2018/01/01 16:00:10,-1.2198 13 | 2018/01/01 16:00:11,-0.8469 14 | 2018/01/01 16:00:12,-0.8469 15 | 2018/01/01 16:00:13,-1.4062 16 | 2018/01/01 16:00:14,-1.779 17 | 2018/01/01 16:00:15,-1.5926 18 | 2018/01/01 16:00:16,-1.2198 19 | 2018/01/01 16:00:17,-0.8469 20 | 2018/01/01 16:00:18,-0.47405 21 | 2018/01/01 16:00:19,-0.47405 22 | 2018/01/01 16:00:20,-0.28763 23 | 2018/01/01 16:00:21,0.085223 24 | 2018/01/01 16:00:22,0.27165 25 | 2018/01/01 16:00:23,0.45807 26 | 2018/01/01 16:00:24,0.45807 27 | 2018/01/01 16:00:25,0.45807 28 | 2018/01/01 16:00:26,0.45807 29 | 2018/01/01 16:00:27,0.45807 30 | 2018/01/01 16:00:28,0.27165 31 | 2018/01/01 16:00:29,0.085223 32 | 2018/01/01 16:00:30,0.085223 33 | 2018/01/01 16:00:31,0.085223 34 | 2018/01/01 16:00:32,0.085223 35 | 2018/01/01 16:00:33,0.27165 36 | 2018/01/01 16:00:34,0.085223 37 | 2018/01/01 16:00:35,0.085223 38 | 2018/01/01 16:00:36,0.085223 39 | 2018/01/01 16:00:37,0.6445 40 | 2018/01/01 16:00:38,1.0173 41 | 2018/01/01 16:00:39,1.3902 42 | 2018/01/01 16:00:40,1.2038 43 | 2018/01/01 16:00:41,1.2038 44 | 2018/01/01 16:00:42,1.763 45 | 2018/01/01 16:00:43,2.1359 46 | 2018/01/01 16:00:44,1.763 47 | 2018/01/01 16:00:45,0.6445 48 | 2018/01/01 16:00:46,-0.1012 49 | 2018/01/01 16:00:47,-0.66048 50 | 2018/01/01 16:00:48,-1.2198 51 | 2018/01/01 16:00:49,-1.2198 52 | 2018/01/01 16:00:50,-0.66048 53 | 2018/01/01 16:00:51,-0.66048 54 | 2018/01/01 16:00:52,-1.4062 55 | 2018/01/01 16:00:53,-1.9655 56 | 2018/01/01 16:00:54,-1.779 57 | 2018/01/01 16:00:55,-1.4062 58 | 2018/01/01 16:00:56,-0.47405 59 | 2018/01/01 16:00:57,0.085223 60 | 2018/01/01 16:00:58,0.45807 61 | 2018/01/01 16:00:59,0.27165 62 | 2018/01/01 16:01:00,-0.28763 63 | 2018/01/01 16:01:01,-0.1012 64 | 2018/01/01 16:01:02,0.085223 65 | 2018/01/01 16:01:03,0.27165 66 | 2018/01/01 16:01:04,0.085223 67 | 2018/01/01 16:01:05,-0.1012 68 | 2018/01/01 16:01:06,-0.1012 69 | 2018/01/01 16:01:07,-0.1012 70 | 2018/01/01 16:01:08,-0.1012 71 | 2018/01/01 16:01:09,-0.1012 -------------------------------------------------------------------------------- /data/sample_data05.csv: -------------------------------------------------------------------------------- 1 | time,data 2 | 2018/01/01 17:00:00,0.8053 3 | 2018/01/01 17:00:01,1.4248 4 | 2018/01/01 17:00:02,1.2183 5 | 2018/01/01 17:00:03,1.8377 6 | 2018/01/01 17:00:04,1.4248 7 | 2018/01/01 17:00:05,0.59882 8 | 2018/01/01 17:00:06,-0.64011 9 | 2018/01/01 17:00:07,-0.8466 10 | 2018/01/01 17:00:08,-0.8466 11 | 2018/01/01 17:00:09,-0.43362 12 | 2018/01/01 17:00:10,-0.43362 13 | 2018/01/01 17:00:11,-1.2596 14 | 2018/01/01 17:00:12,-1.879 15 | 2018/01/01 17:00:13,-2.292 16 | 2018/01/01 17:00:14,-2.0855 17 | 2018/01/01 17:00:15,-1.0531 18 | 2018/01/01 17:00:16,-0.64011 19 | 2018/01/01 17:00:17,-0.8466 20 | 2018/01/01 17:00:18,-0.8466 21 | 2018/01/01 17:00:19,-0.22714 22 | 2018/01/01 17:00:20,0.39233 23 | 2018/01/01 17:00:21,0.8053 24 | 2018/01/01 17:00:22,0.39233 25 | 2018/01/01 17:00:23,0.18584 26 | 2018/01/01 17:00:24,-0.020649 27 | 2018/01/01 17:00:25,-0.020649 28 | 2018/01/01 17:00:26,0.18584 29 | 2018/01/01 17:00:27,0.39233 30 | 2018/01/01 17:00:28,0.39233 31 | 2018/01/01 17:00:29,-0.020649 32 | 2018/01/01 17:00:30,-0.020649 33 | 2018/01/01 17:00:31,-0.020649 34 | 2018/01/01 17:00:32,-0.020649 35 | 2018/01/01 17:00:33,0.39233 36 | 2018/01/01 17:00:34,0.39233 37 | 2018/01/01 17:00:35,0.39233 38 | 2018/01/01 17:00:36,0.59882 39 | 2018/01/01 17:00:37,0.8053 40 | 2018/01/01 17:00:38,0.59882 41 | 2018/01/01 17:00:39,0.59882 42 | 2018/01/01 17:00:40,0.18584 43 | 2018/01/01 17:00:41,-0.020649 44 | 2018/01/01 17:00:42,0.18584 45 | 2018/01/01 17:00:43,1.8377 46 | 2018/01/01 17:00:44,2.4572 47 | 2018/01/01 17:00:45,1.4248 48 | 2018/01/01 17:00:46,0.39233 49 | 2018/01/01 17:00:47,-1.2596 50 | 2018/01/01 17:00:48,-1.2596 51 | 2018/01/01 17:00:49,-0.43362 52 | 2018/01/01 17:00:50,0.18584 53 | 2018/01/01 17:00:51,-0.22714 54 | 2018/01/01 17:00:52,-1.2596 55 | 2018/01/01 17:00:53,-2.0855 56 | 2018/01/01 17:00:54,-2.0855 57 | 2018/01/01 17:00:55,-1.879 58 | 2018/01/01 17:00:56,-1.0531 59 | 2018/01/01 17:00:57,-0.43362 60 | 2018/01/01 17:00:58,-0.22714 61 | 2018/01/01 17:00:59,-0.22714 62 | 2018/01/01 17:01:00,-0.020649 63 | 2018/01/01 17:01:01,0.39233 64 | 2018/01/01 17:01:02,1.0118 65 | 2018/01/01 17:01:03,1.0118 66 | 2018/01/01 17:01:04,1.0118 67 | 2018/01/01 17:01:05,1.0118 68 | 2018/01/01 17:01:06,0.59882 69 | 2018/01/01 17:01:07,0.59882 70 | 2018/01/01 17:01:08,0.39233 71 | 2018/01/01 17:01:09,0.39233 -------------------------------------------------------------------------------- /data/sample_data06.csv: -------------------------------------------------------------------------------- 1 | time,data 2 | 2018/01/01 18:00:00,1.2042 3 | 2018/01/01 18:00:01,2.5392 4 | 2018/01/01 18:00:02,2.5392 5 | 2018/01/01 18:00:03,2.3485 6 | 2018/01/01 18:00:04,1.9671 7 | 2018/01/01 18:00:05,1.0135 8 | 2018/01/01 18:00:06,-0.32149 9 | 2018/01/01 18:00:07,-1.0843 10 | 2018/01/01 18:00:08,-1.4658 11 | 2018/01/01 18:00:09,-1.4658 12 | 2018/01/01 18:00:10,-1.0843 13 | 2018/01/01 18:00:11,-1.2751 14 | 2018/01/01 18:00:12,-1.2751 15 | 2018/01/01 18:00:13,-1.2751 16 | 2018/01/01 18:00:14,-1.2751 17 | 2018/01/01 18:00:15,-1.2751 18 | 2018/01/01 18:00:16,-1.0843 19 | 2018/01/01 18:00:17,-0.89363 20 | 2018/01/01 18:00:18,-0.5122 21 | 2018/01/01 18:00:19,-0.32149 22 | 2018/01/01 18:00:20,-0.13077 23 | 2018/01/01 18:00:21,0.44137 24 | 2018/01/01 18:00:22,0.44137 25 | 2018/01/01 18:00:23,0.25065 26 | 2018/01/01 18:00:24,0.059938 27 | 2018/01/01 18:00:25,0.44137 28 | 2018/01/01 18:00:26,0.44137 29 | 2018/01/01 18:00:27,0.25065 30 | 2018/01/01 18:00:28,0.25065 31 | 2018/01/01 18:00:29,0.25065 32 | 2018/01/01 18:00:30,0.44137 33 | 2018/01/01 18:00:31,0.44137 34 | 2018/01/01 18:00:32,0.44137 35 | 2018/01/01 18:00:33,0.25065 36 | 2018/01/01 18:00:34,0.25065 37 | 2018/01/01 18:00:35,-0.13077 38 | 2018/01/01 18:00:36,0.44137 39 | 2018/01/01 18:00:37,1.0135 40 | 2018/01/01 18:00:38,1.0135 41 | 2018/01/01 18:00:39,1.0135 42 | 2018/01/01 18:00:40,1.2042 43 | 2018/01/01 18:00:41,1.3949 44 | 2018/01/01 18:00:42,1.5856 45 | 2018/01/01 18:00:43,1.2042 46 | 2018/01/01 18:00:44,0.25065 47 | 2018/01/01 18:00:45,-0.32149 48 | 2018/01/01 18:00:46,-0.70291 49 | 2018/01/01 18:00:47,-1.0843 50 | 2018/01/01 18:00:48,-1.2751 51 | 2018/01/01 18:00:49,-0.70291 52 | 2018/01/01 18:00:50,-0.5122 53 | 2018/01/01 18:00:51,-1.0843 54 | 2018/01/01 18:00:52,-1.6565 55 | 2018/01/01 18:00:53,-1.8472 56 | 2018/01/01 18:00:54,-1.4658 57 | 2018/01/01 18:00:55,-0.70291 58 | 2018/01/01 18:00:56,0.059938 59 | 2018/01/01 18:00:57,0.25065 60 | 2018/01/01 18:00:58,0.25065 61 | 2018/01/01 18:00:59,0.059938 62 | 2018/01/01 18:01:00,0.059938 63 | 2018/01/01 18:01:01,0.25065 64 | 2018/01/01 18:01:02,0.25065 65 | 2018/01/01 18:01:03,0.059938 66 | 2018/01/01 18:01:04,-0.32149 67 | 2018/01/01 18:01:05,-0.13077 68 | 2018/01/01 18:01:06,-0.13077 69 | 2018/01/01 18:01:07,0.059938 70 | 2018/01/01 18:01:08,0.059938 71 | 2018/01/01 18:01:09,0.059938 -------------------------------------------------------------------------------- /data/sample_data07.csv: -------------------------------------------------------------------------------- 1 | time,data 2 | 2018/01/01 19:00:00,1.1821 3 | 2018/01/01 19:00:01,2.0396 4 | 2018/01/01 19:00:02,2.0396 5 | 2018/01/01 19:00:03,1.8252 6 | 2018/01/01 19:00:04,0.96773 7 | 2018/01/01 19:00:05,-0.10412 8 | 2018/01/01 19:00:06,-0.74723 9 | 2018/01/01 19:00:07,-1.3903 10 | 2018/01/01 19:00:08,-1.6047 11 | 2018/01/01 19:00:09,-1.176 12 | 2018/01/01 19:00:10,-0.9616 13 | 2018/01/01 19:00:11,-1.176 14 | 2018/01/01 19:00:12,-1.6047 15 | 2018/01/01 19:00:13,-1.8191 16 | 2018/01/01 19:00:14,-1.8191 17 | 2018/01/01 19:00:15,-1.3903 18 | 2018/01/01 19:00:16,-0.9616 19 | 2018/01/01 19:00:17,-0.9616 20 | 2018/01/01 19:00:18,-0.74723 21 | 2018/01/01 19:00:19,-0.31849 22 | 2018/01/01 19:00:20,0.11025 23 | 2018/01/01 19:00:21,0.32462 24 | 2018/01/01 19:00:22,0.53899 25 | 2018/01/01 19:00:23,0.32462 26 | 2018/01/01 19:00:24,0.32462 27 | 2018/01/01 19:00:25,0.32462 28 | 2018/01/01 19:00:26,0.53899 29 | 2018/01/01 19:00:27,0.32462 30 | 2018/01/01 19:00:28,0.11025 31 | 2018/01/01 19:00:29,0.11025 32 | 2018/01/01 19:00:30,0.32462 33 | 2018/01/01 19:00:31,0.32462 34 | 2018/01/01 19:00:32,0.32462 35 | 2018/01/01 19:00:33,-0.10412 36 | 2018/01/01 19:00:34,-0.10412 37 | 2018/01/01 19:00:35,-0.10412 38 | 2018/01/01 19:00:36,0.32462 39 | 2018/01/01 19:00:37,0.96773 40 | 2018/01/01 19:00:38,1.1821 41 | 2018/01/01 19:00:39,0.75336 42 | 2018/01/01 19:00:40,1.1821 43 | 2018/01/01 19:00:41,1.8252 44 | 2018/01/01 19:00:42,2.0396 45 | 2018/01/01 19:00:43,1.3965 46 | 2018/01/01 19:00:44,-0.10412 47 | 2018/01/01 19:00:45,-0.31849 48 | 2018/01/01 19:00:46,-1.176 49 | 2018/01/01 19:00:47,-1.176 50 | 2018/01/01 19:00:48,-0.9616 51 | 2018/01/01 19:00:49,-0.53286 52 | 2018/01/01 19:00:50,0.32462 53 | 2018/01/01 19:00:51,-0.53286 54 | 2018/01/01 19:00:52,-1.6047 55 | 2018/01/01 19:00:53,-2.0335 56 | 2018/01/01 19:00:54,-1.6047 57 | 2018/01/01 19:00:55,-0.53286 58 | 2018/01/01 19:00:56,0.32462 59 | 2018/01/01 19:00:57,0.53899 60 | 2018/01/01 19:00:58,0.53899 61 | 2018/01/01 19:00:59,0.53899 62 | 2018/01/01 19:01:00,0.75336 63 | 2018/01/01 19:01:01,0.75336 64 | 2018/01/01 19:01:02,0.53899 65 | 2018/01/01 19:01:03,0.32462 66 | 2018/01/01 19:01:04,-0.10412 67 | 2018/01/01 19:01:05,-0.10412 68 | 2018/01/01 19:01:06,0.11025 69 | 2018/01/01 19:01:07,0.32462 70 | 2018/01/01 19:01:08,0.53899 71 | 2018/01/01 19:01:09,0.79921 -------------------------------------------------------------------------------- /data/sample_data08.csv: -------------------------------------------------------------------------------- 1 | time,data 2 | 2018/01/01 20:00:00,1.3424 3 | 2018/01/01 20:00:01,2.0666 4 | 2018/01/01 20:00:02,2.0666 5 | 2018/01/01 20:00:03,2.0666 6 | 2018/01/01 20:00:04,0.98026 7 | 2018/01/01 20:00:05,0.25606 8 | 2018/01/01 20:00:06,-0.6492 9 | 2018/01/01 20:00:07,-1.1923 10 | 2018/01/01 20:00:08,-1.3734 11 | 2018/01/01 20:00:09,-1.3734 12 | 2018/01/01 20:00:10,-1.0113 13 | 2018/01/01 20:00:11,-1.0113 14 | 2018/01/01 20:00:12,-1.5544 15 | 2018/01/01 20:00:13,-1.9166 16 | 2018/01/01 20:00:14,-1.9166 17 | 2018/01/01 20:00:15,-1.5544 18 | 2018/01/01 20:00:16,-1.1923 19 | 2018/01/01 20:00:17,-0.83025 20 | 2018/01/01 20:00:18,-0.6492 21 | 2018/01/01 20:00:19,-0.28709 22 | 2018/01/01 20:00:20,-0.10604 23 | 2018/01/01 20:00:21,0.25606 24 | 2018/01/01 20:00:22,0.43711 25 | 2018/01/01 20:00:23,0.43711 26 | 2018/01/01 20:00:24,0.43711 27 | 2018/01/01 20:00:25,0.43711 28 | 2018/01/01 20:00:26,0.43711 29 | 2018/01/01 20:00:27,0.25606 30 | 2018/01/01 20:00:28,0.25606 31 | 2018/01/01 20:00:29,0.25606 32 | 2018/01/01 20:00:30,0.25606 33 | 2018/01/01 20:00:31,-0.10604 34 | 2018/01/01 20:00:32,0.075007 35 | 2018/01/01 20:00:33,0.25606 36 | 2018/01/01 20:00:34,0.075007 37 | 2018/01/01 20:00:35,-0.10604 38 | 2018/01/01 20:00:36,0.075007 39 | 2018/01/01 20:00:37,0.25606 40 | 2018/01/01 20:00:38,0.61816 41 | 2018/01/01 20:00:39,0.98026 42 | 2018/01/01 20:00:40,1.5234 43 | 2018/01/01 20:00:41,2.0666 44 | 2018/01/01 20:00:42,1.8855 45 | 2018/01/01 20:00:43,1.1613 46 | 2018/01/01 20:00:44,-0.10604 47 | 2018/01/01 20:00:45,-0.6492 48 | 2018/01/01 20:00:46,-1.0113 49 | 2018/01/01 20:00:47,-1.1923 50 | 2018/01/01 20:00:48,-1.3734 51 | 2018/01/01 20:00:49,-1.0113 52 | 2018/01/01 20:00:50,-0.46815 53 | 2018/01/01 20:00:51,-0.83025 54 | 2018/01/01 20:00:52,-1.3734 55 | 2018/01/01 20:00:53,-1.7355 56 | 2018/01/01 20:00:54,-1.0113 57 | 2018/01/01 20:00:55,-0.46815 58 | 2018/01/01 20:00:56,0.25606 59 | 2018/01/01 20:00:57,0.61816 60 | 2018/01/01 20:00:58,0.43711 61 | 2018/01/01 20:00:59,0.43711 62 | 2018/01/01 20:01:00,0.43711 63 | 2018/01/01 20:01:01,0.61816 64 | 2018/01/01 20:01:02,0.61816 65 | 2018/01/01 20:01:03,0.43711 66 | 2018/01/01 20:01:04,0.43711 67 | 2018/01/01 20:01:05,0.25606 68 | 2018/01/01 20:01:06,0.25606 69 | 2018/01/01 20:01:07,0.43711 70 | 2018/01/01 20:01:08,0.79921 71 | 2018/01/01 20:01:09,0.79921 -------------------------------------------------------------------------------- /data/sample_data09.csv: -------------------------------------------------------------------------------- 1 | time,data 2 | 2018/01/01 21:00:00,0.93054 3 | 2018/01/01 21:00:01,1.6939 4 | 2018/01/01 21:00:02,2.4572 5 | 2018/01/01 21:00:03,2.4572 6 | 2018/01/01 21:00:04,1.185 7 | 2018/01/01 21:00:05,-0.087238 8 | 2018/01/01 21:00:06,-1.105 9 | 2018/01/01 21:00:07,-1.3595 10 | 2018/01/01 21:00:08,-1.6139 11 | 2018/01/01 21:00:09,-0.85057 12 | 2018/01/01 21:00:10,-0.59613 13 | 2018/01/01 21:00:11,-1.105 14 | 2018/01/01 21:00:12,-1.6139 15 | 2018/01/01 21:00:13,-1.8683 16 | 2018/01/01 21:00:14,-1.3595 17 | 2018/01/01 21:00:15,-1.105 18 | 2018/01/01 21:00:16,-0.85057 19 | 2018/01/01 21:00:17,-0.85057 20 | 2018/01/01 21:00:18,-0.85057 21 | 2018/01/01 21:00:19,-0.59613 22 | 2018/01/01 21:00:20,-0.087238 23 | 2018/01/01 21:00:21,0.16721 24 | 2018/01/01 21:00:22,0.42165 25 | 2018/01/01 21:00:23,0.67609 26 | 2018/01/01 21:00:24,0.67609 27 | 2018/01/01 21:00:25,0.67609 28 | 2018/01/01 21:00:26,0.67609 29 | 2018/01/01 21:00:27,0.67609 30 | 2018/01/01 21:00:28,0.42165 31 | 2018/01/01 21:00:29,0.16721 32 | 2018/01/01 21:00:30,0.42165 33 | 2018/01/01 21:00:31,0.16721 34 | 2018/01/01 21:00:32,0.16721 35 | 2018/01/01 21:00:33,0.16721 36 | 2018/01/01 21:00:34,-0.087238 37 | 2018/01/01 21:00:35,0.16721 38 | 2018/01/01 21:00:36,0.16721 39 | 2018/01/01 21:00:37,0.16721 40 | 2018/01/01 21:00:38,-0.34168 41 | 2018/01/01 21:00:39,0.16721 42 | 2018/01/01 21:00:40,1.185 43 | 2018/01/01 21:00:41,1.9483 44 | 2018/01/01 21:00:42,2.4572 45 | 2018/01/01 21:00:43,1.4394 46 | 2018/01/01 21:00:44,0.16721 47 | 2018/01/01 21:00:45,-0.59613 48 | 2018/01/01 21:00:46,-1.105 49 | 2018/01/01 21:00:47,-0.85057 50 | 2018/01/01 21:00:48,-0.59613 51 | 2018/01/01 21:00:49,-0.087238 52 | 2018/01/01 21:00:50,-0.087238 53 | 2018/01/01 21:00:51,-1.105 54 | 2018/01/01 21:00:52,-1.8683 55 | 2018/01/01 21:00:53,-1.8683 56 | 2018/01/01 21:00:54,-1.3595 57 | 2018/01/01 21:00:55,-0.34168 58 | 2018/01/01 21:00:56,0.42165 59 | 2018/01/01 21:00:57,0.42165 60 | 2018/01/01 21:00:58,-0.087238 61 | 2018/01/01 21:00:59,0.16721 62 | 2018/01/01 21:01:00,0.42165 63 | 2018/01/01 21:01:01,0.93054 64 | 2018/01/01 21:01:02,0.67609 65 | 2018/01/01 21:01:03,0.67609 66 | 2018/01/01 21:01:04,0.42165 67 | 2018/01/01 21:01:05,0.16721 68 | 2018/01/01 21:01:06,0.16721 69 | 2018/01/01 21:01:07,0.16721 70 | 2018/01/01 21:01:08,-0.34168 71 | 2018/01/01 21:01:09,-0.087238 -------------------------------------------------------------------------------- /data/sample_data10.csv: -------------------------------------------------------------------------------- 1 | time,data 2 | 2018/01/01 22:00:00,-0.33813 3 | 2018/01/01 22:00:01,0.01514 4 | 2018/01/01 22:00:02,1.075 5 | 2018/01/01 22:00:03,1.7815 6 | 2018/01/01 22:00:04,1.075 7 | 2018/01/01 22:00:05,0.01514 8 | 2018/01/01 22:00:06,-1.0447 9 | 2018/01/01 22:00:07,-2.1045 10 | 2018/01/01 22:00:08,-1.398 11 | 2018/01/01 22:00:09,-0.33813 12 | 2018/01/01 22:00:10,1.075 13 | 2018/01/01 22:00:11,0.36841 14 | 2018/01/01 22:00:12,-0.69141 15 | 2018/01/01 22:00:13,-1.7512 16 | 2018/01/01 22:00:14,-1.7512 17 | 2018/01/01 22:00:15,-1.0447 18 | 2018/01/01 22:00:16,-0.33813 19 | 2018/01/01 22:00:17,-0.33813 20 | 2018/01/01 22:00:18,0.36841 21 | 2018/01/01 22:00:19,0.36841 22 | 2018/01/01 22:00:20,0.72169 23 | 2018/01/01 22:00:21,1.075 24 | 2018/01/01 22:00:22,0.72169 25 | 2018/01/01 22:00:23,1.4282 26 | 2018/01/01 22:00:24,1.075 27 | 2018/01/01 22:00:25,1.075 28 | 2018/01/01 22:00:26,0.72169 29 | 2018/01/01 22:00:27,0.72169 30 | 2018/01/01 22:00:28,0.36841 31 | 2018/01/01 22:00:29,0.36841 32 | 2018/01/01 22:00:30,0.36841 33 | 2018/01/01 22:00:31,0.01514 34 | 2018/01/01 22:00:32,0.01514 35 | 2018/01/01 22:00:33,-0.33813 36 | 2018/01/01 22:00:34,-0.33813 37 | 2018/01/01 22:00:35,-0.33813 38 | 2018/01/01 22:00:36,0.01514 39 | 2018/01/01 22:00:37,-0.33813 40 | 2018/01/01 22:00:38,-0.33813 41 | 2018/01/01 22:00:39,0.36841 42 | 2018/01/01 22:00:40,0.72169 43 | 2018/01/01 22:00:41,1.075 44 | 2018/01/01 22:00:42,1.075 45 | 2018/01/01 22:00:43,-1.0447 46 | 2018/01/01 22:00:44,-1.0447 47 | 2018/01/01 22:00:45,0.01514 48 | 2018/01/01 22:00:46,0.36841 49 | 2018/01/01 22:00:47,0.72169 50 | 2018/01/01 22:00:48,-0.33813 51 | 2018/01/01 22:00:49,-2.1045 52 | 2018/01/01 22:00:50,-1.7512 53 | 2018/01/01 22:00:51,-1.7512 54 | 2018/01/01 22:00:52,-2.4578 55 | 2018/01/01 22:00:53,-2.1045 56 | 2018/01/01 22:00:54,-0.69141 57 | 2018/01/01 22:00:55,0.72169 58 | 2018/01/01 22:00:56,1.7815 59 | 2018/01/01 22:00:57,2.1348 60 | 2018/01/01 22:00:58,1.075 61 | 2018/01/01 22:00:59,0.01514 62 | 2018/01/01 22:01:00,-0.33813 63 | 2018/01/01 22:01:01,-0.33813 64 | 2018/01/01 22:01:02,0.01514 65 | 2018/01/01 22:01:03,0.01514 66 | 2018/01/01 22:01:04,0.36841 67 | 2018/01/01 22:01:05,0.36841 68 | 2018/01/01 22:01:06,0.36841 -------------------------------------------------------------------------------- /notebook/clustering_waveform_using_kshape.ipynb: -------------------------------------------------------------------------------- 1 | { 2 | "cells": [ 3 | { 4 | "cell_type": "markdown", 5 | "metadata": {}, 6 | "source": [ 7 | "## Clustering using KShape\n", 8 | "- Clustering Methods in Time Series Data\n", 9 | "- Implemented in [tslearn](https://github.com/rtavenar/tslearn)\n", 10 | "\n", 11 | "The algorithm will be executed in the following.\n", 12 | "\n", 13 | "1. Use a distance scale based on cross-correlation measurements (Shape-based distance: SBD)\n", 14 | "2. Compute the center of gravity of time series clusters based on 1. (a new center-of-gravity-based clustering algorithm that preserves the shape of the time series)\n", 15 | "3. The method of dividing the data into clusters is the same as for general kmeans, but 1. and 2. above are used when calculating the distance scale and center of gravity\n", 16 | "\n", 17 | "Please refer to the paper for detailed formulas and theory.\n", 18 | "\n", 19 | "[J. Paparrizos & L. Gravano. k-Shape: Efficient and Accurate Clustering of Time Series. SIGMOD 2015. pp. 1855-1870.](http://www1.cs.columbia.edu/~jopa/Papers/PaparrizosSIGMOD2015.pdf)" 20 | ] 21 | }, 22 | { 23 | "cell_type": "code", 24 | "execution_count": 1, 25 | "metadata": {}, 26 | "outputs": [ 27 | { 28 | "name": "stderr", 29 | "output_type": "stream", 30 | "text": [ 31 | "/Users/masataka/.pyenv/versions/anaconda3-5.3.0/envs/datascience/lib/python3.8/site-packages/sklearn/utils/deprecation.py:143: FutureWarning: The sklearn.cluster.k_means_ module is deprecated in version 0.22 and will be removed in version 0.24. The corresponding classes / functions should instead be imported from sklearn.cluster. Anything that cannot be imported from sklearn.cluster is now part of the private API.\n", 32 | " warnings.warn(message, FutureWarning)\n" 33 | ] 34 | } 35 | ], 36 | "source": [ 37 | "import pandas as pd\n", 38 | "import numpy as np\n", 39 | "import glob\n", 40 | "import matplotlib.pyplot as plt\n", 41 | "from tslearn.clustering import KShape\n", 42 | "from tslearn.preprocessing import TimeSeriesScalerMeanVariance\n", 43 | "\n", 44 | "%matplotlib inline" 45 | ] 46 | }, 47 | { 48 | "cell_type": "code", 49 | "execution_count": 2, 50 | "metadata": {}, 51 | "outputs": [], 52 | "source": [ 53 | "def align_timeseries_dataset(dfs, target_col=None):\n", 54 | " # Load dataframes, turns them into a time-series array, and stores them in a list\n", 55 | " tsdata = []\n", 56 | " for i, df in enumerate(dfs):\n", 57 | " tsdata.append(df[target_col].values.tolist()[:])\n", 58 | " \n", 59 | " # Check the maximum length of each time series data\n", 60 | " len_max = 0\n", 61 | " for ts in tsdata:\n", 62 | " if len(ts) > len_max:\n", 63 | " len_max = len(ts)\n", 64 | " \n", 65 | " # Assign the last data to align the length of the time series data\n", 66 | " for i, ts in enumerate(tsdata):\n", 67 | " len_add = len_max - len(ts)\n", 68 | " tsdata[i] = ts + [ts[-1]] * len_add\n", 69 | " \n", 70 | " tsdata = np.array(tsdata)\n", 71 | " return tsdata\n", 72 | "\n", 73 | "def transform_timeseries_vectors(timeseries_dataset):\n", 74 | " # Transform vectors\n", 75 | " stack_list = []\n", 76 | " for j in range(len(timeseries_dataset)):\n", 77 | " data = np.array(timeseries_dataset[j])\n", 78 | " data = data.reshape((1, len(data))).T\n", 79 | " stack_list.append(data)\n", 80 | " \n", 81 | " # Convert to one-dimensional array\n", 82 | " transformed_data = np.stack(stack_list, axis=0)\n", 83 | " return transformed_data" 84 | ] 85 | }, 86 | { 87 | "cell_type": "markdown", 88 | "metadata": {}, 89 | "source": [ 90 | "## Prepare for datasets" 91 | ] 92 | }, 93 | { 94 | "cell_type": "code", 95 | "execution_count": 3, 96 | "metadata": {}, 97 | "outputs": [], 98 | "source": [ 99 | "# Get some filelists\n", 100 | "filenames = sorted(glob.glob('../data/sample_data*.csv'))" 101 | ] 102 | }, 103 | { 104 | "cell_type": "code", 105 | "execution_count": 4, 106 | "metadata": {}, 107 | "outputs": [], 108 | "source": [ 109 | "# Loads a dataframe from a file and stores it in a list\n", 110 | "df_list = []\n", 111 | "for filename in filenames:\n", 112 | " df = pd.read_csv(filename, index_col=None, header=0)\n", 113 | " df_list.append(df)" 114 | ] 115 | }, 116 | { 117 | "cell_type": "code", 118 | "execution_count": 5, 119 | "metadata": {}, 120 | "outputs": [], 121 | "source": [ 122 | "dfs = align_timeseries_dataset(df_list, target_col='data')\n", 123 | "transformed_data = transform_timeseries_vectors(dfs)" 124 | ] 125 | }, 126 | { 127 | "cell_type": "markdown", 128 | "metadata": {}, 129 | "source": [ 130 | "## Visualize clustering results" 131 | ] 132 | }, 133 | { 134 | "cell_type": "code", 135 | "execution_count": null, 136 | "metadata": {}, 137 | "outputs": [], 138 | "source": [ 139 | "# Set a seed for reproducibility\n", 140 | "seed = 0\n", 141 | "np.random.seed(seed)" 142 | ] 143 | }, 144 | { 145 | "cell_type": "code", 146 | "execution_count": 6, 147 | "metadata": {}, 148 | "outputs": [ 149 | { 150 | "name": "stdout", 151 | "output_type": "stream", 152 | "text": [ 153 | "Init 1\n", 154 | "0.015 --> 0.016 --> \n", 155 | "Init 2\n", 156 | "Resumed because of empty cluster\n", 157 | "Init 2\n", 158 | "0.010 --> 0.012 --> \n", 159 | "Init 3\n", 160 | "0.019 --> Resumed because of empty cluster\n", 161 | "Init 3\n", 162 | "0.018 --> 0.018 --> 0.018 --> \n", 163 | "Init 4\n", 164 | "0.013 --> 0.013 --> 0.022 --> \n", 165 | "Init 5\n", 166 | "Resumed because of empty cluster\n", 167 | "Init 5\n", 168 | "0.006 --> 0.010 --> \n", 169 | "Init 6\n", 170 | "Resumed because of empty cluster\n", 171 | "Init 6\n", 172 | "0.006 --> 0.010 --> \n" 173 | ] 174 | }, 175 | { 176 | "data": { 177 | "image/png": "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\n", 178 | "text/plain": [ 179 | "
" 180 | ] 181 | }, 182 | "metadata": { 183 | "needs_background": "light" 184 | }, 185 | "output_type": "display_data" 186 | } 187 | ], 188 | "source": [ 189 | "# To calculate the cross-correlation, it must be normalized.\n", 190 | "# TimeSeriesScalerMeanVariance will be the class that normalizes the data.\n", 191 | "stack_data = TimeSeriesScalerMeanVariance(mu=0.0, std=1.0).fit_transform(transformed_data)\n", 192 | "\n", 193 | "# Instantiation of KShape class\n", 194 | "ks = KShape(n_clusters=2, n_init=10, verbose=True, random_state=seed)\n", 195 | "y_pred = ks.fit_predict(stack_data)\n", 196 | "\n", 197 | "# Clustering and visualizing results\n", 198 | "plt.figure(figsize=(16, 9))\n", 199 | "for yi in range(2):\n", 200 | " plt.subplot(2, 1, 1 + yi)\n", 201 | " for xx in stack_data[y_pred == yi]:\n", 202 | " plt.plot(xx.ravel(), \"k-\", alpha=.2)\n", 203 | " # plt.plot(ks.cluster_centers_[yi].ravel(), \"r-\")\n", 204 | " plt.title(\"Cluster %d\" % (yi + 1))\n", 205 | "\n", 206 | "plt.tight_layout()\n", 207 | "plt.show()" 208 | ] 209 | }, 210 | { 211 | "cell_type": "markdown", 212 | "metadata": {}, 213 | "source": [ 214 | "## Calculation of the number of clusters using the elbow method\n", 215 | "- What is the elbow method?\n", 216 | " - The sum of the squares of the distances from each point to the assigned cluster center is calculated as the sum of the squares of the within-cluster error squares (SSE).\n", 217 | " - The method is to vary the number of clusters and plot the SSE values for each, with the 'elbow' curved point being the optimal number of clusters.\n", 218 | "- The elbow method is described in wikipedia.\n", 219 | " - https://en.wikipedia.org/wiki/Elbow_method_(clustering)" 220 | ] 221 | }, 222 | { 223 | "cell_type": "code", 224 | "execution_count": 7, 225 | "metadata": { 226 | "scrolled": false 227 | }, 228 | "outputs": [ 229 | { 230 | "name": "stdout", 231 | "output_type": "stream", 232 | "text": [ 233 | "Init 1\n", 234 | "0.023 --> 0.024 --> \n", 235 | "Init 2\n", 236 | "0.022 --> 0.023 --> \n", 237 | "Init 3\n", 238 | "0.022 --> 0.022 --> 0.023 --> \n", 239 | "Init 4\n", 240 | "0.022 --> 0.022 --> 0.023 --> \n", 241 | "Init 5\n", 242 | "0.022 --> 0.023 --> \n", 243 | "Init 6\n", 244 | "0.023 --> 0.023 --> 0.033 --> \n", 245 | "Init 7\n", 246 | "0.022 --> 0.022 --> 0.023 --> \n", 247 | "Init 8\n", 248 | "0.023 --> 0.024 --> \n", 249 | "Init 9\n", 250 | "0.023 --> 0.023 --> \n", 251 | "Init 10\n", 252 | "0.027 --> 0.023 --> 0.024 --> \n", 253 | "Init 1\n", 254 | "0.015 --> 0.016 --> \n", 255 | "Init 2\n", 256 | "Resumed because of empty cluster\n", 257 | "Init 2\n", 258 | "0.010 --> 0.012 --> \n", 259 | "Init 3\n", 260 | "0.019 --> Resumed because of empty cluster\n", 261 | "Init 3\n", 262 | "0.018 --> 0.018 --> 0.018 --> \n", 263 | "Init 4\n", 264 | "0.013 --> 0.013 --> 0.022 --> \n", 265 | "Init 5\n", 266 | "Resumed because of empty cluster\n", 267 | "Init 5\n", 268 | "0.006 --> 0.010 --> \n", 269 | "Init 6\n", 270 | "Resumed because of empty cluster\n", 271 | "Init 6\n", 272 | "0.006 --> 0.010 --> \n", 273 | "Init 1\n", 274 | "0.009 --> Resumed because of empty cluster\n", 275 | "Init 1\n", 276 | "0.008 --> 0.009 --> \n", 277 | "Init 2\n", 278 | "0.015 --> Resumed because of empty cluster\n", 279 | "Init 2\n", 280 | "0.010 --> 0.009 --> 0.016 --> \n", 281 | "Init 3\n", 282 | "Resumed because of empty cluster\n", 283 | "Init 3\n", 284 | "Resumed because of empty cluster\n", 285 | "Init 3\n", 286 | "0.004 --> 0.007 --> \n", 287 | "Init 4\n", 288 | "Resumed because of empty cluster\n", 289 | "Init 4\n", 290 | "0.007 --> 0.003 --> 0.007 --> \n", 291 | "Init 5\n", 292 | "0.009 --> Resumed because of empty cluster\n", 293 | "Init 1\n", 294 | "Resumed because of empty cluster\n", 295 | "Init 1\n", 296 | "0.006 --> Resumed because of empty cluster\n", 297 | "Init 1\n", 298 | "0.007 --> 0.005 --> 0.008 --> \n", 299 | "Init 2\n", 300 | "Resumed because of empty cluster\n", 301 | "Init 2\n", 302 | "Resumed because of empty cluster\n", 303 | "Init 2\n", 304 | "Resumed because of empty cluster\n", 305 | "Init 2\n", 306 | "Resumed because of empty cluster\n", 307 | "Init 2\n", 308 | "0.008 --> Resumed because of empty cluster\n", 309 | "Init 2\n", 310 | "Resumed because of empty cluster\n", 311 | "Init 2\n", 312 | "0.004 --> 0.006 --> \n", 313 | "Init 1\n", 314 | "Resumed because of empty cluster\n", 315 | "Init 1\n", 316 | "0.001 --> Resumed because of empty cluster\n", 317 | "Init 1\n", 318 | "Resumed because of empty cluster\n", 319 | "Init 1\n", 320 | "Resumed because of empty cluster\n", 321 | "Init 1\n", 322 | "Resumed because of empty cluster\n", 323 | "Init 1\n", 324 | "Resumed because of empty cluster\n", 325 | "Init 1\n", 326 | "Resumed because of empty cluster\n", 327 | "Init 1\n", 328 | "Resumed because of empty cluster\n", 329 | "Init 1\n", 330 | "Resumed because of empty cluster\n", 331 | "Init 1\n", 332 | "Resumed because of empty cluster\n", 333 | "Init 1\n", 334 | "Resumed because of empty cluster\n", 335 | "Init 1\n", 336 | "0.002 --> 0.002 --> Resumed because of empty cluster\n", 337 | "Init 1\n", 338 | "Resumed because of empty cluster\n", 339 | "Init 1\n", 340 | "Resumed because of empty cluster\n", 341 | "Init 1\n", 342 | "Resumed because of empty cluster\n", 343 | "Init 1\n", 344 | "Resumed because of empty cluster\n", 345 | "Init 1\n", 346 | "Resumed because of empty cluster\n", 347 | "Init 1\n", 348 | "Resumed because of empty cluster\n", 349 | "Init 1\n", 350 | "Resumed because of empty cluster\n", 351 | "Init 1\n", 352 | "Resumed because of empty cluster\n", 353 | "Init 1\n", 354 | "Resumed because of empty cluster\n", 355 | "Init 1\n", 356 | "Resumed because of empty cluster\n", 357 | "Init 1\n", 358 | "Resumed because of empty cluster\n", 359 | "Init 1\n", 360 | "Resumed because of empty cluster\n", 361 | "Init 1\n", 362 | "Resumed because of empty cluster\n", 363 | "Init 1\n", 364 | "Resumed because of empty cluster\n", 365 | "Init 1\n", 366 | "Resumed because of empty cluster\n", 367 | "Init 1\n", 368 | "Resumed because of empty cluster\n", 369 | "Init 1\n", 370 | "Resumed because of empty cluster\n", 371 | "Init 1\n", 372 | "Resumed because of empty cluster\n", 373 | "Init 1\n", 374 | "Resumed because of empty cluster\n", 375 | "Init 1\n", 376 | "Resumed because of empty cluster\n", 377 | "Init 1\n", 378 | "Resumed because of empty cluster\n", 379 | "Init 1\n", 380 | "Resumed because of empty cluster\n", 381 | "Init 1\n", 382 | "Resumed because of empty cluster\n", 383 | "Init 1\n", 384 | "Resumed because of empty cluster\n", 385 | "Init 1\n", 386 | "Resumed because of empty cluster\n", 387 | "Init 1\n", 388 | "Resumed because of empty cluster\n", 389 | "Init 1\n", 390 | "Resumed because of empty cluster\n", 391 | "Init 1\n", 392 | "Resumed because of empty cluster\n", 393 | "Init 1\n", 394 | "Resumed because of empty cluster\n", 395 | "Init 1\n", 396 | "Resumed because of empty cluster\n", 397 | "Init 1\n", 398 | "Resumed because of empty cluster\n", 399 | "Init 1\n", 400 | "Resumed because of empty cluster\n", 401 | "Init 1\n", 402 | "Resumed because of empty cluster\n", 403 | "Init 1\n", 404 | "Resumed because of empty cluster\n", 405 | "Init 1\n", 406 | "Resumed because of empty cluster\n", 407 | "Init 1\n", 408 | "Resumed because of empty cluster\n", 409 | "Init 1\n", 410 | "Resumed because of empty cluster\n", 411 | "Init 1\n", 412 | "Resumed because of empty cluster\n", 413 | "Init 1\n", 414 | "Resumed because of empty cluster\n", 415 | "Init 1\n", 416 | "Resumed because of empty cluster\n", 417 | "Init 1\n", 418 | "Resumed because of empty cluster\n", 419 | "Init 1\n", 420 | "Resumed because of empty cluster\n", 421 | "Init 1\n", 422 | "Resumed because of empty cluster\n", 423 | "Init 1\n", 424 | "Resumed because of empty cluster\n", 425 | "Init 1\n", 426 | "Resumed because of empty cluster\n", 427 | "Init 1\n", 428 | "Resumed because of empty cluster\n", 429 | "Init 1\n", 430 | "Resumed because of empty cluster\n", 431 | "Init 1\n", 432 | "Resumed because of empty cluster\n" 433 | ] 434 | }, 435 | { 436 | "data": { 437 | "image/png": "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\n", 438 | "text/plain": [ 439 | "
" 440 | ] 441 | }, 442 | "metadata": { 443 | "needs_background": "light" 444 | }, 445 | "output_type": "display_data" 446 | } 447 | ], 448 | "source": [ 449 | "distortions = []\n", 450 | "\n", 451 | "# Calculated from 1~10 clusters\n", 452 | "for i in range(1,11):\n", 453 | " ks = KShape(n_clusters=i, n_init=10, verbose=True, random_state=seed)\n", 454 | " # Run clustering calculations\n", 455 | " ks.fit(stack_data)\n", 456 | " # ks.fit gives ks.inertia_\n", 457 | " distortions.append(ks.inertia_)\n", 458 | "\n", 459 | "plt.plot(range(1,11), distortions, marker='o')\n", 460 | "plt.xlabel('Number of clusters')\n", 461 | "plt.ylabel('Distortion')\n", 462 | "plt.show()" 463 | ] 464 | }, 465 | { 466 | "cell_type": "code", 467 | "execution_count": null, 468 | "metadata": {}, 469 | "outputs": [], 470 | "source": [] 471 | } 472 | ], 473 | "metadata": { 474 | "kernelspec": { 475 | "display_name": "Python 3", 476 | "language": "python", 477 | "name": "python3" 478 | }, 479 | "language_info": { 480 | "codemirror_mode": { 481 | "name": "ipython", 482 | "version": 3 483 | }, 484 | "file_extension": ".py", 485 | "mimetype": "text/x-python", 486 | "name": "python", 487 | "nbconvert_exporter": "python", 488 | "pygments_lexer": "ipython3", 489 | "version": "3.8.3" 490 | } 491 | }, 492 | "nbformat": 4, 493 | "nbformat_minor": 2 494 | } 495 | --------------------------------------------------------------------------------