├── .gitignore
├── LICENSE
├── README.md
├── gpu
    ├── 01-Intro_to_cuDF.ipynb
    ├── 01-Intro_to_cuGraph.ipynb
    ├── 01-Introduction-LinearRegression-Hyperparam.ipynb
    ├── 02-Intro_to_cuDF_UDFs.ipynb
    ├── 02-LogisticRegression.ipynb
    ├── 02-Louvain.ipynb
    ├── 03-Pagerank.ipynb
    ├── 03-UMAP.ipynb
    ├── Introduction.ipynb
    ├── README.md
    ├── data
    │   └── karate-data.csv
    └── index.ipynb
└── modern_time_series_analysis
    ├── ModernTimeSeriesAnalysis
        ├── DeepLearning
        │   ├── Electricity
        │   │   ├── 5_Forecasting_electric_use_with_mxnet_INSTRUCTOR.ipynb
        │   │   ├── 5_Forecasting_electric_use_with_mxnet_STUDENT.ipynb
        │   │   ├── electricity.diff.txt
        │   │   ├── models.json
        │   │   ├── perf.py
        │   │   └── run.json
        │   └── Stocks
        │   │   ├── 6_Stocks_INSTRUCTOR.ipynb
        │   │   ├── 6_Stocks_STUDENT.ipynb
        │   │   └── sp500.csv
        ├── MachineLearning
        │   ├── 3_Trees_for_Classification_and_Prediction_INSTRUCTOR.ipynb
        │   ├── 3_Trees_for_Classification_and_Prediction_STUDENT.ipynb
        │   ├── 4_Clustering_INSTRUCTOR.ipynb
        │   ├── 4_Clustering_STUDENT.ipynb
        │   ├── data
        │   │   ├── 50words.csv
        │   │   ├── AirPassengers.csv
        │   │   ├── featurized_hists.csv
        │   │   ├── featurized_words.csv
        │   │   ├── full_eeg_data_features.csv
        │   │   ├── pairwise_word_distances.npy
        │   │   └── training_eeg.csv
        │   ├── dtaidistance
        │   │   ├── __init__.py
        │   │   ├── alignment.py
        │   │   ├── clustering.py
        │   │   ├── dp.py
        │   │   ├── dtw.py
        │   │   ├── dtw_c.pyx
        │   │   ├── dtw_ndim.py
        │   │   ├── dtw_ndim_visualisation.py
        │   │   ├── dtw_visualisation.py
        │   │   ├── dtw_weighted.py
        │   │   └── util.py
        │   └── full_eeg_data_features.csv
        └── StateSpaceModels
        │   ├── 1_Structural_Time_Series_INSTRUCTOR.ipynb
        │   ├── 1_Structural_Time_Series_STUDENT.ipynb
        │   ├── 2_Gaussian_HMM_INSTRUCTOR.ipynb
        │   ├── 2_Gaussian_HMM_STUDENT.ipynb
        │   ├── Nile.csv
        │   └── global_temps.csv
    ├── README.md
    ├── SciPyModernTimeSeries.pdf
    └── requirements.txt


/.gitignore:
--------------------------------------------------------------------------------
  1 | # Byte-compiled / optimized / DLL files
  2 | __pycache__/
  3 | *.py[cod]
  4 | *$py.class
  5 | 
  6 | # C extensions
  7 | *.so
  8 | 
  9 | # Distribution / packaging
 10 | .Python
 11 | build/
 12 | develop-eggs/
 13 | dist/
 14 | downloads/
 15 | eggs/
 16 | .eggs/
 17 | lib/
 18 | lib64/
 19 | parts/
 20 | sdist/
 21 | var/
 22 | wheels/
 23 | *.egg-info/
 24 | .installed.cfg
 25 | *.egg
 26 | MANIFEST
 27 | 
 28 | # PyInstaller
 29 | #  Usually these files are written by a python script from a template
 30 | #  before PyInstaller builds the exe, so as to inject date/other infos into it.
 31 | *.manifest
 32 | *.spec
 33 | 
 34 | # Installer logs
 35 | pip-log.txt
 36 | pip-delete-this-directory.txt
 37 | 
 38 | # Unit test / coverage reports
 39 | htmlcov/
 40 | .tox/
 41 | .coverage
 42 | .coverage.*
 43 | .cache
 44 | nosetests.xml
 45 | coverage.xml
 46 | *.cover
 47 | .hypothesis/
 48 | .pytest_cache/
 49 | 
 50 | # Translations
 51 | *.mo
 52 | *.pot
 53 | 
 54 | # Django stuff:
 55 | *.log
 56 | local_settings.py
 57 | db.sqlite3
 58 | 
 59 | # Flask stuff:
 60 | instance/
 61 | .webassets-cache
 62 | 
 63 | # Scrapy stuff:
 64 | .scrapy
 65 | 
 66 | # Sphinx documentation
 67 | docs/_build/
 68 | 
 69 | # PyBuilder
 70 | target/
 71 | 
 72 | # Jupyter Notebook
 73 | .ipynb_checkpoints
 74 | 
 75 | # pyenv
 76 | .python-version
 77 | 
 78 | # celery beat schedule file
 79 | celerybeat-schedule
 80 | 
 81 | # SageMath parsed files
 82 | *.sage.py
 83 | 
 84 | # Environments
 85 | .env
 86 | .venv
 87 | env/
 88 | venv/
 89 | ENV/
 90 | env.bak/
 91 | venv.bak/
 92 | 
 93 | # Spyder project settings
 94 | .spyderproject
 95 | .spyproject
 96 | 
 97 | # Rope project settings
 98 | .ropeproject
 99 | 
100 | # mkdocs documentation
101 | /site
102 | 
103 | # mypy
104 | .mypy_cache/
105 | 


--------------------------------------------------------------------------------
/LICENSE:
--------------------------------------------------------------------------------
 1 | MIT License
 2 | 
 3 | Copyright (c) 2019 John Stilley
 4 | 
 5 | Permission is hereby granted, free of charge, to any person obtaining a copy
 6 | of this software and associated documentation files (the "Software"), to deal
 7 | in the Software without restriction, including without limitation the rights
 8 | to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
 9 | copies of the Software, and to permit persons to whom the Software is
10 | furnished to do so, subject to the following conditions:
11 | 
12 | The above copyright notice and this permission notice shall be included in all
13 | copies or substantial portions of the Software.
14 | 
15 | THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
16 | IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
17 | FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
18 | AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
19 | LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
20 | OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
21 | SOFTWARE.
22 | 


--------------------------------------------------------------------------------
/README.md:
--------------------------------------------------------------------------------
 1 | # Tutorial Sessions for SciPy Con 2019
 2 | 
 3 | I couldn't attend every tutorial, but these seemed like fun:
 4 | 
 5 | 
 6 | ## Monday, 8AM (Room 203) Modern Time Series Analysis
 7 | 
 8 | * [Here](https://github.com/john-science/scipy_con_2019/tree/main/modern_time_series_analysis/) is my version of this tutorial, since it was not available online.
 9 | 
10 | ## Monday, 8AM (Room 106) PyTest
11 | 
12 | * [Here](https://leemangeophysicalllc.github.io/testing-with-python/) is a link to their official GitHub.
13 | 
14 | ## Monday 1PM (Room 203) Bayesian Data Science: Probabilistic Programming
15 | 
16 | * [Here](https://github.com/john-science/bayesian-stats-modelling-tutorial) is my fork of their GitHub.
17 | 
18 | 
19 | ## Tuesday 8AM (Room 106) RAPIDS: Open GPU Data Science
20 | 
21 | * [Here](https://github.com/john-science/scipy_con_2019/tree/main/gpu) is my version of their tutorial, since it was not available online.
22 | 
23 | ## Tuesday 1PM (Room 104) Escape from Auto-manual Testing with Hypothesis!
24 | 
25 | * [Here](https://github.com/john-science/escape-from-automanual-testing) is my fork of their GitHub.
26 | 


--------------------------------------------------------------------------------
/gpu/01-Intro_to_cuDF.ipynb:
--------------------------------------------------------------------------------
  1 | <!DOCTYPE HTML>
  2 | <html>
  3 | 
  4 | <head>
  5 |     <meta charset="utf-8">
  6 | 
  7 |     <title>Jupyter Notebook</title>
  8 |     <link id="favicon" rel="shortcut icon" type="image/x-icon" href="/static/base/images/favicon.ico?v=97c6417ed01bdc0ae3ef32ae4894fd03">
  9 |     <meta http-equiv="X-UA-Compatible" content="IE=edge" />
 10 |     <link rel="stylesheet" href="/static/components/jquery-ui/themes/smoothness/jquery-ui.min.css?v=3c2a865c832a1322285c55c6ed99abb2" type="text/css" />
 11 |     <link rel="stylesheet" href="/static/components/jquery-typeahead/dist/jquery.typeahead.min.css?v=7afb461de36accb1aa133a1710f5bc56" type="text/css" />
 12 |     <meta name="viewport" content="width=device-width, initial-scale=1.0">
 13 |     
 14 |     
 15 | 
 16 |     <link rel="stylesheet" href="/static/style/style.min.css?v=56cad67b8b1ff3ed9b48850e9efcffb3" type="text/css"/>
 17 |     
 18 | <link rel="stylesheet" href="/static/auth/css/override.css?v=19ec59d2c4f1203c49fe47462028cd9a" type="text/css" />
 19 | 
 20 |     <link rel="stylesheet" href="/custom/custom.css" type="text/css" />
 21 |     <script src="/static/components/es6-promise/promise.min.js?v=f004a16cb856e0ff11781d01ec5ca8fe" type="text/javascript" charset="utf-8"></script>
 22 |     <script src="/static/components/preact/index.js?v=00a2fac73c670ce39ac53d26640eb542" type="text/javascript"></script>
 23 |     <script src="/static/components/proptypes/index.js?v=c40890eb04df9811fcc4d47e53a29604" type="text/javascript"></script>
 24 |     <script src="/static/components/preact-compat/index.js?v=aea8f6660e54b18ace8d84a9b9654c1c" type="text/javascript"></script>
 25 |     <script src="/static/components/requirejs/require.js?v=951f856e81496aaeec2e71a1c2c0d51f" type="text/javascript" charset="utf-8"></script>
 26 |     <script>
 27 |       require.config({
 28 |           
 29 |           urlArgs: "v=20190709124529",
 30 |           
 31 |           baseUrl: '/static/',
 32 |           paths: {
 33 |             'auth/js/main': 'auth/js/main.min',
 34 |             custom : '/custom',
 35 |             nbextensions : '/nbextensions',
 36 |             kernelspecs : '/kernelspecs',
 37 |             underscore : 'components/underscore/underscore-min',
 38 |             backbone : 'components/backbone/backbone-min',
 39 |             jed: 'components/jed/jed',
 40 |             jquery: 'components/jquery/jquery.min',
 41 |             json: 'components/requirejs-plugins/src/json',
 42 |             text: 'components/requirejs-text/text',
 43 |             bootstrap: 'components/bootstrap/dist/js/bootstrap.min',
 44 |             bootstraptour: 'components/bootstrap-tour/build/js/bootstrap-tour.min',
 45 |             'jquery-ui': 'components/jquery-ui/jquery-ui.min',
 46 |             moment: 'components/moment/min/moment-with-locales',
 47 |             codemirror: 'components/codemirror',
 48 |             termjs: 'components/xterm.js/xterm',
 49 |             typeahead: 'components/jquery-typeahead/dist/jquery.typeahead.min',
 50 |           },
 51 |           map: { // for backward compatibility
 52 |               "*": {
 53 |                   "jqueryui": "jquery-ui",
 54 |               }
 55 |           },
 56 |           shim: {
 57 |             typeahead: {
 58 |               deps: ["jquery"],
 59 |               exports: "typeahead"
 60 |             },
 61 |             underscore: {
 62 |               exports: '_'
 63 |             },
 64 |             backbone: {
 65 |               deps: ["underscore", "jquery"],
 66 |               exports: "Backbone"
 67 |             },
 68 |             bootstrap: {
 69 |               deps: ["jquery"],
 70 |               exports: "bootstrap"
 71 |             },
 72 |             bootstraptour: {
 73 |               deps: ["bootstrap"],
 74 |               exports: "Tour"
 75 |             },
 76 |             "jquery-ui": {
 77 |               deps: ["jquery"],
 78 |               exports: "$"
 79 |             }
 80 |           },
 81 |           waitSeconds: 30,
 82 |       });
 83 | 
 84 |       require.config({
 85 |           map: {
 86 |               '*':{
 87 |                 'contents': 'services/contents',
 88 |               }
 89 |           }
 90 |       });
 91 | 
 92 |       // error-catching custom.js shim.
 93 |       define("custom", function (require, exports, module) {
 94 |           try {
 95 |               var custom = require('custom/custom');
 96 |               console.debug('loaded custom.js');
 97 |               return custom;
 98 |           } catch (e) {
 99 |               console.error("error loading custom.js", e);
100 |               return {};
101 |           }
102 |       })
103 | 
104 |     document.nbjs_translations = {"domain": "nbjs", "locale_data": {"nbjs": {"": {"domain": "nbjs"}}}};
105 |     document.documentElement.lang = navigator.language.toLowerCase();
106 |     </script>
107 | 
108 |     
109 |     
110 | 
111 | </head>
112 | 
113 | <body class=""
114 |  
115 |   
116 |  
117 | dir="ltr">
118 | 
119 | <noscript>
120 |     <div id='noscript'>
121 |       Jupyter Notebook requires JavaScript.<br>
122 |       Please enable it to proceed. 
123 |   </div>
124 | </noscript>
125 | 
126 | <div id="header">
127 |   <div id="header-container" class="container">
128 |   <div id="ipython_notebook" class="nav navbar-brand"><a href="/lab" title='dashboard'>
129 |       <img src='/static/base/images/logo.png?v=641991992878ee24c6f3826e81054a0f' alt='Jupyter Notebook'/>
130 |   </a></div>
131 | 
132 |   
133 |   
134 |   
135 |   
136 |   
137 |   
138 | 
139 | 
140 |   
141 |   
142 |   </div>
143 |   <div class="header-bar"></div>
144 | 
145 |   
146 |   
147 | </div>
148 | 
149 | <div id="site">
150 | 
151 | 
152 | <div id="ipython-main-app" class="container">
153 | 
154 |     
155 |     
156 |     <div class="row">
157 |     <div class="navbar col-sm-8">
158 |       <div class="navbar-inner">
159 |         <div class="container">
160 |           <div class="center-nav">
161 |             <form action="/login?next=%2Flab%2Ftree%2Fcudf%2F01-Intro%2520to%2520cuDF.ipynb" method="post" class="navbar-form pull-left">
162 |               <input type="hidden" name="_xsrf" value="2|37a4ae7b|f42c9c96933f9b02156532f989cbd65a|1562679386"/>
163 |               
164 |                 <label for="password_input"><strong>Password:</strong></label>
165 |               
166 |               <input type="password" name="password" id="password_input" class="form-control">
167 |               <button type="submit" class="btn btn-default" id="login_submit">Log in</button>
168 |             </form>
169 |           </div>
170 |         </div>
171 |       </div>
172 |     </div>
173 |     </div>
174 |     
175 |     
176 |     
177 | </div>
178 | 
179 | 
180 | </div>
181 | 
182 | 
183 | 
184 | 
185 | 
186 | 
187 | 
188 | 
189 | <script type="text/javascript">
190 |   require(["auth/js/main"], function (auth) {
191 |     auth.login_main();
192 |   });
193 | </script>
194 | 
195 | 
196 | 
197 | <script type='text/javascript'>
198 |   function _remove_token_from_url() {
199 |     if (window.location.search.length <= 1) {
200 |       return;
201 |     }
202 |     var search_parameters = window.location.search.slice(1).split('&');
203 |     for (var i = 0; i < search_parameters.length; i++) {
204 |       if (search_parameters[i].split('=')[0] === 'token') {
205 |         // remote token from search parameters
206 |         search_parameters.splice(i, 1);
207 |         var new_search = '';
208 |         if (search_parameters.length) {
209 |           new_search = '?' + search_parameters.join('&');
210 |         }
211 |         var new_url = window.location.origin + 
212 |                       window.location.pathname + 
213 |                       new_search + 
214 |                       window.location.hash;
215 |         window.history.replaceState({}, "", new_url);
216 |         return;
217 |       }
218 |     }
219 |   }
220 |   _remove_token_from_url();
221 | </script>
222 | </body>
223 | 
224 | </html>


--------------------------------------------------------------------------------
/gpu/01-Intro_to_cuGraph.ipynb:
--------------------------------------------------------------------------------
  1 | <!DOCTYPE HTML>
  2 | <html>
  3 | 
  4 | <head>
  5 |     <meta charset="utf-8">
  6 | 
  7 |     <title>Jupyter Notebook</title>
  8 |     <link id="favicon" rel="shortcut icon" type="image/x-icon" href="/static/base/images/favicon.ico?v=97c6417ed01bdc0ae3ef32ae4894fd03">
  9 |     <meta http-equiv="X-UA-Compatible" content="IE=edge" />
 10 |     <link rel="stylesheet" href="/static/components/jquery-ui/themes/smoothness/jquery-ui.min.css?v=3c2a865c832a1322285c55c6ed99abb2" type="text/css" />
 11 |     <link rel="stylesheet" href="/static/components/jquery-typeahead/dist/jquery.typeahead.min.css?v=7afb461de36accb1aa133a1710f5bc56" type="text/css" />
 12 |     <meta name="viewport" content="width=device-width, initial-scale=1.0">
 13 |     
 14 |     
 15 | 
 16 |     <link rel="stylesheet" href="/static/style/style.min.css?v=56cad67b8b1ff3ed9b48850e9efcffb3" type="text/css"/>
 17 |     
 18 | <link rel="stylesheet" href="/static/auth/css/override.css?v=19ec59d2c4f1203c49fe47462028cd9a" type="text/css" />
 19 | 
 20 |     <link rel="stylesheet" href="/custom/custom.css" type="text/css" />
 21 |     <script src="/static/components/es6-promise/promise.min.js?v=f004a16cb856e0ff11781d01ec5ca8fe" type="text/javascript" charset="utf-8"></script>
 22 |     <script src="/static/components/preact/index.js?v=00a2fac73c670ce39ac53d26640eb542" type="text/javascript"></script>
 23 |     <script src="/static/components/proptypes/index.js?v=c40890eb04df9811fcc4d47e53a29604" type="text/javascript"></script>
 24 |     <script src="/static/components/preact-compat/index.js?v=aea8f6660e54b18ace8d84a9b9654c1c" type="text/javascript"></script>
 25 |     <script src="/static/components/requirejs/require.js?v=951f856e81496aaeec2e71a1c2c0d51f" type="text/javascript" charset="utf-8"></script>
 26 |     <script>
 27 |       require.config({
 28 |           
 29 |           urlArgs: "v=20190709124529",
 30 |           
 31 |           baseUrl: '/static/',
 32 |           paths: {
 33 |             'auth/js/main': 'auth/js/main.min',
 34 |             custom : '/custom',
 35 |             nbextensions : '/nbextensions',
 36 |             kernelspecs : '/kernelspecs',
 37 |             underscore : 'components/underscore/underscore-min',
 38 |             backbone : 'components/backbone/backbone-min',
 39 |             jed: 'components/jed/jed',
 40 |             jquery: 'components/jquery/jquery.min',
 41 |             json: 'components/requirejs-plugins/src/json',
 42 |             text: 'components/requirejs-text/text',
 43 |             bootstrap: 'components/bootstrap/dist/js/bootstrap.min',
 44 |             bootstraptour: 'components/bootstrap-tour/build/js/bootstrap-tour.min',
 45 |             'jquery-ui': 'components/jquery-ui/jquery-ui.min',
 46 |             moment: 'components/moment/min/moment-with-locales',
 47 |             codemirror: 'components/codemirror',
 48 |             termjs: 'components/xterm.js/xterm',
 49 |             typeahead: 'components/jquery-typeahead/dist/jquery.typeahead.min',
 50 |           },
 51 |           map: { // for backward compatibility
 52 |               "*": {
 53 |                   "jqueryui": "jquery-ui",
 54 |               }
 55 |           },
 56 |           shim: {
 57 |             typeahead: {
 58 |               deps: ["jquery"],
 59 |               exports: "typeahead"
 60 |             },
 61 |             underscore: {
 62 |               exports: '_'
 63 |             },
 64 |             backbone: {
 65 |               deps: ["underscore", "jquery"],
 66 |               exports: "Backbone"
 67 |             },
 68 |             bootstrap: {
 69 |               deps: ["jquery"],
 70 |               exports: "bootstrap"
 71 |             },
 72 |             bootstraptour: {
 73 |               deps: ["bootstrap"],
 74 |               exports: "Tour"
 75 |             },
 76 |             "jquery-ui": {
 77 |               deps: ["jquery"],
 78 |               exports: "$"
 79 |             }
 80 |           },
 81 |           waitSeconds: 30,
 82 |       });
 83 | 
 84 |       require.config({
 85 |           map: {
 86 |               '*':{
 87 |                 'contents': 'services/contents',
 88 |               }
 89 |           }
 90 |       });
 91 | 
 92 |       // error-catching custom.js shim.
 93 |       define("custom", function (require, exports, module) {
 94 |           try {
 95 |               var custom = require('custom/custom');
 96 |               console.debug('loaded custom.js');
 97 |               return custom;
 98 |           } catch (e) {
 99 |               console.error("error loading custom.js", e);
100 |               return {};
101 |           }
102 |       })
103 | 
104 |     document.nbjs_translations = {"domain": "nbjs", "locale_data": {"nbjs": {"": {"domain": "nbjs"}}}};
105 |     document.documentElement.lang = navigator.language.toLowerCase();
106 |     </script>
107 | 
108 |     
109 |     
110 | 
111 | </head>
112 | 
113 | <body class=""
114 |  
115 |   
116 |  
117 | dir="ltr">
118 | 
119 | <noscript>
120 |     <div id='noscript'>
121 |       Jupyter Notebook requires JavaScript.<br>
122 |       Please enable it to proceed. 
123 |   </div>
124 | </noscript>
125 | 
126 | <div id="header">
127 |   <div id="header-container" class="container">
128 |   <div id="ipython_notebook" class="nav navbar-brand"><a href="/lab" title='dashboard'>
129 |       <img src='/static/base/images/logo.png?v=641991992878ee24c6f3826e81054a0f' alt='Jupyter Notebook'/>
130 |   </a></div>
131 | 
132 |   
133 |   
134 |   
135 |   
136 |   
137 |   
138 | 
139 | 
140 |   
141 |   
142 |   </div>
143 |   <div class="header-bar"></div>
144 | 
145 |   
146 |   
147 | </div>
148 | 
149 | <div id="site">
150 | 
151 | 
152 | <div id="ipython-main-app" class="container">
153 | 
154 |     
155 |     
156 |     <div class="row">
157 |     <div class="navbar col-sm-8">
158 |       <div class="navbar-inner">
159 |         <div class="container">
160 |           <div class="center-nav">
161 |             <form action="/login?next=%2Flab%2Ftree%2Fcugraph%2F01-Intro%2520to%2520cuGraph.ipynb" method="post" class="navbar-form pull-left">
162 |               <input type="hidden" name="_xsrf" value="2|37c78d09|0fbc66b95a84c049290457456cae2f7a|1562680568"/>
163 |               
164 |                 <label for="password_input"><strong>Password:</strong></label>
165 |               
166 |               <input type="password" name="password" id="password_input" class="form-control">
167 |               <button type="submit" class="btn btn-default" id="login_submit">Log in</button>
168 |             </form>
169 |           </div>
170 |         </div>
171 |       </div>
172 |     </div>
173 |     </div>
174 |     
175 |     
176 |     
177 | </div>
178 | 
179 | 
180 | </div>
181 | 
182 | 
183 | 
184 | 
185 | 
186 | 
187 | 
188 | 
189 | <script type="text/javascript">
190 |   require(["auth/js/main"], function (auth) {
191 |     auth.login_main();
192 |   });
193 | </script>
194 | 
195 | 
196 | 
197 | <script type='text/javascript'>
198 |   function _remove_token_from_url() {
199 |     if (window.location.search.length <= 1) {
200 |       return;
201 |     }
202 |     var search_parameters = window.location.search.slice(1).split('&');
203 |     for (var i = 0; i < search_parameters.length; i++) {
204 |       if (search_parameters[i].split('=')[0] === 'token') {
205 |         // remote token from search parameters
206 |         search_parameters.splice(i, 1);
207 |         var new_search = '';
208 |         if (search_parameters.length) {
209 |           new_search = '?' + search_parameters.join('&');
210 |         }
211 |         var new_url = window.location.origin + 
212 |                       window.location.pathname + 
213 |                       new_search + 
214 |                       window.location.hash;
215 |         window.history.replaceState({}, "", new_url);
216 |         return;
217 |       }
218 |     }
219 |   }
220 |   _remove_token_from_url();
221 | </script>
222 | </body>
223 | 
224 | </html>


--------------------------------------------------------------------------------
/gpu/01-Introduction-LinearRegression-Hyperparam.ipynb:
--------------------------------------------------------------------------------
  1 | <!DOCTYPE HTML>
  2 | <html>
  3 | 
  4 | <head>
  5 |     <meta charset="utf-8">
  6 | 
  7 |     <title>Jupyter Notebook</title>
  8 |     <link id="favicon" rel="shortcut icon" type="image/x-icon" href="/static/base/images/favicon.ico?v=97c6417ed01bdc0ae3ef32ae4894fd03">
  9 |     <meta http-equiv="X-UA-Compatible" content="IE=edge" />
 10 |     <link rel="stylesheet" href="/static/components/jquery-ui/themes/smoothness/jquery-ui.min.css?v=3c2a865c832a1322285c55c6ed99abb2" type="text/css" />
 11 |     <link rel="stylesheet" href="/static/components/jquery-typeahead/dist/jquery.typeahead.min.css?v=7afb461de36accb1aa133a1710f5bc56" type="text/css" />
 12 |     <meta name="viewport" content="width=device-width, initial-scale=1.0">
 13 |     
 14 |     
 15 | 
 16 |     <link rel="stylesheet" href="/static/style/style.min.css?v=56cad67b8b1ff3ed9b48850e9efcffb3" type="text/css"/>
 17 |     
 18 | <link rel="stylesheet" href="/static/auth/css/override.css?v=19ec59d2c4f1203c49fe47462028cd9a" type="text/css" />
 19 | 
 20 |     <link rel="stylesheet" href="/custom/custom.css" type="text/css" />
 21 |     <script src="/static/components/es6-promise/promise.min.js?v=f004a16cb856e0ff11781d01ec5ca8fe" type="text/javascript" charset="utf-8"></script>
 22 |     <script src="/static/components/preact/index.js?v=00a2fac73c670ce39ac53d26640eb542" type="text/javascript"></script>
 23 |     <script src="/static/components/proptypes/index.js?v=c40890eb04df9811fcc4d47e53a29604" type="text/javascript"></script>
 24 |     <script src="/static/components/preact-compat/index.js?v=aea8f6660e54b18ace8d84a9b9654c1c" type="text/javascript"></script>
 25 |     <script src="/static/components/requirejs/require.js?v=951f856e81496aaeec2e71a1c2c0d51f" type="text/javascript" charset="utf-8"></script>
 26 |     <script>
 27 |       require.config({
 28 |           
 29 |           urlArgs: "v=20190709124529",
 30 |           
 31 |           baseUrl: '/static/',
 32 |           paths: {
 33 |             'auth/js/main': 'auth/js/main.min',
 34 |             custom : '/custom',
 35 |             nbextensions : '/nbextensions',
 36 |             kernelspecs : '/kernelspecs',
 37 |             underscore : 'components/underscore/underscore-min',
 38 |             backbone : 'components/backbone/backbone-min',
 39 |             jed: 'components/jed/jed',
 40 |             jquery: 'components/jquery/jquery.min',
 41 |             json: 'components/requirejs-plugins/src/json',
 42 |             text: 'components/requirejs-text/text',
 43 |             bootstrap: 'components/bootstrap/dist/js/bootstrap.min',
 44 |             bootstraptour: 'components/bootstrap-tour/build/js/bootstrap-tour.min',
 45 |             'jquery-ui': 'components/jquery-ui/jquery-ui.min',
 46 |             moment: 'components/moment/min/moment-with-locales',
 47 |             codemirror: 'components/codemirror',
 48 |             termjs: 'components/xterm.js/xterm',
 49 |             typeahead: 'components/jquery-typeahead/dist/jquery.typeahead.min',
 50 |           },
 51 |           map: { // for backward compatibility
 52 |               "*": {
 53 |                   "jqueryui": "jquery-ui",
 54 |               }
 55 |           },
 56 |           shim: {
 57 |             typeahead: {
 58 |               deps: ["jquery"],
 59 |               exports: "typeahead"
 60 |             },
 61 |             underscore: {
 62 |               exports: '_'
 63 |             },
 64 |             backbone: {
 65 |               deps: ["underscore", "jquery"],
 66 |               exports: "Backbone"
 67 |             },
 68 |             bootstrap: {
 69 |               deps: ["jquery"],
 70 |               exports: "bootstrap"
 71 |             },
 72 |             bootstraptour: {
 73 |               deps: ["bootstrap"],
 74 |               exports: "Tour"
 75 |             },
 76 |             "jquery-ui": {
 77 |               deps: ["jquery"],
 78 |               exports: "$"
 79 |             }
 80 |           },
 81 |           waitSeconds: 30,
 82 |       });
 83 | 
 84 |       require.config({
 85 |           map: {
 86 |               '*':{
 87 |                 'contents': 'services/contents',
 88 |               }
 89 |           }
 90 |       });
 91 | 
 92 |       // error-catching custom.js shim.
 93 |       define("custom", function (require, exports, module) {
 94 |           try {
 95 |               var custom = require('custom/custom');
 96 |               console.debug('loaded custom.js');
 97 |               return custom;
 98 |           } catch (e) {
 99 |               console.error("error loading custom.js", e);
100 |               return {};
101 |           }
102 |       })
103 | 
104 |     document.nbjs_translations = {"domain": "nbjs", "locale_data": {"nbjs": {"": {"domain": "nbjs"}}}};
105 |     document.documentElement.lang = navigator.language.toLowerCase();
106 |     </script>
107 | 
108 |     
109 |     
110 | 
111 | </head>
112 | 
113 | <body class=""
114 |  
115 |   
116 |  
117 | dir="ltr">
118 | 
119 | <noscript>
120 |     <div id='noscript'>
121 |       Jupyter Notebook requires JavaScript.<br>
122 |       Please enable it to proceed. 
123 |   </div>
124 | </noscript>
125 | 
126 | <div id="header">
127 |   <div id="header-container" class="container">
128 |   <div id="ipython_notebook" class="nav navbar-brand"><a href="/lab" title='dashboard'>
129 |       <img src='/static/base/images/logo.png?v=641991992878ee24c6f3826e81054a0f' alt='Jupyter Notebook'/>
130 |   </a></div>
131 | 
132 |   
133 |   
134 |   
135 |   
136 |   
137 |   
138 | 
139 | 
140 |   
141 |   
142 |   </div>
143 |   <div class="header-bar"></div>
144 | 
145 |   
146 |   
147 | </div>
148 | 
149 | <div id="site">
150 | 
151 | 
152 | <div id="ipython-main-app" class="container">
153 | 
154 |     
155 |     
156 |     <div class="row">
157 |     <div class="navbar col-sm-8">
158 |       <div class="navbar-inner">
159 |         <div class="container">
160 |           <div class="center-nav">
161 |             <form action="/login?next=%2Flab%2Ftree%2Fcuml%2F01-Introduction-LinearRegression-Hyperparam.ipynb" method="post" class="navbar-form pull-left">
162 |               <input type="hidden" name="_xsrf" value="2|a774466e|098d3bb37ccf67d834f9b50b8211efe9|1562680653"/>
163 |               
164 |                 <label for="password_input"><strong>Password:</strong></label>
165 |               
166 |               <input type="password" name="password" id="password_input" class="form-control">
167 |               <button type="submit" class="btn btn-default" id="login_submit">Log in</button>
168 |             </form>
169 |           </div>
170 |         </div>
171 |       </div>
172 |     </div>
173 |     </div>
174 |     
175 |     
176 |     
177 | </div>
178 | 
179 | 
180 | </div>
181 | 
182 | 
183 | 
184 | 
185 | 
186 | 
187 | 
188 | 
189 | <script type="text/javascript">
190 |   require(["auth/js/main"], function (auth) {
191 |     auth.login_main();
192 |   });
193 | </script>
194 | 
195 | 
196 | 
197 | <script type='text/javascript'>
198 |   function _remove_token_from_url() {
199 |     if (window.location.search.length <= 1) {
200 |       return;
201 |     }
202 |     var search_parameters = window.location.search.slice(1).split('&');
203 |     for (var i = 0; i < search_parameters.length; i++) {
204 |       if (search_parameters[i].split('=')[0] === 'token') {
205 |         // remote token from search parameters
206 |         search_parameters.splice(i, 1);
207 |         var new_search = '';
208 |         if (search_parameters.length) {
209 |           new_search = '?' + search_parameters.join('&');
210 |         }
211 |         var new_url = window.location.origin + 
212 |                       window.location.pathname + 
213 |                       new_search + 
214 |                       window.location.hash;
215 |         window.history.replaceState({}, "", new_url);
216 |         return;
217 |       }
218 |     }
219 |   }
220 |   _remove_token_from_url();
221 | </script>
222 | </body>
223 | 
224 | </html>


--------------------------------------------------------------------------------
/gpu/02-Intro_to_cuDF_UDFs.ipynb:
--------------------------------------------------------------------------------
  1 | <!DOCTYPE HTML>
  2 | <html>
  3 | 
  4 | <head>
  5 |     <meta charset="utf-8">
  6 | 
  7 |     <title>Jupyter Notebook</title>
  8 |     <link id="favicon" rel="shortcut icon" type="image/x-icon" href="/static/base/images/favicon.ico?v=97c6417ed01bdc0ae3ef32ae4894fd03">
  9 |     <meta http-equiv="X-UA-Compatible" content="IE=edge" />
 10 |     <link rel="stylesheet" href="/static/components/jquery-ui/themes/smoothness/jquery-ui.min.css?v=3c2a865c832a1322285c55c6ed99abb2" type="text/css" />
 11 |     <link rel="stylesheet" href="/static/components/jquery-typeahead/dist/jquery.typeahead.min.css?v=7afb461de36accb1aa133a1710f5bc56" type="text/css" />
 12 |     <meta name="viewport" content="width=device-width, initial-scale=1.0">
 13 |     
 14 |     
 15 | 
 16 |     <link rel="stylesheet" href="/static/style/style.min.css?v=56cad67b8b1ff3ed9b48850e9efcffb3" type="text/css"/>
 17 |     
 18 | <link rel="stylesheet" href="/static/auth/css/override.css?v=19ec59d2c4f1203c49fe47462028cd9a" type="text/css" />
 19 | 
 20 |     <link rel="stylesheet" href="/custom/custom.css" type="text/css" />
 21 |     <script src="/static/components/es6-promise/promise.min.js?v=f004a16cb856e0ff11781d01ec5ca8fe" type="text/javascript" charset="utf-8"></script>
 22 |     <script src="/static/components/preact/index.js?v=00a2fac73c670ce39ac53d26640eb542" type="text/javascript"></script>
 23 |     <script src="/static/components/proptypes/index.js?v=c40890eb04df9811fcc4d47e53a29604" type="text/javascript"></script>
 24 |     <script src="/static/components/preact-compat/index.js?v=aea8f6660e54b18ace8d84a9b9654c1c" type="text/javascript"></script>
 25 |     <script src="/static/components/requirejs/require.js?v=951f856e81496aaeec2e71a1c2c0d51f" type="text/javascript" charset="utf-8"></script>
 26 |     <script>
 27 |       require.config({
 28 |           
 29 |           urlArgs: "v=20190709124529",
 30 |           
 31 |           baseUrl: '/static/',
 32 |           paths: {
 33 |             'auth/js/main': 'auth/js/main.min',
 34 |             custom : '/custom',
 35 |             nbextensions : '/nbextensions',
 36 |             kernelspecs : '/kernelspecs',
 37 |             underscore : 'components/underscore/underscore-min',
 38 |             backbone : 'components/backbone/backbone-min',
 39 |             jed: 'components/jed/jed',
 40 |             jquery: 'components/jquery/jquery.min',
 41 |             json: 'components/requirejs-plugins/src/json',
 42 |             text: 'components/requirejs-text/text',
 43 |             bootstrap: 'components/bootstrap/dist/js/bootstrap.min',
 44 |             bootstraptour: 'components/bootstrap-tour/build/js/bootstrap-tour.min',
 45 |             'jquery-ui': 'components/jquery-ui/jquery-ui.min',
 46 |             moment: 'components/moment/min/moment-with-locales',
 47 |             codemirror: 'components/codemirror',
 48 |             termjs: 'components/xterm.js/xterm',
 49 |             typeahead: 'components/jquery-typeahead/dist/jquery.typeahead.min',
 50 |           },
 51 |           map: { // for backward compatibility
 52 |               "*": {
 53 |                   "jqueryui": "jquery-ui",
 54 |               }
 55 |           },
 56 |           shim: {
 57 |             typeahead: {
 58 |               deps: ["jquery"],
 59 |               exports: "typeahead"
 60 |             },
 61 |             underscore: {
 62 |               exports: '_'
 63 |             },
 64 |             backbone: {
 65 |               deps: ["underscore", "jquery"],
 66 |               exports: "Backbone"
 67 |             },
 68 |             bootstrap: {
 69 |               deps: ["jquery"],
 70 |               exports: "bootstrap"
 71 |             },
 72 |             bootstraptour: {
 73 |               deps: ["bootstrap"],
 74 |               exports: "Tour"
 75 |             },
 76 |             "jquery-ui": {
 77 |               deps: ["jquery"],
 78 |               exports: "$"
 79 |             }
 80 |           },
 81 |           waitSeconds: 30,
 82 |       });
 83 | 
 84 |       require.config({
 85 |           map: {
 86 |               '*':{
 87 |                 'contents': 'services/contents',
 88 |               }
 89 |           }
 90 |       });
 91 | 
 92 |       // error-catching custom.js shim.
 93 |       define("custom", function (require, exports, module) {
 94 |           try {
 95 |               var custom = require('custom/custom');
 96 |               console.debug('loaded custom.js');
 97 |               return custom;
 98 |           } catch (e) {
 99 |               console.error("error loading custom.js", e);
100 |               return {};
101 |           }
102 |       })
103 | 
104 |     document.nbjs_translations = {"domain": "nbjs", "locale_data": {"nbjs": {"": {"domain": "nbjs"}}}};
105 |     document.documentElement.lang = navigator.language.toLowerCase();
106 |     </script>
107 | 
108 |     
109 |     
110 | 
111 | </head>
112 | 
113 | <body class=""
114 |  
115 |   
116 |  
117 | dir="ltr">
118 | 
119 | <noscript>
120 |     <div id='noscript'>
121 |       Jupyter Notebook requires JavaScript.<br>
122 |       Please enable it to proceed. 
123 |   </div>
124 | </noscript>
125 | 
126 | <div id="header">
127 |   <div id="header-container" class="container">
128 |   <div id="ipython_notebook" class="nav navbar-brand"><a href="/lab" title='dashboard'>
129 |       <img src='/static/base/images/logo.png?v=641991992878ee24c6f3826e81054a0f' alt='Jupyter Notebook'/>
130 |   </a></div>
131 | 
132 |   
133 |   
134 |   
135 |   
136 |   
137 |   
138 | 
139 | 
140 |   
141 |   
142 |   </div>
143 |   <div class="header-bar"></div>
144 | 
145 |   
146 |   
147 | </div>
148 | 
149 | <div id="site">
150 | 
151 | 
152 | <div id="ipython-main-app" class="container">
153 | 
154 |     
155 |     
156 |     <div class="row">
157 |     <div class="navbar col-sm-8">
158 |       <div class="navbar-inner">
159 |         <div class="container">
160 |           <div class="center-nav">
161 |             <form action="/login?next=%2Flab%2Ftree%2Fcudf%2F02-Intro%2520to%2520cuDF%2520UDFs.ipynb" method="post" class="navbar-form pull-left">
162 |               <input type="hidden" name="_xsrf" value="2|f1184747|c33072bcef514f631450fe2a7eae6bc8|1562679407"/>
163 |               
164 |                 <label for="password_input"><strong>Password:</strong></label>
165 |               
166 |               <input type="password" name="password" id="password_input" class="form-control">
167 |               <button type="submit" class="btn btn-default" id="login_submit">Log in</button>
168 |             </form>
169 |           </div>
170 |         </div>
171 |       </div>
172 |     </div>
173 |     </div>
174 |     
175 |     
176 |     
177 | </div>
178 | 
179 | 
180 | </div>
181 | 
182 | 
183 | 
184 | 
185 | 
186 | 
187 | 
188 | 
189 | <script type="text/javascript">
190 |   require(["auth/js/main"], function (auth) {
191 |     auth.login_main();
192 |   });
193 | </script>
194 | 
195 | 
196 | 
197 | <script type='text/javascript'>
198 |   function _remove_token_from_url() {
199 |     if (window.location.search.length <= 1) {
200 |       return;
201 |     }
202 |     var search_parameters = window.location.search.slice(1).split('&');
203 |     for (var i = 0; i < search_parameters.length; i++) {
204 |       if (search_parameters[i].split('=')[0] === 'token') {
205 |         // remote token from search parameters
206 |         search_parameters.splice(i, 1);
207 |         var new_search = '';
208 |         if (search_parameters.length) {
209 |           new_search = '?' + search_parameters.join('&');
210 |         }
211 |         var new_url = window.location.origin + 
212 |                       window.location.pathname + 
213 |                       new_search + 
214 |                       window.location.hash;
215 |         window.history.replaceState({}, "", new_url);
216 |         return;
217 |       }
218 |     }
219 |   }
220 |   _remove_token_from_url();
221 | </script>
222 | </body>
223 | 
224 | </html>


--------------------------------------------------------------------------------
/gpu/02-LogisticRegression.ipynb:
--------------------------------------------------------------------------------
  1 | <!DOCTYPE HTML>
  2 | <html>
  3 | 
  4 | <head>
  5 |     <meta charset="utf-8">
  6 | 
  7 |     <title>Jupyter Notebook</title>
  8 |     <link id="favicon" rel="shortcut icon" type="image/x-icon" href="/static/base/images/favicon.ico?v=97c6417ed01bdc0ae3ef32ae4894fd03">
  9 |     <meta http-equiv="X-UA-Compatible" content="IE=edge" />
 10 |     <link rel="stylesheet" href="/static/components/jquery-ui/themes/smoothness/jquery-ui.min.css?v=3c2a865c832a1322285c55c6ed99abb2" type="text/css" />
 11 |     <link rel="stylesheet" href="/static/components/jquery-typeahead/dist/jquery.typeahead.min.css?v=7afb461de36accb1aa133a1710f5bc56" type="text/css" />
 12 |     <meta name="viewport" content="width=device-width, initial-scale=1.0">
 13 |     
 14 |     
 15 | 
 16 |     <link rel="stylesheet" href="/static/style/style.min.css?v=56cad67b8b1ff3ed9b48850e9efcffb3" type="text/css"/>
 17 |     
 18 | <link rel="stylesheet" href="/static/auth/css/override.css?v=19ec59d2c4f1203c49fe47462028cd9a" type="text/css" />
 19 | 
 20 |     <link rel="stylesheet" href="/custom/custom.css" type="text/css" />
 21 |     <script src="/static/components/es6-promise/promise.min.js?v=f004a16cb856e0ff11781d01ec5ca8fe" type="text/javascript" charset="utf-8"></script>
 22 |     <script src="/static/components/preact/index.js?v=00a2fac73c670ce39ac53d26640eb542" type="text/javascript"></script>
 23 |     <script src="/static/components/proptypes/index.js?v=c40890eb04df9811fcc4d47e53a29604" type="text/javascript"></script>
 24 |     <script src="/static/components/preact-compat/index.js?v=aea8f6660e54b18ace8d84a9b9654c1c" type="text/javascript"></script>
 25 |     <script src="/static/components/requirejs/require.js?v=951f856e81496aaeec2e71a1c2c0d51f" type="text/javascript" charset="utf-8"></script>
 26 |     <script>
 27 |       require.config({
 28 |           
 29 |           urlArgs: "v=20190709124529",
 30 |           
 31 |           baseUrl: '/static/',
 32 |           paths: {
 33 |             'auth/js/main': 'auth/js/main.min',
 34 |             custom : '/custom',
 35 |             nbextensions : '/nbextensions',
 36 |             kernelspecs : '/kernelspecs',
 37 |             underscore : 'components/underscore/underscore-min',
 38 |             backbone : 'components/backbone/backbone-min',
 39 |             jed: 'components/jed/jed',
 40 |             jquery: 'components/jquery/jquery.min',
 41 |             json: 'components/requirejs-plugins/src/json',
 42 |             text: 'components/requirejs-text/text',
 43 |             bootstrap: 'components/bootstrap/dist/js/bootstrap.min',
 44 |             bootstraptour: 'components/bootstrap-tour/build/js/bootstrap-tour.min',
 45 |             'jquery-ui': 'components/jquery-ui/jquery-ui.min',
 46 |             moment: 'components/moment/min/moment-with-locales',
 47 |             codemirror: 'components/codemirror',
 48 |             termjs: 'components/xterm.js/xterm',
 49 |             typeahead: 'components/jquery-typeahead/dist/jquery.typeahead.min',
 50 |           },
 51 |           map: { // for backward compatibility
 52 |               "*": {
 53 |                   "jqueryui": "jquery-ui",
 54 |               }
 55 |           },
 56 |           shim: {
 57 |             typeahead: {
 58 |               deps: ["jquery"],
 59 |               exports: "typeahead"
 60 |             },
 61 |             underscore: {
 62 |               exports: '_'
 63 |             },
 64 |             backbone: {
 65 |               deps: ["underscore", "jquery"],
 66 |               exports: "Backbone"
 67 |             },
 68 |             bootstrap: {
 69 |               deps: ["jquery"],
 70 |               exports: "bootstrap"
 71 |             },
 72 |             bootstraptour: {
 73 |               deps: ["bootstrap"],
 74 |               exports: "Tour"
 75 |             },
 76 |             "jquery-ui": {
 77 |               deps: ["jquery"],
 78 |               exports: "$"
 79 |             }
 80 |           },
 81 |           waitSeconds: 30,
 82 |       });
 83 | 
 84 |       require.config({
 85 |           map: {
 86 |               '*':{
 87 |                 'contents': 'services/contents',
 88 |               }
 89 |           }
 90 |       });
 91 | 
 92 |       // error-catching custom.js shim.
 93 |       define("custom", function (require, exports, module) {
 94 |           try {
 95 |               var custom = require('custom/custom');
 96 |               console.debug('loaded custom.js');
 97 |               return custom;
 98 |           } catch (e) {
 99 |               console.error("error loading custom.js", e);
100 |               return {};
101 |           }
102 |       })
103 | 
104 |     document.nbjs_translations = {"domain": "nbjs", "locale_data": {"nbjs": {"": {"domain": "nbjs"}}}};
105 |     document.documentElement.lang = navigator.language.toLowerCase();
106 |     </script>
107 | 
108 |     
109 |     
110 | 
111 | </head>
112 | 
113 | <body class=""
114 |  
115 |   
116 |  
117 | dir="ltr">
118 | 
119 | <noscript>
120 |     <div id='noscript'>
121 |       Jupyter Notebook requires JavaScript.<br>
122 |       Please enable it to proceed. 
123 |   </div>
124 | </noscript>
125 | 
126 | <div id="header">
127 |   <div id="header-container" class="container">
128 |   <div id="ipython_notebook" class="nav navbar-brand"><a href="/lab" title='dashboard'>
129 |       <img src='/static/base/images/logo.png?v=641991992878ee24c6f3826e81054a0f' alt='Jupyter Notebook'/>
130 |   </a></div>
131 | 
132 |   
133 |   
134 |   
135 |   
136 |   
137 |   
138 | 
139 | 
140 |   
141 |   
142 |   </div>
143 |   <div class="header-bar"></div>
144 | 
145 |   
146 |   
147 | </div>
148 | 
149 | <div id="site">
150 | 
151 | 
152 | <div id="ipython-main-app" class="container">
153 | 
154 |     
155 |     
156 |     <div class="row">
157 |     <div class="navbar col-sm-8">
158 |       <div class="navbar-inner">
159 |         <div class="container">
160 |           <div class="center-nav">
161 |             <form action="/login?next=%2Flab%2Ftree%2Fcuml%2F02-LogisticRegression.ipynb" method="post" class="navbar-form pull-left">
162 |               <input type="hidden" name="_xsrf" value="2|0f57ddb2|4e96990420c0d32e7030e433e572e8a2|1562680666"/>
163 |               
164 |                 <label for="password_input"><strong>Password:</strong></label>
165 |               
166 |               <input type="password" name="password" id="password_input" class="form-control">
167 |               <button type="submit" class="btn btn-default" id="login_submit">Log in</button>
168 |             </form>
169 |           </div>
170 |         </div>
171 |       </div>
172 |     </div>
173 |     </div>
174 |     
175 |     
176 |     
177 | </div>
178 | 
179 | 
180 | </div>
181 | 
182 | 
183 | 
184 | 
185 | 
186 | 
187 | 
188 | 
189 | <script type="text/javascript">
190 |   require(["auth/js/main"], function (auth) {
191 |     auth.login_main();
192 |   });
193 | </script>
194 | 
195 | 
196 | 
197 | <script type='text/javascript'>
198 |   function _remove_token_from_url() {
199 |     if (window.location.search.length <= 1) {
200 |       return;
201 |     }
202 |     var search_parameters = window.location.search.slice(1).split('&');
203 |     for (var i = 0; i < search_parameters.length; i++) {
204 |       if (search_parameters[i].split('=')[0] === 'token') {
205 |         // remote token from search parameters
206 |         search_parameters.splice(i, 1);
207 |         var new_search = '';
208 |         if (search_parameters.length) {
209 |           new_search = '?' + search_parameters.join('&');
210 |         }
211 |         var new_url = window.location.origin + 
212 |                       window.location.pathname + 
213 |                       new_search + 
214 |                       window.location.hash;
215 |         window.history.replaceState({}, "", new_url);
216 |         return;
217 |       }
218 |     }
219 |   }
220 |   _remove_token_from_url();
221 | </script>
222 | </body>
223 | 
224 | </html>


--------------------------------------------------------------------------------
/gpu/02-Louvain.ipynb:
--------------------------------------------------------------------------------
  1 | <!DOCTYPE HTML>
  2 | <html>
  3 | 
  4 | <head>
  5 |     <meta charset="utf-8">
  6 | 
  7 |     <title>Jupyter Notebook</title>
  8 |     <link id="favicon" rel="shortcut icon" type="image/x-icon" href="/static/base/images/favicon.ico?v=97c6417ed01bdc0ae3ef32ae4894fd03">
  9 |     <meta http-equiv="X-UA-Compatible" content="IE=edge" />
 10 |     <link rel="stylesheet" href="/static/components/jquery-ui/themes/smoothness/jquery-ui.min.css?v=3c2a865c832a1322285c55c6ed99abb2" type="text/css" />
 11 |     <link rel="stylesheet" href="/static/components/jquery-typeahead/dist/jquery.typeahead.min.css?v=7afb461de36accb1aa133a1710f5bc56" type="text/css" />
 12 |     <meta name="viewport" content="width=device-width, initial-scale=1.0">
 13 |     
 14 |     
 15 | 
 16 |     <link rel="stylesheet" href="/static/style/style.min.css?v=56cad67b8b1ff3ed9b48850e9efcffb3" type="text/css"/>
 17 |     
 18 | <link rel="stylesheet" href="/static/auth/css/override.css?v=19ec59d2c4f1203c49fe47462028cd9a" type="text/css" />
 19 | 
 20 |     <link rel="stylesheet" href="/custom/custom.css" type="text/css" />
 21 |     <script src="/static/components/es6-promise/promise.min.js?v=f004a16cb856e0ff11781d01ec5ca8fe" type="text/javascript" charset="utf-8"></script>
 22 |     <script src="/static/components/preact/index.js?v=00a2fac73c670ce39ac53d26640eb542" type="text/javascript"></script>
 23 |     <script src="/static/components/proptypes/index.js?v=c40890eb04df9811fcc4d47e53a29604" type="text/javascript"></script>
 24 |     <script src="/static/components/preact-compat/index.js?v=aea8f6660e54b18ace8d84a9b9654c1c" type="text/javascript"></script>
 25 |     <script src="/static/components/requirejs/require.js?v=951f856e81496aaeec2e71a1c2c0d51f" type="text/javascript" charset="utf-8"></script>
 26 |     <script>
 27 |       require.config({
 28 |           
 29 |           urlArgs: "v=20190709124529",
 30 |           
 31 |           baseUrl: '/static/',
 32 |           paths: {
 33 |             'auth/js/main': 'auth/js/main.min',
 34 |             custom : '/custom',
 35 |             nbextensions : '/nbextensions',
 36 |             kernelspecs : '/kernelspecs',
 37 |             underscore : 'components/underscore/underscore-min',
 38 |             backbone : 'components/backbone/backbone-min',
 39 |             jed: 'components/jed/jed',
 40 |             jquery: 'components/jquery/jquery.min',
 41 |             json: 'components/requirejs-plugins/src/json',
 42 |             text: 'components/requirejs-text/text',
 43 |             bootstrap: 'components/bootstrap/dist/js/bootstrap.min',
 44 |             bootstraptour: 'components/bootstrap-tour/build/js/bootstrap-tour.min',
 45 |             'jquery-ui': 'components/jquery-ui/jquery-ui.min',
 46 |             moment: 'components/moment/min/moment-with-locales',
 47 |             codemirror: 'components/codemirror',
 48 |             termjs: 'components/xterm.js/xterm',
 49 |             typeahead: 'components/jquery-typeahead/dist/jquery.typeahead.min',
 50 |           },
 51 |           map: { // for backward compatibility
 52 |               "*": {
 53 |                   "jqueryui": "jquery-ui",
 54 |               }
 55 |           },
 56 |           shim: {
 57 |             typeahead: {
 58 |               deps: ["jquery"],
 59 |               exports: "typeahead"
 60 |             },
 61 |             underscore: {
 62 |               exports: '_'
 63 |             },
 64 |             backbone: {
 65 |               deps: ["underscore", "jquery"],
 66 |               exports: "Backbone"
 67 |             },
 68 |             bootstrap: {
 69 |               deps: ["jquery"],
 70 |               exports: "bootstrap"
 71 |             },
 72 |             bootstraptour: {
 73 |               deps: ["bootstrap"],
 74 |               exports: "Tour"
 75 |             },
 76 |             "jquery-ui": {
 77 |               deps: ["jquery"],
 78 |               exports: "$"
 79 |             }
 80 |           },
 81 |           waitSeconds: 30,
 82 |       });
 83 | 
 84 |       require.config({
 85 |           map: {
 86 |               '*':{
 87 |                 'contents': 'services/contents',
 88 |               }
 89 |           }
 90 |       });
 91 | 
 92 |       // error-catching custom.js shim.
 93 |       define("custom", function (require, exports, module) {
 94 |           try {
 95 |               var custom = require('custom/custom');
 96 |               console.debug('loaded custom.js');
 97 |               return custom;
 98 |           } catch (e) {
 99 |               console.error("error loading custom.js", e);
100 |               return {};
101 |           }
102 |       })
103 | 
104 |     document.nbjs_translations = {"domain": "nbjs", "locale_data": {"nbjs": {"": {"domain": "nbjs"}}}};
105 |     document.documentElement.lang = navigator.language.toLowerCase();
106 |     </script>
107 | 
108 |     
109 |     
110 | 
111 | </head>
112 | 
113 | <body class=""
114 |  
115 |   
116 |  
117 | dir="ltr">
118 | 
119 | <noscript>
120 |     <div id='noscript'>
121 |       Jupyter Notebook requires JavaScript.<br>
122 |       Please enable it to proceed. 
123 |   </div>
124 | </noscript>
125 | 
126 | <div id="header">
127 |   <div id="header-container" class="container">
128 |   <div id="ipython_notebook" class="nav navbar-brand"><a href="/lab" title='dashboard'>
129 |       <img src='/static/base/images/logo.png?v=641991992878ee24c6f3826e81054a0f' alt='Jupyter Notebook'/>
130 |   </a></div>
131 | 
132 |   
133 |   
134 |   
135 |   
136 |   
137 |   
138 | 
139 | 
140 |   
141 |   
142 |   </div>
143 |   <div class="header-bar"></div>
144 | 
145 |   
146 |   
147 | </div>
148 | 
149 | <div id="site">
150 | 
151 | 
152 | <div id="ipython-main-app" class="container">
153 | 
154 |     
155 |     
156 |     <div class="row">
157 |     <div class="navbar col-sm-8">
158 |       <div class="navbar-inner">
159 |         <div class="container">
160 |           <div class="center-nav">
161 |             <form action="/login?next=%2Flab%2Ftree%2Fcugraph%2F02-Louvain.ipynb" method="post" class="navbar-form pull-left">
162 |               <input type="hidden" name="_xsrf" value="2|656fa83e|f11d75c62306e2b8adccf9f2928fcedc|1562680580"/>
163 |               
164 |                 <label for="password_input"><strong>Password:</strong></label>
165 |               
166 |               <input type="password" name="password" id="password_input" class="form-control">
167 |               <button type="submit" class="btn btn-default" id="login_submit">Log in</button>
168 |             </form>
169 |           </div>
170 |         </div>
171 |       </div>
172 |     </div>
173 |     </div>
174 |     
175 |     
176 |     
177 | </div>
178 | 
179 | 
180 | </div>
181 | 
182 | 
183 | 
184 | 
185 | 
186 | 
187 | 
188 | 
189 | <script type="text/javascript">
190 |   require(["auth/js/main"], function (auth) {
191 |     auth.login_main();
192 |   });
193 | </script>
194 | 
195 | 
196 | 
197 | <script type='text/javascript'>
198 |   function _remove_token_from_url() {
199 |     if (window.location.search.length <= 1) {
200 |       return;
201 |     }
202 |     var search_parameters = window.location.search.slice(1).split('&');
203 |     for (var i = 0; i < search_parameters.length; i++) {
204 |       if (search_parameters[i].split('=')[0] === 'token') {
205 |         // remote token from search parameters
206 |         search_parameters.splice(i, 1);
207 |         var new_search = '';
208 |         if (search_parameters.length) {
209 |           new_search = '?' + search_parameters.join('&');
210 |         }
211 |         var new_url = window.location.origin + 
212 |                       window.location.pathname + 
213 |                       new_search + 
214 |                       window.location.hash;
215 |         window.history.replaceState({}, "", new_url);
216 |         return;
217 |       }
218 |     }
219 |   }
220 |   _remove_token_from_url();
221 | </script>
222 | </body>
223 | 
224 | </html>


--------------------------------------------------------------------------------
/gpu/03-Pagerank.ipynb:
--------------------------------------------------------------------------------
  1 | <!DOCTYPE HTML>
  2 | <html>
  3 | 
  4 | <head>
  5 |     <meta charset="utf-8">
  6 | 
  7 |     <title>Jupyter Notebook</title>
  8 |     <link id="favicon" rel="shortcut icon" type="image/x-icon" href="/static/base/images/favicon.ico?v=97c6417ed01bdc0ae3ef32ae4894fd03">
  9 |     <meta http-equiv="X-UA-Compatible" content="IE=edge" />
 10 |     <link rel="stylesheet" href="/static/components/jquery-ui/themes/smoothness/jquery-ui.min.css?v=3c2a865c832a1322285c55c6ed99abb2" type="text/css" />
 11 |     <link rel="stylesheet" href="/static/components/jquery-typeahead/dist/jquery.typeahead.min.css?v=7afb461de36accb1aa133a1710f5bc56" type="text/css" />
 12 |     <meta name="viewport" content="width=device-width, initial-scale=1.0">
 13 |     
 14 |     
 15 | 
 16 |     <link rel="stylesheet" href="/static/style/style.min.css?v=56cad67b8b1ff3ed9b48850e9efcffb3" type="text/css"/>
 17 |     
 18 | <link rel="stylesheet" href="/static/auth/css/override.css?v=19ec59d2c4f1203c49fe47462028cd9a" type="text/css" />
 19 | 
 20 |     <link rel="stylesheet" href="/custom/custom.css" type="text/css" />
 21 |     <script src="/static/components/es6-promise/promise.min.js?v=f004a16cb856e0ff11781d01ec5ca8fe" type="text/javascript" charset="utf-8"></script>
 22 |     <script src="/static/components/preact/index.js?v=00a2fac73c670ce39ac53d26640eb542" type="text/javascript"></script>
 23 |     <script src="/static/components/proptypes/index.js?v=c40890eb04df9811fcc4d47e53a29604" type="text/javascript"></script>
 24 |     <script src="/static/components/preact-compat/index.js?v=aea8f6660e54b18ace8d84a9b9654c1c" type="text/javascript"></script>
 25 |     <script src="/static/components/requirejs/require.js?v=951f856e81496aaeec2e71a1c2c0d51f" type="text/javascript" charset="utf-8"></script>
 26 |     <script>
 27 |       require.config({
 28 |           
 29 |           urlArgs: "v=20190709124529",
 30 |           
 31 |           baseUrl: '/static/',
 32 |           paths: {
 33 |             'auth/js/main': 'auth/js/main.min',
 34 |             custom : '/custom',
 35 |             nbextensions : '/nbextensions',
 36 |             kernelspecs : '/kernelspecs',
 37 |             underscore : 'components/underscore/underscore-min',
 38 |             backbone : 'components/backbone/backbone-min',
 39 |             jed: 'components/jed/jed',
 40 |             jquery: 'components/jquery/jquery.min',
 41 |             json: 'components/requirejs-plugins/src/json',
 42 |             text: 'components/requirejs-text/text',
 43 |             bootstrap: 'components/bootstrap/dist/js/bootstrap.min',
 44 |             bootstraptour: 'components/bootstrap-tour/build/js/bootstrap-tour.min',
 45 |             'jquery-ui': 'components/jquery-ui/jquery-ui.min',
 46 |             moment: 'components/moment/min/moment-with-locales',
 47 |             codemirror: 'components/codemirror',
 48 |             termjs: 'components/xterm.js/xterm',
 49 |             typeahead: 'components/jquery-typeahead/dist/jquery.typeahead.min',
 50 |           },
 51 |           map: { // for backward compatibility
 52 |               "*": {
 53 |                   "jqueryui": "jquery-ui",
 54 |               }
 55 |           },
 56 |           shim: {
 57 |             typeahead: {
 58 |               deps: ["jquery"],
 59 |               exports: "typeahead"
 60 |             },
 61 |             underscore: {
 62 |               exports: '_'
 63 |             },
 64 |             backbone: {
 65 |               deps: ["underscore", "jquery"],
 66 |               exports: "Backbone"
 67 |             },
 68 |             bootstrap: {
 69 |               deps: ["jquery"],
 70 |               exports: "bootstrap"
 71 |             },
 72 |             bootstraptour: {
 73 |               deps: ["bootstrap"],
 74 |               exports: "Tour"
 75 |             },
 76 |             "jquery-ui": {
 77 |               deps: ["jquery"],
 78 |               exports: "$"
 79 |             }
 80 |           },
 81 |           waitSeconds: 30,
 82 |       });
 83 | 
 84 |       require.config({
 85 |           map: {
 86 |               '*':{
 87 |                 'contents': 'services/contents',
 88 |               }
 89 |           }
 90 |       });
 91 | 
 92 |       // error-catching custom.js shim.
 93 |       define("custom", function (require, exports, module) {
 94 |           try {
 95 |               var custom = require('custom/custom');
 96 |               console.debug('loaded custom.js');
 97 |               return custom;
 98 |           } catch (e) {
 99 |               console.error("error loading custom.js", e);
100 |               return {};
101 |           }
102 |       })
103 | 
104 |     document.nbjs_translations = {"domain": "nbjs", "locale_data": {"nbjs": {"": {"domain": "nbjs"}}}};
105 |     document.documentElement.lang = navigator.language.toLowerCase();
106 |     </script>
107 | 
108 |     
109 |     
110 | 
111 | </head>
112 | 
113 | <body class=""
114 |  
115 |   
116 |  
117 | dir="ltr">
118 | 
119 | <noscript>
120 |     <div id='noscript'>
121 |       Jupyter Notebook requires JavaScript.<br>
122 |       Please enable it to proceed. 
123 |   </div>
124 | </noscript>
125 | 
126 | <div id="header">
127 |   <div id="header-container" class="container">
128 |   <div id="ipython_notebook" class="nav navbar-brand"><a href="/lab" title='dashboard'>
129 |       <img src='/static/base/images/logo.png?v=641991992878ee24c6f3826e81054a0f' alt='Jupyter Notebook'/>
130 |   </a></div>
131 | 
132 |   
133 |   
134 |   
135 |   
136 |   
137 |   
138 | 
139 | 
140 |   
141 |   
142 |   </div>
143 |   <div class="header-bar"></div>
144 | 
145 |   
146 |   
147 | </div>
148 | 
149 | <div id="site">
150 | 
151 | 
152 | <div id="ipython-main-app" class="container">
153 | 
154 |     
155 |     
156 |     <div class="row">
157 |     <div class="navbar col-sm-8">
158 |       <div class="navbar-inner">
159 |         <div class="container">
160 |           <div class="center-nav">
161 |             <form action="/login?next=%2Flab%2Ftree%2Fcugraph%2F03-Pagerank.ipynb" method="post" class="navbar-form pull-left">
162 |               <input type="hidden" name="_xsrf" value="2|0372cd46|96dcd1ef5ee49cda92e0a6e6710ac192|1562680595"/>
163 |               
164 |                 <label for="password_input"><strong>Password:</strong></label>
165 |               
166 |               <input type="password" name="password" id="password_input" class="form-control">
167 |               <button type="submit" class="btn btn-default" id="login_submit">Log in</button>
168 |             </form>
169 |           </div>
170 |         </div>
171 |       </div>
172 |     </div>
173 |     </div>
174 |     
175 |     
176 |     
177 | </div>
178 | 
179 | 
180 | </div>
181 | 
182 | 
183 | 
184 | 
185 | 
186 | 
187 | 
188 | 
189 | <script type="text/javascript">
190 |   require(["auth/js/main"], function (auth) {
191 |     auth.login_main();
192 |   });
193 | </script>
194 | 
195 | 
196 | 
197 | <script type='text/javascript'>
198 |   function _remove_token_from_url() {
199 |     if (window.location.search.length <= 1) {
200 |       return;
201 |     }
202 |     var search_parameters = window.location.search.slice(1).split('&');
203 |     for (var i = 0; i < search_parameters.length; i++) {
204 |       if (search_parameters[i].split('=')[0] === 'token') {
205 |         // remote token from search parameters
206 |         search_parameters.splice(i, 1);
207 |         var new_search = '';
208 |         if (search_parameters.length) {
209 |           new_search = '?' + search_parameters.join('&');
210 |         }
211 |         var new_url = window.location.origin + 
212 |                       window.location.pathname + 
213 |                       new_search + 
214 |                       window.location.hash;
215 |         window.history.replaceState({}, "", new_url);
216 |         return;
217 |       }
218 |     }
219 |   }
220 |   _remove_token_from_url();
221 | </script>
222 | </body>
223 | 
224 | </html>


--------------------------------------------------------------------------------
/gpu/03-UMAP.ipynb:
--------------------------------------------------------------------------------
  1 | <!DOCTYPE HTML>
  2 | <html>
  3 | 
  4 | <head>
  5 |     <meta charset="utf-8">
  6 | 
  7 |     <title>Jupyter Notebook</title>
  8 |     <link id="favicon" rel="shortcut icon" type="image/x-icon" href="/static/base/images/favicon.ico?v=97c6417ed01bdc0ae3ef32ae4894fd03">
  9 |     <meta http-equiv="X-UA-Compatible" content="IE=edge" />
 10 |     <link rel="stylesheet" href="/static/components/jquery-ui/themes/smoothness/jquery-ui.min.css?v=3c2a865c832a1322285c55c6ed99abb2" type="text/css" />
 11 |     <link rel="stylesheet" href="/static/components/jquery-typeahead/dist/jquery.typeahead.min.css?v=7afb461de36accb1aa133a1710f5bc56" type="text/css" />
 12 |     <meta name="viewport" content="width=device-width, initial-scale=1.0">
 13 |     
 14 |     
 15 | 
 16 |     <link rel="stylesheet" href="/static/style/style.min.css?v=56cad67b8b1ff3ed9b48850e9efcffb3" type="text/css"/>
 17 |     
 18 | <link rel="stylesheet" href="/static/auth/css/override.css?v=19ec59d2c4f1203c49fe47462028cd9a" type="text/css" />
 19 | 
 20 |     <link rel="stylesheet" href="/custom/custom.css" type="text/css" />
 21 |     <script src="/static/components/es6-promise/promise.min.js?v=f004a16cb856e0ff11781d01ec5ca8fe" type="text/javascript" charset="utf-8"></script>
 22 |     <script src="/static/components/preact/index.js?v=00a2fac73c670ce39ac53d26640eb542" type="text/javascript"></script>
 23 |     <script src="/static/components/proptypes/index.js?v=c40890eb04df9811fcc4d47e53a29604" type="text/javascript"></script>
 24 |     <script src="/static/components/preact-compat/index.js?v=aea8f6660e54b18ace8d84a9b9654c1c" type="text/javascript"></script>
 25 |     <script src="/static/components/requirejs/require.js?v=951f856e81496aaeec2e71a1c2c0d51f" type="text/javascript" charset="utf-8"></script>
 26 |     <script>
 27 |       require.config({
 28 |           
 29 |           urlArgs: "v=20190709124529",
 30 |           
 31 |           baseUrl: '/static/',
 32 |           paths: {
 33 |             'auth/js/main': 'auth/js/main.min',
 34 |             custom : '/custom',
 35 |             nbextensions : '/nbextensions',
 36 |             kernelspecs : '/kernelspecs',
 37 |             underscore : 'components/underscore/underscore-min',
 38 |             backbone : 'components/backbone/backbone-min',
 39 |             jed: 'components/jed/jed',
 40 |             jquery: 'components/jquery/jquery.min',
 41 |             json: 'components/requirejs-plugins/src/json',
 42 |             text: 'components/requirejs-text/text',
 43 |             bootstrap: 'components/bootstrap/dist/js/bootstrap.min',
 44 |             bootstraptour: 'components/bootstrap-tour/build/js/bootstrap-tour.min',
 45 |             'jquery-ui': 'components/jquery-ui/jquery-ui.min',
 46 |             moment: 'components/moment/min/moment-with-locales',
 47 |             codemirror: 'components/codemirror',
 48 |             termjs: 'components/xterm.js/xterm',
 49 |             typeahead: 'components/jquery-typeahead/dist/jquery.typeahead.min',
 50 |           },
 51 |           map: { // for backward compatibility
 52 |               "*": {
 53 |                   "jqueryui": "jquery-ui",
 54 |               }
 55 |           },
 56 |           shim: {
 57 |             typeahead: {
 58 |               deps: ["jquery"],
 59 |               exports: "typeahead"
 60 |             },
 61 |             underscore: {
 62 |               exports: '_'
 63 |             },
 64 |             backbone: {
 65 |               deps: ["underscore", "jquery"],
 66 |               exports: "Backbone"
 67 |             },
 68 |             bootstrap: {
 69 |               deps: ["jquery"],
 70 |               exports: "bootstrap"
 71 |             },
 72 |             bootstraptour: {
 73 |               deps: ["bootstrap"],
 74 |               exports: "Tour"
 75 |             },
 76 |             "jquery-ui": {
 77 |               deps: ["jquery"],
 78 |               exports: "$"
 79 |             }
 80 |           },
 81 |           waitSeconds: 30,
 82 |       });
 83 | 
 84 |       require.config({
 85 |           map: {
 86 |               '*':{
 87 |                 'contents': 'services/contents',
 88 |               }
 89 |           }
 90 |       });
 91 | 
 92 |       // error-catching custom.js shim.
 93 |       define("custom", function (require, exports, module) {
 94 |           try {
 95 |               var custom = require('custom/custom');
 96 |               console.debug('loaded custom.js');
 97 |               return custom;
 98 |           } catch (e) {
 99 |               console.error("error loading custom.js", e);
100 |               return {};
101 |           }
102 |       })
103 | 
104 |     document.nbjs_translations = {"domain": "nbjs", "locale_data": {"nbjs": {"": {"domain": "nbjs"}}}};
105 |     document.documentElement.lang = navigator.language.toLowerCase();
106 |     </script>
107 | 
108 |     
109 |     
110 | 
111 | </head>
112 | 
113 | <body class=""
114 |  
115 |   
116 |  
117 | dir="ltr">
118 | 
119 | <noscript>
120 |     <div id='noscript'>
121 |       Jupyter Notebook requires JavaScript.<br>
122 |       Please enable it to proceed. 
123 |   </div>
124 | </noscript>
125 | 
126 | <div id="header">
127 |   <div id="header-container" class="container">
128 |   <div id="ipython_notebook" class="nav navbar-brand"><a href="/lab" title='dashboard'>
129 |       <img src='/static/base/images/logo.png?v=641991992878ee24c6f3826e81054a0f' alt='Jupyter Notebook'/>
130 |   </a></div>
131 | 
132 |   
133 |   
134 |   
135 |   
136 |   
137 |   
138 | 
139 | 
140 |   
141 |   
142 |   </div>
143 |   <div class="header-bar"></div>
144 | 
145 |   
146 |   
147 | </div>
148 | 
149 | <div id="site">
150 | 
151 | 
152 | <div id="ipython-main-app" class="container">
153 | 
154 |     
155 |     
156 |     <div class="row">
157 |     <div class="navbar col-sm-8">
158 |       <div class="navbar-inner">
159 |         <div class="container">
160 |           <div class="center-nav">
161 |             <form action="/login?next=%2Flab%2Ftree%2Fcuml%2F03-UMAP.ipynb" method="post" class="navbar-form pull-left">
162 |               <input type="hidden" name="_xsrf" value="2|6362e97e|255984b632f8bb93c451c76780b8987a|1562680677"/>
163 |               
164 |                 <label for="password_input"><strong>Password:</strong></label>
165 |               
166 |               <input type="password" name="password" id="password_input" class="form-control">
167 |               <button type="submit" class="btn btn-default" id="login_submit">Log in</button>
168 |             </form>
169 |           </div>
170 |         </div>
171 |       </div>
172 |     </div>
173 |     </div>
174 |     
175 |     
176 |     
177 | </div>
178 | 
179 | 
180 | </div>
181 | 
182 | 
183 | 
184 | 
185 | 
186 | 
187 | 
188 | 
189 | <script type="text/javascript">
190 |   require(["auth/js/main"], function (auth) {
191 |     auth.login_main();
192 |   });
193 | </script>
194 | 
195 | 
196 | 
197 | <script type='text/javascript'>
198 |   function _remove_token_from_url() {
199 |     if (window.location.search.length <= 1) {
200 |       return;
201 |     }
202 |     var search_parameters = window.location.search.slice(1).split('&');
203 |     for (var i = 0; i < search_parameters.length; i++) {
204 |       if (search_parameters[i].split('=')[0] === 'token') {
205 |         // remote token from search parameters
206 |         search_parameters.splice(i, 1);
207 |         var new_search = '';
208 |         if (search_parameters.length) {
209 |           new_search = '?' + search_parameters.join('&');
210 |         }
211 |         var new_url = window.location.origin + 
212 |                       window.location.pathname + 
213 |                       new_search + 
214 |                       window.location.hash;
215 |         window.history.replaceState({}, "", new_url);
216 |         return;
217 |       }
218 |     }
219 |   }
220 |   _remove_token_from_url();
221 | </script>
222 | </body>
223 | 
224 | </html>


--------------------------------------------------------------------------------
/gpu/Introduction.ipynb:
--------------------------------------------------------------------------------
  1 | <!DOCTYPE HTML>
  2 | <html>
  3 | 
  4 | <head>
  5 |     <meta charset="utf-8">
  6 | 
  7 |     <title>Jupyter Notebook</title>
  8 |     <link id="favicon" rel="shortcut icon" type="image/x-icon" href="/static/base/images/favicon.ico?v=97c6417ed01bdc0ae3ef32ae4894fd03">
  9 |     <meta http-equiv="X-UA-Compatible" content="IE=edge" />
 10 |     <link rel="stylesheet" href="/static/components/jquery-ui/themes/smoothness/jquery-ui.min.css?v=3c2a865c832a1322285c55c6ed99abb2" type="text/css" />
 11 |     <link rel="stylesheet" href="/static/components/jquery-typeahead/dist/jquery.typeahead.min.css?v=7afb461de36accb1aa133a1710f5bc56" type="text/css" />
 12 |     <meta name="viewport" content="width=device-width, initial-scale=1.0">
 13 |     
 14 |     
 15 | 
 16 |     <link rel="stylesheet" href="/static/style/style.min.css?v=56cad67b8b1ff3ed9b48850e9efcffb3" type="text/css"/>
 17 |     
 18 | <link rel="stylesheet" href="/static/auth/css/override.css?v=19ec59d2c4f1203c49fe47462028cd9a" type="text/css" />
 19 | 
 20 |     <link rel="stylesheet" href="/custom/custom.css" type="text/css" />
 21 |     <script src="/static/components/es6-promise/promise.min.js?v=f004a16cb856e0ff11781d01ec5ca8fe" type="text/javascript" charset="utf-8"></script>
 22 |     <script src="/static/components/preact/index.js?v=00a2fac73c670ce39ac53d26640eb542" type="text/javascript"></script>
 23 |     <script src="/static/components/proptypes/index.js?v=c40890eb04df9811fcc4d47e53a29604" type="text/javascript"></script>
 24 |     <script src="/static/components/preact-compat/index.js?v=aea8f6660e54b18ace8d84a9b9654c1c" type="text/javascript"></script>
 25 |     <script src="/static/components/requirejs/require.js?v=951f856e81496aaeec2e71a1c2c0d51f" type="text/javascript" charset="utf-8"></script>
 26 |     <script>
 27 |       require.config({
 28 |           
 29 |           urlArgs: "v=20190709124529",
 30 |           
 31 |           baseUrl: '/static/',
 32 |           paths: {
 33 |             'auth/js/main': 'auth/js/main.min',
 34 |             custom : '/custom',
 35 |             nbextensions : '/nbextensions',
 36 |             kernelspecs : '/kernelspecs',
 37 |             underscore : 'components/underscore/underscore-min',
 38 |             backbone : 'components/backbone/backbone-min',
 39 |             jed: 'components/jed/jed',
 40 |             jquery: 'components/jquery/jquery.min',
 41 |             json: 'components/requirejs-plugins/src/json',
 42 |             text: 'components/requirejs-text/text',
 43 |             bootstrap: 'components/bootstrap/dist/js/bootstrap.min',
 44 |             bootstraptour: 'components/bootstrap-tour/build/js/bootstrap-tour.min',
 45 |             'jquery-ui': 'components/jquery-ui/jquery-ui.min',
 46 |             moment: 'components/moment/min/moment-with-locales',
 47 |             codemirror: 'components/codemirror',
 48 |             termjs: 'components/xterm.js/xterm',
 49 |             typeahead: 'components/jquery-typeahead/dist/jquery.typeahead.min',
 50 |           },
 51 |           map: { // for backward compatibility
 52 |               "*": {
 53 |                   "jqueryui": "jquery-ui",
 54 |               }
 55 |           },
 56 |           shim: {
 57 |             typeahead: {
 58 |               deps: ["jquery"],
 59 |               exports: "typeahead"
 60 |             },
 61 |             underscore: {
 62 |               exports: '_'
 63 |             },
 64 |             backbone: {
 65 |               deps: ["underscore", "jquery"],
 66 |               exports: "Backbone"
 67 |             },
 68 |             bootstrap: {
 69 |               deps: ["jquery"],
 70 |               exports: "bootstrap"
 71 |             },
 72 |             bootstraptour: {
 73 |               deps: ["bootstrap"],
 74 |               exports: "Tour"
 75 |             },
 76 |             "jquery-ui": {
 77 |               deps: ["jquery"],
 78 |               exports: "$"
 79 |             }
 80 |           },
 81 |           waitSeconds: 30,
 82 |       });
 83 | 
 84 |       require.config({
 85 |           map: {
 86 |               '*':{
 87 |                 'contents': 'services/contents',
 88 |               }
 89 |           }
 90 |       });
 91 | 
 92 |       // error-catching custom.js shim.
 93 |       define("custom", function (require, exports, module) {
 94 |           try {
 95 |               var custom = require('custom/custom');
 96 |               console.debug('loaded custom.js');
 97 |               return custom;
 98 |           } catch (e) {
 99 |               console.error("error loading custom.js", e);
100 |               return {};
101 |           }
102 |       })
103 | 
104 |     document.nbjs_translations = {"domain": "nbjs", "locale_data": {"nbjs": {"": {"domain": "nbjs"}}}};
105 |     document.documentElement.lang = navigator.language.toLowerCase();
106 |     </script>
107 | 
108 |     
109 |     
110 | 
111 | </head>
112 | 
113 | <body class=""
114 |  
115 |   
116 |  
117 | dir="ltr">
118 | 
119 | <noscript>
120 |     <div id='noscript'>
121 |       Jupyter Notebook requires JavaScript.<br>
122 |       Please enable it to proceed. 
123 |   </div>
124 | </noscript>
125 | 
126 | <div id="header">
127 |   <div id="header-container" class="container">
128 |   <div id="ipython_notebook" class="nav navbar-brand"><a href="/lab" title='dashboard'>
129 |       <img src='/static/base/images/logo.png?v=641991992878ee24c6f3826e81054a0f' alt='Jupyter Notebook'/>
130 |   </a></div>
131 | 
132 |   
133 |   
134 |   
135 |   
136 |   
137 |   
138 | 
139 | 
140 |   
141 |   
142 |   </div>
143 |   <div class="header-bar"></div>
144 | 
145 |   
146 |   
147 | </div>
148 | 
149 | <div id="site">
150 | 
151 | 
152 | <div id="ipython-main-app" class="container">
153 | 
154 |     
155 |     
156 |     <div class="row">
157 |     <div class="navbar col-sm-8">
158 |       <div class="navbar-inner">
159 |         <div class="container">
160 |           <div class="center-nav">
161 |             <form action="/login?next=%2Flab%2Ftree%2FIntroduction.ipynb" method="post" class="navbar-form pull-left">
162 |               <input type="hidden" name="_xsrf" value="2|3640f761|eaa6d847c229c19a83a96034e99cc820|1562679159"/>
163 |               
164 |                 <label for="password_input"><strong>Password:</strong></label>
165 |               
166 |               <input type="password" name="password" id="password_input" class="form-control">
167 |               <button type="submit" class="btn btn-default" id="login_submit">Log in</button>
168 |             </form>
169 |           </div>
170 |         </div>
171 |       </div>
172 |     </div>
173 |     </div>
174 |     
175 |     
176 |     
177 | </div>
178 | 
179 | 
180 | </div>
181 | 
182 | 
183 | 
184 | 
185 | 
186 | 
187 | 
188 | 
189 | <script type="text/javascript">
190 |   require(["auth/js/main"], function (auth) {
191 |     auth.login_main();
192 |   });
193 | </script>
194 | 
195 | 
196 | 
197 | <script type='text/javascript'>
198 |   function _remove_token_from_url() {
199 |     if (window.location.search.length <= 1) {
200 |       return;
201 |     }
202 |     var search_parameters = window.location.search.slice(1).split('&');
203 |     for (var i = 0; i < search_parameters.length; i++) {
204 |       if (search_parameters[i].split('=')[0] === 'token') {
205 |         // remote token from search parameters
206 |         search_parameters.splice(i, 1);
207 |         var new_search = '';
208 |         if (search_parameters.length) {
209 |           new_search = '?' + search_parameters.join('&');
210 |         }
211 |         var new_url = window.location.origin + 
212 |                       window.location.pathname + 
213 |                       new_search + 
214 |                       window.location.hash;
215 |         window.history.replaceState({}, "", new_url);
216 |         return;
217 |       }
218 |     }
219 |   }
220 |   _remove_token_from_url();
221 | </script>
222 | </body>
223 | 
224 | </html>


--------------------------------------------------------------------------------
/gpu/README.md:
--------------------------------------------------------------------------------
 1 | # GPU Data Science
 2 | 
 3 | > Okay, so they are giving us cloud IP addresses to log into, so I may not have much to share here, except my notes.
 4 | 
 5 | ## Notes
 6 | 
 7 | * Fun! They provide a cloud instance for everyone in the room, so we can play with a machine with sufficient GPU power.
 8 | * The old solution: Hadoop, but that had too much IO. So people moved to Spark.
 9 | * RAPIDS is Nvidia's GPU architecture
10 | * RAPIDS uses Apache Arrow (which tries to standardize the storage of columnar data in memory)
11 | * cuDF has nearly identical syntax to Pandas DataFrames
12 |    * The API for cuDF is super handy, I'll give them that.
13 |    * I guess that means you can also use cuDF as a drop-in replacement in your code base... I see.
14 | * String Example: big string replace operations on pd.DataFrame
15 |    * 50x speed-up using cuDF and fixed-length NumPy strings
16 |    * TODO: Whoops. At work, are we not converting our strings in pd.DataFrames to NumPy string arrays?
17 | * [RAPIDS docs](https://docs.rapids.ai/)
18 | * [Play around with RAPIDS](https://rapids.ai/start.html)
19 | * [cuDF GitHub](https://github.com/rapidsai/cudf)
20 | * [cupy](https://cupy.chainer.org/) is a replacement for NumPy, but in the GPU
21 | 
22 | 


--------------------------------------------------------------------------------
/gpu/data/karate-data.csv:
--------------------------------------------------------------------------------
  1 | <!DOCTYPE HTML>
  2 | <html>
  3 | 
  4 | <head>
  5 |     <meta charset="utf-8">
  6 | 
  7 |     <title>Jupyter Notebook</title>
  8 |     <link id="favicon" rel="shortcut icon" type="image/x-icon" href="/static/base/images/favicon.ico?v=97c6417ed01bdc0ae3ef32ae4894fd03">
  9 |     <meta http-equiv="X-UA-Compatible" content="IE=edge" />
 10 |     <link rel="stylesheet" href="/static/components/jquery-ui/themes/smoothness/jquery-ui.min.css?v=3c2a865c832a1322285c55c6ed99abb2" type="text/css" />
 11 |     <link rel="stylesheet" href="/static/components/jquery-typeahead/dist/jquery.typeahead.min.css?v=7afb461de36accb1aa133a1710f5bc56" type="text/css" />
 12 |     <meta name="viewport" content="width=device-width, initial-scale=1.0">
 13 |     
 14 |     
 15 | 
 16 |     <link rel="stylesheet" href="/static/style/style.min.css?v=56cad67b8b1ff3ed9b48850e9efcffb3" type="text/css"/>
 17 |     
 18 | <link rel="stylesheet" href="/static/auth/css/override.css?v=19ec59d2c4f1203c49fe47462028cd9a" type="text/css" />
 19 | 
 20 |     <link rel="stylesheet" href="/custom/custom.css" type="text/css" />
 21 |     <script src="/static/components/es6-promise/promise.min.js?v=f004a16cb856e0ff11781d01ec5ca8fe" type="text/javascript" charset="utf-8"></script>
 22 |     <script src="/static/components/preact/index.js?v=00a2fac73c670ce39ac53d26640eb542" type="text/javascript"></script>
 23 |     <script src="/static/components/proptypes/index.js?v=c40890eb04df9811fcc4d47e53a29604" type="text/javascript"></script>
 24 |     <script src="/static/components/preact-compat/index.js?v=aea8f6660e54b18ace8d84a9b9654c1c" type="text/javascript"></script>
 25 |     <script src="/static/components/requirejs/require.js?v=951f856e81496aaeec2e71a1c2c0d51f" type="text/javascript" charset="utf-8"></script>
 26 |     <script>
 27 |       require.config({
 28 |           
 29 |           urlArgs: "v=20190709124529",
 30 |           
 31 |           baseUrl: '/static/',
 32 |           paths: {
 33 |             'auth/js/main': 'auth/js/main.min',
 34 |             custom : '/custom',
 35 |             nbextensions : '/nbextensions',
 36 |             kernelspecs : '/kernelspecs',
 37 |             underscore : 'components/underscore/underscore-min',
 38 |             backbone : 'components/backbone/backbone-min',
 39 |             jed: 'components/jed/jed',
 40 |             jquery: 'components/jquery/jquery.min',
 41 |             json: 'components/requirejs-plugins/src/json',
 42 |             text: 'components/requirejs-text/text',
 43 |             bootstrap: 'components/bootstrap/dist/js/bootstrap.min',
 44 |             bootstraptour: 'components/bootstrap-tour/build/js/bootstrap-tour.min',
 45 |             'jquery-ui': 'components/jquery-ui/jquery-ui.min',
 46 |             moment: 'components/moment/min/moment-with-locales',
 47 |             codemirror: 'components/codemirror',
 48 |             termjs: 'components/xterm.js/xterm',
 49 |             typeahead: 'components/jquery-typeahead/dist/jquery.typeahead.min',
 50 |           },
 51 |           map: { // for backward compatibility
 52 |               "*": {
 53 |                   "jqueryui": "jquery-ui",
 54 |               }
 55 |           },
 56 |           shim: {
 57 |             typeahead: {
 58 |               deps: ["jquery"],
 59 |               exports: "typeahead"
 60 |             },
 61 |             underscore: {
 62 |               exports: '_'
 63 |             },
 64 |             backbone: {
 65 |               deps: ["underscore", "jquery"],
 66 |               exports: "Backbone"
 67 |             },
 68 |             bootstrap: {
 69 |               deps: ["jquery"],
 70 |               exports: "bootstrap"
 71 |             },
 72 |             bootstraptour: {
 73 |               deps: ["bootstrap"],
 74 |               exports: "Tour"
 75 |             },
 76 |             "jquery-ui": {
 77 |               deps: ["jquery"],
 78 |               exports: "$"
 79 |             }
 80 |           },
 81 |           waitSeconds: 30,
 82 |       });
 83 | 
 84 |       require.config({
 85 |           map: {
 86 |               '*':{
 87 |                 'contents': 'services/contents',
 88 |               }
 89 |           }
 90 |       });
 91 | 
 92 |       // error-catching custom.js shim.
 93 |       define("custom", function (require, exports, module) {
 94 |           try {
 95 |               var custom = require('custom/custom');
 96 |               console.debug('loaded custom.js');
 97 |               return custom;
 98 |           } catch (e) {
 99 |               console.error("error loading custom.js", e);
100 |               return {};
101 |           }
102 |       })
103 | 
104 |     document.nbjs_translations = {"domain": "nbjs", "locale_data": {"nbjs": {"": {"domain": "nbjs"}}}};
105 |     document.documentElement.lang = navigator.language.toLowerCase();
106 |     </script>
107 | 
108 |     
109 |     
110 | 
111 | </head>
112 | 
113 | <body class=""
114 |  
115 |   
116 |  
117 | dir="ltr">
118 | 
119 | <noscript>
120 |     <div id='noscript'>
121 |       Jupyter Notebook requires JavaScript.<br>
122 |       Please enable it to proceed. 
123 |   </div>
124 | </noscript>
125 | 
126 | <div id="header">
127 |   <div id="header-container" class="container">
128 |   <div id="ipython_notebook" class="nav navbar-brand"><a href="/lab" title='dashboard'>
129 |       <img src='/static/base/images/logo.png?v=641991992878ee24c6f3826e81054a0f' alt='Jupyter Notebook'/>
130 |   </a></div>
131 | 
132 |   
133 |   
134 |   
135 |   
136 |   
137 |   
138 | 
139 | 
140 |   
141 |   
142 |   </div>
143 |   <div class="header-bar"></div>
144 | 
145 |   
146 |   
147 | </div>
148 | 
149 | <div id="site">
150 | 
151 | 
152 | <div id="ipython-main-app" class="container">
153 | 
154 |     
155 |     
156 |     <div class="row">
157 |     <div class="navbar col-sm-8">
158 |       <div class="navbar-inner">
159 |         <div class="container">
160 |           <div class="center-nav">
161 |             <form action="/login?next=%2Flab%2Ftree%2Fcugraph%2Fdata%2Fkarate-data.csv" method="post" class="navbar-form pull-left">
162 |               <input type="hidden" name="_xsrf" value="2|bf3b752d|ec20157b555c28374d637870018d05f4|1562680614"/>
163 |               
164 |                 <label for="password_input"><strong>Password:</strong></label>
165 |               
166 |               <input type="password" name="password" id="password_input" class="form-control">
167 |               <button type="submit" class="btn btn-default" id="login_submit">Log in</button>
168 |             </form>
169 |           </div>
170 |         </div>
171 |       </div>
172 |     </div>
173 |     </div>
174 |     
175 |     
176 |     
177 | </div>
178 | 
179 | 
180 | </div>
181 | 
182 | 
183 | 
184 | 
185 | 
186 | 
187 | 
188 | 
189 | <script type="text/javascript">
190 |   require(["auth/js/main"], function (auth) {
191 |     auth.login_main();
192 |   });
193 | </script>
194 | 
195 | 
196 | 
197 | <script type='text/javascript'>
198 |   function _remove_token_from_url() {
199 |     if (window.location.search.length <= 1) {
200 |       return;
201 |     }
202 |     var search_parameters = window.location.search.slice(1).split('&');
203 |     for (var i = 0; i < search_parameters.length; i++) {
204 |       if (search_parameters[i].split('=')[0] === 'token') {
205 |         // remote token from search parameters
206 |         search_parameters.splice(i, 1);
207 |         var new_search = '';
208 |         if (search_parameters.length) {
209 |           new_search = '?' + search_parameters.join('&');
210 |         }
211 |         var new_url = window.location.origin + 
212 |                       window.location.pathname + 
213 |                       new_search + 
214 |                       window.location.hash;
215 |         window.history.replaceState({}, "", new_url);
216 |         return;
217 |       }
218 |     }
219 |   }
220 |   _remove_token_from_url();
221 | </script>
222 | </body>
223 | 
224 | </html>


--------------------------------------------------------------------------------
/gpu/index.ipynb:
--------------------------------------------------------------------------------
  1 | <!DOCTYPE HTML>
  2 | <html>
  3 | 
  4 | <head>
  5 |     <meta charset="utf-8">
  6 | 
  7 |     <title>Jupyter Notebook</title>
  8 |     <link id="favicon" rel="shortcut icon" type="image/x-icon" href="/static/base/images/favicon.ico?v=97c6417ed01bdc0ae3ef32ae4894fd03">
  9 |     <meta http-equiv="X-UA-Compatible" content="IE=edge" />
 10 |     <link rel="stylesheet" href="/static/components/jquery-ui/themes/smoothness/jquery-ui.min.css?v=3c2a865c832a1322285c55c6ed99abb2" type="text/css" />
 11 |     <link rel="stylesheet" href="/static/components/jquery-typeahead/dist/jquery.typeahead.min.css?v=7afb461de36accb1aa133a1710f5bc56" type="text/css" />
 12 |     <meta name="viewport" content="width=device-width, initial-scale=1.0">
 13 |     
 14 |     
 15 | 
 16 |     <link rel="stylesheet" href="/static/style/style.min.css?v=56cad67b8b1ff3ed9b48850e9efcffb3" type="text/css"/>
 17 |     
 18 | <link rel="stylesheet" href="/static/auth/css/override.css?v=19ec59d2c4f1203c49fe47462028cd9a" type="text/css" />
 19 | 
 20 |     <link rel="stylesheet" href="/custom/custom.css" type="text/css" />
 21 |     <script src="/static/components/es6-promise/promise.min.js?v=f004a16cb856e0ff11781d01ec5ca8fe" type="text/javascript" charset="utf-8"></script>
 22 |     <script src="/static/components/preact/index.js?v=00a2fac73c670ce39ac53d26640eb542" type="text/javascript"></script>
 23 |     <script src="/static/components/proptypes/index.js?v=c40890eb04df9811fcc4d47e53a29604" type="text/javascript"></script>
 24 |     <script src="/static/components/preact-compat/index.js?v=aea8f6660e54b18ace8d84a9b9654c1c" type="text/javascript"></script>
 25 |     <script src="/static/components/requirejs/require.js?v=951f856e81496aaeec2e71a1c2c0d51f" type="text/javascript" charset="utf-8"></script>
 26 |     <script>
 27 |       require.config({
 28 |           
 29 |           urlArgs: "v=20190709124529",
 30 |           
 31 |           baseUrl: '/static/',
 32 |           paths: {
 33 |             'auth/js/main': 'auth/js/main.min',
 34 |             custom : '/custom',
 35 |             nbextensions : '/nbextensions',
 36 |             kernelspecs : '/kernelspecs',
 37 |             underscore : 'components/underscore/underscore-min',
 38 |             backbone : 'components/backbone/backbone-min',
 39 |             jed: 'components/jed/jed',
 40 |             jquery: 'components/jquery/jquery.min',
 41 |             json: 'components/requirejs-plugins/src/json',
 42 |             text: 'components/requirejs-text/text',
 43 |             bootstrap: 'components/bootstrap/dist/js/bootstrap.min',
 44 |             bootstraptour: 'components/bootstrap-tour/build/js/bootstrap-tour.min',
 45 |             'jquery-ui': 'components/jquery-ui/jquery-ui.min',
 46 |             moment: 'components/moment/min/moment-with-locales',
 47 |             codemirror: 'components/codemirror',
 48 |             termjs: 'components/xterm.js/xterm',
 49 |             typeahead: 'components/jquery-typeahead/dist/jquery.typeahead.min',
 50 |           },
 51 |           map: { // for backward compatibility
 52 |               "*": {
 53 |                   "jqueryui": "jquery-ui",
 54 |               }
 55 |           },
 56 |           shim: {
 57 |             typeahead: {
 58 |               deps: ["jquery"],
 59 |               exports: "typeahead"
 60 |             },
 61 |             underscore: {
 62 |               exports: '_'
 63 |             },
 64 |             backbone: {
 65 |               deps: ["underscore", "jquery"],
 66 |               exports: "Backbone"
 67 |             },
 68 |             bootstrap: {
 69 |               deps: ["jquery"],
 70 |               exports: "bootstrap"
 71 |             },
 72 |             bootstraptour: {
 73 |               deps: ["bootstrap"],
 74 |               exports: "Tour"
 75 |             },
 76 |             "jquery-ui": {
 77 |               deps: ["jquery"],
 78 |               exports: "$"
 79 |             }
 80 |           },
 81 |           waitSeconds: 30,
 82 |       });
 83 | 
 84 |       require.config({
 85 |           map: {
 86 |               '*':{
 87 |                 'contents': 'services/contents',
 88 |               }
 89 |           }
 90 |       });
 91 | 
 92 |       // error-catching custom.js shim.
 93 |       define("custom", function (require, exports, module) {
 94 |           try {
 95 |               var custom = require('custom/custom');
 96 |               console.debug('loaded custom.js');
 97 |               return custom;
 98 |           } catch (e) {
 99 |               console.error("error loading custom.js", e);
100 |               return {};
101 |           }
102 |       })
103 | 
104 |     document.nbjs_translations = {"domain": "nbjs", "locale_data": {"nbjs": {"": {"domain": "nbjs"}}}};
105 |     document.documentElement.lang = navigator.language.toLowerCase();
106 |     </script>
107 | 
108 |     
109 |     
110 | 
111 | </head>
112 | 
113 | <body class=""
114 |  
115 |   
116 |  
117 | dir="ltr">
118 | 
119 | <noscript>
120 |     <div id='noscript'>
121 |       Jupyter Notebook requires JavaScript.<br>
122 |       Please enable it to proceed. 
123 |   </div>
124 | </noscript>
125 | 
126 | <div id="header">
127 |   <div id="header-container" class="container">
128 |   <div id="ipython_notebook" class="nav navbar-brand"><a href="/lab" title='dashboard'>
129 |       <img src='/static/base/images/logo.png?v=641991992878ee24c6f3826e81054a0f' alt='Jupyter Notebook'/>
130 |   </a></div>
131 | 
132 |   
133 |   
134 |   
135 |   
136 |   
137 |   
138 | 
139 | 
140 |   
141 |   
142 |   </div>
143 |   <div class="header-bar"></div>
144 | 
145 |   
146 |   
147 | </div>
148 | 
149 | <div id="site">
150 | 
151 | 
152 | <div id="ipython-main-app" class="container">
153 | 
154 |     
155 |     
156 |     <div class="row">
157 |     <div class="navbar col-sm-8">
158 |       <div class="navbar-inner">
159 |         <div class="container">
160 |           <div class="center-nav">
161 |             <form action="/login?next=%2Flab%2Ftree%2Findex.ipynb" method="post" class="navbar-form pull-left">
162 |               <input type="hidden" name="_xsrf" value="2|b7945cec|1d6b7e75b5c4a689d93b7357f82cc988|1562679134"/>
163 |               
164 |                 <label for="password_input"><strong>Password:</strong></label>
165 |               
166 |               <input type="password" name="password" id="password_input" class="form-control">
167 |               <button type="submit" class="btn btn-default" id="login_submit">Log in</button>
168 |             </form>
169 |           </div>
170 |         </div>
171 |       </div>
172 |     </div>
173 |     </div>
174 |     
175 |     
176 |     
177 | </div>
178 | 
179 | 
180 | </div>
181 | 
182 | 
183 | 
184 | 
185 | 
186 | 
187 | 
188 | 
189 | <script type="text/javascript">
190 |   require(["auth/js/main"], function (auth) {
191 |     auth.login_main();
192 |   });
193 | </script>
194 | 
195 | 
196 | 
197 | <script type='text/javascript'>
198 |   function _remove_token_from_url() {
199 |     if (window.location.search.length <= 1) {
200 |       return;
201 |     }
202 |     var search_parameters = window.location.search.slice(1).split('&');
203 |     for (var i = 0; i < search_parameters.length; i++) {
204 |       if (search_parameters[i].split('=')[0] === 'token') {
205 |         // remote token from search parameters
206 |         search_parameters.splice(i, 1);
207 |         var new_search = '';
208 |         if (search_parameters.length) {
209 |           new_search = '?' + search_parameters.join('&');
210 |         }
211 |         var new_url = window.location.origin + 
212 |                       window.location.pathname + 
213 |                       new_search + 
214 |                       window.location.hash;
215 |         window.history.replaceState({}, "", new_url);
216 |         return;
217 |       }
218 |     }
219 |   }
220 |   _remove_token_from_url();
221 | </script>
222 | </body>
223 | 
224 | </html>


--------------------------------------------------------------------------------
/modern_time_series_analysis/ModernTimeSeriesAnalysis/DeepLearning/Electricity/models.json:
--------------------------------------------------------------------------------
 1 | [
 2 |     {
 3 |         "scenarios": {
 4 |             "drop": [0.2],            
 5 |             "batch-n": [128, 256],
 6 |             "lr": [0.0001],
 7 |             "n-epochs":[10],
 8 |             "model": [ "simple_lstnet_mode",  "lstnet_mode", "rnn_model", "cnn_model", "fc_model"],
 9 |             "data_src": [
10 |                 "electricity.txt", 
11 |                 "electricity.diff.txt", 
12 |             ],
13 |             "win": [128, 64,],
14 |             "seasonal_period": [24]
15 |         },
16 | 
17 |         "drop": "$",
18 |         "win": "$",
19 |         "batch-n": "$",
20 |         "lr": "$",
21 |         "n-epochs": "$",
22 |         "data-file": "/data/${data_src}",
23 |         "save-dir": "${model}_${lr}_${batch-n}_${win}_${drop}/${data_src}"
24 |     },
25 | 
26 | ]
27 | 


--------------------------------------------------------------------------------
/modern_time_series_analysis/ModernTimeSeriesAnalysis/DeepLearning/Electricity/perf.py:
--------------------------------------------------------------------------------
 1 | 
 2 | import numpy as np
 3 | import mxnet as mx
 4 | import pdb
 5 | 
 6 | np.seterr(divide='ignore', invalid='ignore')
 7 | 
 8 | ## for saving
 9 | import pandas as pd
10 | import os
11 | 
12 | def COR(label, pred):
13 |     label_demeaned = label - label.mean(0)
14 |     label_sumsquares = np.sum(np.square(label_demeaned), 0)
15 | 
16 |     pred_demeaned = pred - pred.mean(0)
17 |     pred_sumsquares = np.sum(np.square(pred_demeaned), 0)
18 | 
19 |     cor_coef =  np.diagonal(np.dot(label_demeaned.T, pred_demeaned)) / \
20 |                 np.sqrt(label_sumsquares * pred_sumsquares)                   
21 | 
22 |     return np.nanmean(cor_coef)
23 | 
24 | 
25 | def write_eval(pred, label, save_dir, mode, epoch):
26 |     if not os.path.exists(save_dir):
27 |         os.makedirs(save_dir)
28 |         
29 |     pred_df  = pd.DataFrame(pred)
30 |     label_df = pd.DataFrame(label)
31 |     if epoch < 10:
32 |         pred_df.to_csv( os.path.join(save_dir, '%s_pred_0%d.csv'  % (mode, epoch)))
33 |         label_df.to_csv(os.path.join(save_dir, '%s_label_0%d.csv' % (mode, epoch)))
34 |     else :        
35 |         pred_df.to_csv( os.path.join(save_dir, '%s_pred_%d.csv'  % (mode, epoch)))
36 |         label_df.to_csv(os.path.join(save_dir, '%s_label_%d.csv' % (mode, epoch)))
37 | 
38 |     return { 'COR': COR(label,pred) }
39 |     
40 | 
41 | 
42 | 
43 | 


--------------------------------------------------------------------------------
/modern_time_series_analysis/ModernTimeSeriesAnalysis/DeepLearning/Electricity/run.json:
--------------------------------------------------------------------------------
1 | {
2 |     "script": "/demo/fits.py",
3 |     "models": "/demo/models.json",
4 | }
5 | 


--------------------------------------------------------------------------------
/modern_time_series_analysis/ModernTimeSeriesAnalysis/MachineLearning/3_Trees_for_Classification_and_Prediction_STUDENT.ipynb:
--------------------------------------------------------------------------------
  1 | {
  2 |  "cells": [
  3 |   {
  4 |    "cell_type": "code",
  5 |    "execution_count": null,
  6 |    "metadata": {},
  7 |    "outputs": [],
  8 |    "source": [
  9 |     "import matplotlib.pyplot as plt\n",
 10 |     "plt.rcParams['figure.figsize'] = [10, 10]"
 11 |    ]
 12 |   },
 13 |   {
 14 |    "cell_type": "code",
 15 |    "execution_count": null,
 16 |    "metadata": {},
 17 |    "outputs": [],
 18 |    "source": [
 19 |     "import cesium\n",
 20 |     "import xgboost as xgb\n",
 21 |     "import numpy as np\n",
 22 |     "import matplotlib.pyplot as plt\n",
 23 |     "import pandas as pd\n",
 24 |     "import time\n",
 25 |     "\n",
 26 |     "from cesium import datasets\n",
 27 |     "from cesium import featurize as ft\n",
 28 |     "\n",
 29 |     "import scipy\n",
 30 |     "from scipy.stats import pearsonr, spearmanr\n",
 31 |     "from scipy.stats import skew\n",
 32 |     "\n",
 33 |     "import sklearn\n",
 34 |     "from sklearn.ensemble import RandomForestClassifier\n",
 35 |     "from sklearn.metrics import accuracy_score\n",
 36 |     "from sklearn.model_selection import train_test_split"
 37 |    ]
 38 |   },
 39 |   {
 40 |    "cell_type": "code",
 41 |    "execution_count": null,
 42 |    "metadata": {},
 43 |    "outputs": [],
 44 |    "source": [
 45 |     "print(cesium.__version__)\n",
 46 |     "print(xgb.__version__)\n",
 47 |     "print(scipy.__version__)\n",
 48 |     "print(sklearn.__version__)"
 49 |    ]
 50 |   },
 51 |   {
 52 |    "cell_type": "markdown",
 53 |    "metadata": {},
 54 |    "source": [
 55 |     "## Load data and generate some features of interest"
 56 |    ]
 57 |   },
 58 |   {
 59 |    "cell_type": "code",
 60 |    "execution_count": null,
 61 |    "metadata": {},
 62 |    "outputs": [],
 63 |    "source": [
 64 |     "eeg = datasets.fetch_andrzejak()"
 65 |    ]
 66 |   },
 67 |   {
 68 |    "cell_type": "code",
 69 |    "execution_count": null,
 70 |    "metadata": {},
 71 |    "outputs": [],
 72 |    "source": [
 73 |     "type(eeg)"
 74 |    ]
 75 |   },
 76 |   {
 77 |    "cell_type": "code",
 78 |    "execution_count": null,
 79 |    "metadata": {},
 80 |    "outputs": [],
 81 |    "source": [
 82 |     "eeg.keys()"
 83 |    ]
 84 |   },
 85 |   {
 86 |    "cell_type": "markdown",
 87 |    "metadata": {},
 88 |    "source": [
 89 |     "### Visually inspect"
 90 |    ]
 91 |   },
 92 |   {
 93 |    "cell_type": "code",
 94 |    "execution_count": null,
 95 |    "metadata": {},
 96 |    "outputs": [],
 97 |    "source": [
 98 |     "plt.subplot(3, 1, 1)\n",
 99 |     "plt.plot(eeg[\"measurements\"][0])\n",
100 |     "plt.legend(eeg['classes'][0])\n",
101 |     "plt.subplot(3, 1, 2)\n",
102 |     "plt.plot(eeg[\"measurements\"][300])\n",
103 |     "plt.legend(eeg['classes'][300])\n",
104 |     "plt.subplot(3, 1, 3)\n",
105 |     "plt.plot(eeg[\"measurements\"][450])\n",
106 |     "plt.legend(eeg['classes'][450])"
107 |    ]
108 |   },
109 |   {
110 |    "cell_type": "code",
111 |    "execution_count": null,
112 |    "metadata": {},
113 |    "outputs": [],
114 |    "source": [
115 |     "type(eeg[\"measurements\"][0])"
116 |    ]
117 |   },
118 |   {
119 |    "cell_type": "code",
120 |    "execution_count": null,
121 |    "metadata": {},
122 |    "outputs": [],
123 |    "source": [
124 |     "type(eeg)"
125 |    ]
126 |   },
127 |   {
128 |    "cell_type": "code",
129 |    "execution_count": null,
130 |    "metadata": {},
131 |    "outputs": [],
132 |    "source": [
133 |     "eeg.keys()"
134 |    ]
135 |   },
136 |   {
137 |    "cell_type": "code",
138 |    "execution_count": null,
139 |    "metadata": {},
140 |    "outputs": [],
141 |    "source": [
142 |     "type(eeg['measurements'])"
143 |    ]
144 |   },
145 |   {
146 |    "cell_type": "code",
147 |    "execution_count": null,
148 |    "metadata": {},
149 |    "outputs": [],
150 |    "source": [
151 |     "len(eeg['measurements'])"
152 |    ]
153 |   },
154 |   {
155 |    "cell_type": "code",
156 |    "execution_count": null,
157 |    "metadata": {},
158 |    "outputs": [],
159 |    "source": [
160 |     "eeg['measurements'][0].shape"
161 |    ]
162 |   },
163 |   {
164 |    "cell_type": "markdown",
165 |    "metadata": {},
166 |    "source": [
167 |     "## Generate the features"
168 |    ]
169 |   },
170 |   {
171 |    "cell_type": "code",
172 |    "execution_count": null,
173 |    "metadata": {},
174 |    "outputs": [],
175 |    "source": [
176 |     "# from cesium import featurize as ft\n",
177 |     "# features_to_use = [\"amplitude\",\n",
178 |     "#                    \"percent_beyond_1_std\",\n",
179 |     "#                    \"percent_close_to_median\",\n",
180 |     "#                   \"skew\",\n",
181 |     "#                   \"max_slope\"]\n",
182 |     "# fset_cesium = ft.featurize_time_series(times=eeg[\"times\"],\n",
183 |     "#                                               values=eeg[\"measurements\"],\n",
184 |     "#                                               errors=None,\n",
185 |     "#                                               features_to_use=features_to_use,\n",
186 |     "#                                              scheduler = None)"
187 |    ]
188 |   },
189 |   {
190 |    "cell_type": "code",
191 |    "execution_count": null,
192 |    "metadata": {},
193 |    "outputs": [],
194 |    "source": [
195 |     "fset_cesium = pd.read_csv(\"data/full_eeg_data_features.csv\", header = [0, 1])"
196 |    ]
197 |   },
198 |   {
199 |    "cell_type": "code",
200 |    "execution_count": null,
201 |    "metadata": {
202 |     "scrolled": true
203 |    },
204 |    "outputs": [],
205 |    "source": [
206 |     "fset_cesium.head()"
207 |    ]
208 |   },
209 |   {
210 |    "cell_type": "code",
211 |    "execution_count": null,
212 |    "metadata": {},
213 |    "outputs": [],
214 |    "source": [
215 |     "# fset_cesium.to_csv(\"full_eeg_data_features.csv\")"
216 |    ]
217 |   },
218 |   {
219 |    "cell_type": "code",
220 |    "execution_count": null,
221 |    "metadata": {},
222 |    "outputs": [],
223 |    "source": [
224 |     "fset_cesium.shape"
225 |    ]
226 |   },
227 |   {
228 |    "cell_type": "markdown",
229 |    "metadata": {},
230 |    "source": [
231 |     "## Exercise: validate/calculate these features by hand\n",
232 |     "#### look up feature definitions here: http://cesium-ml.org/docs/feature_table.html\n",
233 |     "confirm the values by hand coding these features for the first EEG measurement\n",
234 |     "(that is eeg[\"measurements\"][0])"
235 |    ]
236 |   },
237 |   {
238 |    "cell_type": "code",
239 |    "execution_count": null,
240 |    "metadata": {},
241 |    "outputs": [],
242 |    "source": [
243 |     "ex = eeg[\"measurements\"][0]\n",
244 |     "ex_mean = np.mean(ex)\n",
245 |     "ex_std  = np.std(ex)"
246 |    ]
247 |   },
248 |   {
249 |    "cell_type": "code",
250 |    "execution_count": null,
251 |    "metadata": {},
252 |    "outputs": [],
253 |    "source": []
254 |   },
255 |   {
256 |    "cell_type": "markdown",
257 |    "metadata": {},
258 |    "source": [
259 |     "## Prepare data for training"
260 |    ]
261 |   },
262 |   {
263 |    "cell_type": "code",
264 |    "execution_count": null,
265 |    "metadata": {},
266 |    "outputs": [],
267 |    "source": [
268 |     "X_train, X_test, y_train, y_test = train_test_split(\n",
269 |     "     fset_cesium.iloc[:, 1:6].values, eeg[\"classes\"], random_state=21)"
270 |    ]
271 |   },
272 |   {
273 |    "cell_type": "markdown",
274 |    "metadata": {},
275 |    "source": [
276 |     "## Try a random forest with these features"
277 |    ]
278 |   },
279 |   {
280 |    "cell_type": "code",
281 |    "execution_count": null,
282 |    "metadata": {},
283 |    "outputs": [],
284 |    "source": [
285 |     "clf = RandomForestClassifier(n_estimators=10, max_depth=3,\n",
286 |     "                              random_state=21)"
287 |    ]
288 |   },
289 |   {
290 |    "cell_type": "code",
291 |    "execution_count": null,
292 |    "metadata": {},
293 |    "outputs": [],
294 |    "source": [
295 |     "clf.fit(X_train, y_train)"
296 |    ]
297 |   },
298 |   {
299 |    "cell_type": "code",
300 |    "execution_count": null,
301 |    "metadata": {},
302 |    "outputs": [],
303 |    "source": [
304 |     "clf.score(X_train, y_train)"
305 |    ]
306 |   },
307 |   {
308 |    "cell_type": "code",
309 |    "execution_count": null,
310 |    "metadata": {},
311 |    "outputs": [],
312 |    "source": [
313 |     "clf.score(X_test, y_test)"
314 |    ]
315 |   },
316 |   {
317 |    "cell_type": "code",
318 |    "execution_count": null,
319 |    "metadata": {},
320 |    "outputs": [],
321 |    "source": [
322 |     "np.unique(y_test, return_counts=True)"
323 |    ]
324 |   },
325 |   {
326 |    "cell_type": "code",
327 |    "execution_count": null,
328 |    "metadata": {},
329 |    "outputs": [],
330 |    "source": [
331 |     "y_test"
332 |    ]
333 |   },
334 |   {
335 |    "cell_type": "code",
336 |    "execution_count": null,
337 |    "metadata": {},
338 |    "outputs": [],
339 |    "source": [
340 |     "y_test.shape"
341 |    ]
342 |   },
343 |   {
344 |    "cell_type": "code",
345 |    "execution_count": null,
346 |    "metadata": {},
347 |    "outputs": [],
348 |    "source": [
349 |     "y_train.shape"
350 |    ]
351 |   },
352 |   {
353 |    "cell_type": "markdown",
354 |    "metadata": {},
355 |    "source": [
356 |     "## Try XGBoost with these features"
357 |    ]
358 |   },
359 |   {
360 |    "cell_type": "code",
361 |    "execution_count": null,
362 |    "metadata": {},
363 |    "outputs": [],
364 |    "source": [
365 |     "model = xgb.XGBClassifier(n_estimators=10, max_depth=3,\n",
366 |     "                              random_state=21)\n",
367 |     "model.fit(X_train, y_train)"
368 |    ]
369 |   },
370 |   {
371 |    "cell_type": "code",
372 |    "execution_count": null,
373 |    "metadata": {},
374 |    "outputs": [],
375 |    "source": [
376 |     "model.score(X_test, y_test)"
377 |    ]
378 |   },
379 |   {
380 |    "cell_type": "code",
381 |    "execution_count": null,
382 |    "metadata": {},
383 |    "outputs": [],
384 |    "source": [
385 |     "model.score(X_train, y_train)"
386 |    ]
387 |   },
388 |   {
389 |    "cell_type": "code",
390 |    "execution_count": null,
391 |    "metadata": {},
392 |    "outputs": [],
393 |    "source": [
394 |     "xgb.plot_importance(model)"
395 |    ]
396 |   },
397 |   {
398 |    "cell_type": "markdown",
399 |    "metadata": {},
400 |    "source": [
401 |     "## Time Series Forecasting with Decision Trees"
402 |    ]
403 |   },
404 |   {
405 |    "cell_type": "markdown",
406 |    "metadata": {},
407 |    "source": [
408 |     "### Explore the data"
409 |    ]
410 |   },
411 |   {
412 |    "cell_type": "code",
413 |    "execution_count": null,
414 |    "metadata": {},
415 |    "outputs": [],
416 |    "source": [
417 |     "ap = pd.read_csv(\"data/AirPassengers.csv\", parse_dates=[0])"
418 |    ]
419 |   },
420 |   {
421 |    "cell_type": "code",
422 |    "execution_count": null,
423 |    "metadata": {},
424 |    "outputs": [],
425 |    "source": [
426 |     "ap.head()"
427 |    ]
428 |   },
429 |   {
430 |    "cell_type": "code",
431 |    "execution_count": null,
432 |    "metadata": {},
433 |    "outputs": [],
434 |    "source": [
435 |     "ap.set_index('Month', inplace=True)"
436 |    ]
437 |   },
438 |   {
439 |    "cell_type": "code",
440 |    "execution_count": null,
441 |    "metadata": {},
442 |    "outputs": [],
443 |    "source": [
444 |     "ap.head()"
445 |    ]
446 |   },
447 |   {
448 |    "cell_type": "code",
449 |    "execution_count": null,
450 |    "metadata": {},
451 |    "outputs": [],
452 |    "source": [
453 |     "plt.plot(ap)"
454 |    ]
455 |   },
456 |   {
457 |    "cell_type": "code",
458 |    "execution_count": null,
459 |    "metadata": {},
460 |    "outputs": [],
461 |    "source": [
462 |     "plt.plot(np.diff(np.log(ap.values[:, 0])))"
463 |    ]
464 |   },
465 |   {
466 |    "cell_type": "code",
467 |    "execution_count": null,
468 |    "metadata": {},
469 |    "outputs": [],
470 |    "source": [
471 |     "ts = np.diff(np.log(ap.values[:, 0]))"
472 |    ]
473 |   },
474 |   {
475 |    "cell_type": "markdown",
476 |    "metadata": {},
477 |    "source": [
478 |     "## Exercise: now that we have 1 time series, how can we convert it to many samples?"
479 |    ]
480 |   },
481 |   {
482 |    "cell_type": "code",
483 |    "execution_count": 5,
484 |    "metadata": {},
485 |    "outputs": [],
486 |    "source": [
487 |     "NSTEPS = 12"
488 |    ]
489 |   },
490 |   {
491 |    "cell_type": "code",
492 |    "execution_count": null,
493 |    "metadata": {},
494 |    "outputs": [],
495 |    "source": []
496 |   },
497 |   {
498 |    "cell_type": "markdown",
499 |    "metadata": {},
500 |    "source": [
501 |     "## Exercise: now that we have the time series broken down into a set of samples, how to featurize?"
502 |    ]
503 |   },
504 |   {
505 |    "cell_type": "code",
506 |    "execution_count": null,
507 |    "metadata": {},
508 |    "outputs": [],
509 |    "source": [
510 |     "measures = [vals[i][0:(NSTEPS - 1)] for i in range(vals.shape[0])]"
511 |    ]
512 |   },
513 |   {
514 |    "cell_type": "code",
515 |    "execution_count": null,
516 |    "metadata": {},
517 |    "outputs": [],
518 |    "source": [
519 |     "times = [[j for j in range(NSTEPS - 1)] for i in range(vals.shape[0])]"
520 |    ]
521 |   },
522 |   {
523 |    "cell_type": "code",
524 |    "execution_count": null,
525 |    "metadata": {},
526 |    "outputs": [],
527 |    "source": []
528 |   },
529 |   {
530 |    "cell_type": "markdown",
531 |    "metadata": {},
532 |    "source": [
533 |     "## Exercise: can you fit an XGBRegressor to this problem? Let's use the first 100 'time series' as the training data"
534 |    ]
535 |   },
536 |   {
537 |    "cell_type": "code",
538 |    "execution_count": null,
539 |    "metadata": {},
540 |    "outputs": [],
541 |    "source": []
542 |   },
543 |   {
544 |    "cell_type": "markdown",
545 |    "metadata": {},
546 |    "source": [
547 |     "### RMSE can be hard to digest .... Use other assessments to determine how well the model performs"
548 |    ]
549 |   },
550 |   {
551 |    "cell_type": "code",
552 |    "execution_count": null,
553 |    "metadata": {},
554 |    "outputs": [],
555 |    "source": []
556 |   },
557 |   {
558 |    "cell_type": "markdown",
559 |    "metadata": {},
560 |    "source": [
561 |     "### What went wrong? Let's revisit the feature set"
562 |    ]
563 |   },
564 |   {
565 |    "cell_type": "code",
566 |    "execution_count": null,
567 |    "metadata": {},
568 |    "outputs": [],
569 |    "source": [
570 |     "fset_ap.head()"
571 |    ]
572 |   },
573 |   {
574 |    "cell_type": "code",
575 |    "execution_count": null,
576 |    "metadata": {},
577 |    "outputs": [],
578 |    "source": [
579 |     "plt.plot(vals[0])\n",
580 |     "plt.plot(vals[1])\n",
581 |     "plt.plot(vals[2])"
582 |    ]
583 |   },
584 |   {
585 |    "cell_type": "markdown",
586 |    "metadata": {},
587 |    "source": [
588 |     "## We need to find a way to generate features that encode positional information"
589 |    ]
590 |   },
591 |   {
592 |    "cell_type": "markdown",
593 |    "metadata": {},
594 |    "source": [
595 |     "### now we will generate our own features"
596 |    ]
597 |   },
598 |   {
599 |    "cell_type": "code",
600 |    "execution_count": null,
601 |    "metadata": {},
602 |    "outputs": [],
603 |    "source": [
604 |     "vals.shape"
605 |    ]
606 |   },
607 |   {
608 |    "cell_type": "code",
609 |    "execution_count": null,
610 |    "metadata": {},
611 |    "outputs": [],
612 |    "source": [
613 |     "feats = np.zeros( (vals.shape[0], 6), dtype = np.float32)\n",
614 |     "for i in range(vals.shape[0]):\n",
615 |     "    feats[i, 0] = np.where(vals[i] == np.max(vals[i]))[0][0]\n",
616 |     "    feats[i, 1] = np.where(vals[i] == np.min(vals[i]))[0][0]\n",
617 |     "    feats[i, 2] = feats[i, 0] - feats[i, 1]\n",
618 |     "    feats[i, 3] = np.max(vals[i][-3:])\n",
619 |     "    feats[i, 4] = vals[i][-1] - vals[i][-2]\n",
620 |     "    feats[i, 5] = vals[i][-1] - vals[i][-3]"
621 |    ]
622 |   },
623 |   {
624 |    "cell_type": "code",
625 |    "execution_count": null,
626 |    "metadata": {},
627 |    "outputs": [],
628 |    "source": [
629 |     "feats[0:3]"
630 |    ]
631 |   },
632 |   {
633 |    "cell_type": "markdown",
634 |    "metadata": {},
635 |    "source": [
636 |     "### How do these look compared to the first set of features?"
637 |    ]
638 |   },
639 |   {
640 |    "cell_type": "code",
641 |    "execution_count": null,
642 |    "metadata": {},
643 |    "outputs": [],
644 |    "source": [
645 |     "pd.DataFrame(feats[0:3])"
646 |    ]
647 |   },
648 |   {
649 |    "cell_type": "code",
650 |    "execution_count": null,
651 |    "metadata": {},
652 |    "outputs": [],
653 |    "source": [
654 |     "X_train, y_train = feats[:100, :], outcomes[:100]\n",
655 |     "X_test, y_test   = feats[100:, :], outcomes[100:]"
656 |    ]
657 |   },
658 |   {
659 |    "cell_type": "code",
660 |    "execution_count": null,
661 |    "metadata": {},
662 |    "outputs": [],
663 |    "source": [
664 |     "model = xgb.XGBRegressor(n_estimators=20, max_depth=2,\n",
665 |     "                              random_state=21)\n",
666 |     "eval_set = [(X_test, y_test)]\n",
667 |     "model.fit(X_train, y_train, eval_metric=\"rmse\", eval_set=eval_set, verbose=True)"
668 |    ]
669 |   },
670 |   {
671 |    "cell_type": "code",
672 |    "execution_count": null,
673 |    "metadata": {},
674 |    "outputs": [],
675 |    "source": [
676 |     "plt.scatter(model.predict(X_test), y_test)"
677 |    ]
678 |   },
679 |   {
680 |    "cell_type": "code",
681 |    "execution_count": null,
682 |    "metadata": {},
683 |    "outputs": [],
684 |    "source": [
685 |     "print(pearsonr(model.predict(X_test), y_test))\n",
686 |     "print(spearmanr(model.predict(X_test), y_test))"
687 |    ]
688 |   },
689 |   {
690 |    "cell_type": "code",
691 |    "execution_count": null,
692 |    "metadata": {},
693 |    "outputs": [],
694 |    "source": [
695 |     "plt.scatter(model.predict(X_train), y_train)"
696 |    ]
697 |   },
698 |   {
699 |    "cell_type": "code",
700 |    "execution_count": null,
701 |    "metadata": {},
702 |    "outputs": [],
703 |    "source": [
704 |     "print(pearsonr(model.predict(X_train), y_train))\n",
705 |     "print(spearmanr(model.predict(X_train), y_train))"
706 |    ]
707 |   },
708 |   {
709 |    "cell_type": "code",
710 |    "execution_count": null,
711 |    "metadata": {},
712 |    "outputs": [],
713 |    "source": []
714 |   }
715 |  ],
716 |  "metadata": {
717 |   "kernelspec": {
718 |    "display_name": "Python 3",
719 |    "language": "python",
720 |    "name": "python3"
721 |   },
722 |   "language_info": {
723 |    "codemirror_mode": {
724 |     "name": "ipython",
725 |     "version": 3
726 |    },
727 |    "file_extension": ".py",
728 |    "mimetype": "text/x-python",
729 |    "name": "python",
730 |    "nbconvert_exporter": "python",
731 |    "pygments_lexer": "ipython3",
732 |    "version": "3.6.8"
733 |   }
734 |  },
735 |  "nbformat": 4,
736 |  "nbformat_minor": 2
737 | }
738 | 


--------------------------------------------------------------------------------
/modern_time_series_analysis/ModernTimeSeriesAnalysis/MachineLearning/data/50words.csv:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/john-science/scipy_con_2019/7280bc1949f90b151048c0ed127bdd656064c2cb/modern_time_series_analysis/ModernTimeSeriesAnalysis/MachineLearning/data/50words.csv


--------------------------------------------------------------------------------
/modern_time_series_analysis/ModernTimeSeriesAnalysis/MachineLearning/data/AirPassengers.csv:
--------------------------------------------------------------------------------
  1 | Month,#Passengers
  2 | 1949-01,112
  3 | 1949-02,118
  4 | 1949-03,132
  5 | 1949-04,129
  6 | 1949-05,121
  7 | 1949-06,135
  8 | 1949-07,148
  9 | 1949-08,148
 10 | 1949-09,136
 11 | 1949-10,119
 12 | 1949-11,104
 13 | 1949-12,118
 14 | 1950-01,115
 15 | 1950-02,126
 16 | 1950-03,141
 17 | 1950-04,135
 18 | 1950-05,125
 19 | 1950-06,149
 20 | 1950-07,170
 21 | 1950-08,170
 22 | 1950-09,158
 23 | 1950-10,133
 24 | 1950-11,114
 25 | 1950-12,140
 26 | 1951-01,145
 27 | 1951-02,150
 28 | 1951-03,178
 29 | 1951-04,163
 30 | 1951-05,172
 31 | 1951-06,178
 32 | 1951-07,199
 33 | 1951-08,199
 34 | 1951-09,184
 35 | 1951-10,162
 36 | 1951-11,146
 37 | 1951-12,166
 38 | 1952-01,171
 39 | 1952-02,180
 40 | 1952-03,193
 41 | 1952-04,181
 42 | 1952-05,183
 43 | 1952-06,218
 44 | 1952-07,230
 45 | 1952-08,242
 46 | 1952-09,209
 47 | 1952-10,191
 48 | 1952-11,172
 49 | 1952-12,194
 50 | 1953-01,196
 51 | 1953-02,196
 52 | 1953-03,236
 53 | 1953-04,235
 54 | 1953-05,229
 55 | 1953-06,243
 56 | 1953-07,264
 57 | 1953-08,272
 58 | 1953-09,237
 59 | 1953-10,211
 60 | 1953-11,180
 61 | 1953-12,201
 62 | 1954-01,204
 63 | 1954-02,188
 64 | 1954-03,235
 65 | 1954-04,227
 66 | 1954-05,234
 67 | 1954-06,264
 68 | 1954-07,302
 69 | 1954-08,293
 70 | 1954-09,259
 71 | 1954-10,229
 72 | 1954-11,203
 73 | 1954-12,229
 74 | 1955-01,242
 75 | 1955-02,233
 76 | 1955-03,267
 77 | 1955-04,269
 78 | 1955-05,270
 79 | 1955-06,315
 80 | 1955-07,364
 81 | 1955-08,347
 82 | 1955-09,312
 83 | 1955-10,274
 84 | 1955-11,237
 85 | 1955-12,278
 86 | 1956-01,284
 87 | 1956-02,277
 88 | 1956-03,317
 89 | 1956-04,313
 90 | 1956-05,318
 91 | 1956-06,374
 92 | 1956-07,413
 93 | 1956-08,405
 94 | 1956-09,355
 95 | 1956-10,306
 96 | 1956-11,271
 97 | 1956-12,306
 98 | 1957-01,315
 99 | 1957-02,301
100 | 1957-03,356
101 | 1957-04,348
102 | 1957-05,355
103 | 1957-06,422
104 | 1957-07,465
105 | 1957-08,467
106 | 1957-09,404
107 | 1957-10,347
108 | 1957-11,305
109 | 1957-12,336
110 | 1958-01,340
111 | 1958-02,318
112 | 1958-03,362
113 | 1958-04,348
114 | 1958-05,363
115 | 1958-06,435
116 | 1958-07,491
117 | 1958-08,505
118 | 1958-09,404
119 | 1958-10,359
120 | 1958-11,310
121 | 1958-12,337
122 | 1959-01,360
123 | 1959-02,342
124 | 1959-03,406
125 | 1959-04,396
126 | 1959-05,420
127 | 1959-06,472
128 | 1959-07,548
129 | 1959-08,559
130 | 1959-09,463
131 | 1959-10,407
132 | 1959-11,362
133 | 1959-12,405
134 | 1960-01,417
135 | 1960-02,391
136 | 1960-03,419
137 | 1960-04,461
138 | 1960-05,472
139 | 1960-06,535
140 | 1960-07,622
141 | 1960-08,606
142 | 1960-09,508
143 | 1960-10,461
144 | 1960-11,390
145 | 1960-12,432


--------------------------------------------------------------------------------
/modern_time_series_analysis/ModernTimeSeriesAnalysis/MachineLearning/data/pairwise_word_distances.npy:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/john-science/scipy_con_2019/7280bc1949f90b151048c0ed127bdd656064c2cb/modern_time_series_analysis/ModernTimeSeriesAnalysis/MachineLearning/data/pairwise_word_distances.npy


--------------------------------------------------------------------------------
/modern_time_series_analysis/ModernTimeSeriesAnalysis/MachineLearning/dtaidistance/__init__.py:
--------------------------------------------------------------------------------
 1 | # -*- coding: UTF-8 -*-
 2 | """
 3 | dtaidistance
 4 | ~~~~~~~~~~~~
 5 | 
 6 | Time series distance methods.
 7 | 
 8 | :author: Wannes Meert
 9 | :copyright: Copyright 2017 KU Leuven, DTAI Research Group.
10 | :license: Apache License, Version 2.0, see LICENSE for details.
11 | 
12 | """
13 | import logging
14 | 
15 | 
16 | logger = logging.getLogger("be.kuleuven.dtai.distance")
17 | 
18 | 
19 | from . import dtw
20 | try:
21 |     from . import dtw_c
22 | except ImportError:
23 |     # Try to compile automatically
24 |     # try:
25 |     #     import numpy as np
26 |     #     import pyximport
27 |     #     pyximport.install(setup_args={'include_dirs': np.get_include()})
28 |     #     from . import dtw_c
29 |     # except ImportError:
30 |     # logger.warning("\nDTW C variant not available.\n\n" +
31 |     #                "If you want to use the C libraries (not required, depends on cython), " +
32 |     #                "then run `cd {};python3 setup.py build_ext --inplace`.".format(dtaidistance_dir))
33 |     dtw_c = None
34 | 
35 | __version__ = "1.2.2"
36 | __author__ = "Wannes Meert"
37 | __copyright__ = "Copyright 2017-2019 KU Leuven, DTAI Research Group"
38 | __license__ = "Apache License, Version 2.0"
39 | 


--------------------------------------------------------------------------------
/modern_time_series_analysis/ModernTimeSeriesAnalysis/MachineLearning/dtaidistance/alignment.py:
--------------------------------------------------------------------------------
  1 | # -*- coding: UTF-8 -*-
  2 | """
  3 | dtaidistance.alignment
  4 | ~~~~~~~~~~~~~~~~~~~~~~
  5 | 
  6 | Sequence alignment (e.g. Needleman–Wunsch).
  7 | 
  8 | :author: Wannes Meert
  9 | :copyright: Copyright 2017-2018 KU Leuven, DTAI Research Group.
 10 | :license: Apache License, Version 2.0, see LICENSE for details.
 11 | 
 12 | """
 13 | import logging
 14 | import math
 15 | import numpy as np
 16 | 
 17 | from .dp import dp
 18 | 
 19 | 
 20 | def needleman_wunsch(s1, s2, window=None, max_dist=None,
 21 |                      max_step=None, max_length_diff=None, psi=None):
 22 |     """Needleman-Wunsch global sequence alignment.
 23 | 
 24 |     Example:
 25 | 
 26 |         >> s1 = "GATTACA"
 27 |         >> s2 = "GCATGCU"
 28 |         >> value, matrix = alignment.needleman_wunsch(s1, s2)
 29 |         >> algn, s1a, s2a = alignment.best_alignment(matrix, s1, s2)
 30 |         >> print(matrix)
 31 |            [[-0., -1., -2., -3., -4., -5., -6., -7.],
 32 |             [-1.,  1., -0., -1., -2., -3., -4., -5.],
 33 |             [-2., -0., -0.,  1., -0., -1., -2., -3.],
 34 |             [-3., -1., -1., -0.,  2.,  1., -0., -1.],
 35 |             [-4., -2., -2., -1.,  1.,  1., -0., -1.],
 36 |             [-5., -3., -3., -1., -0., -0., -0., -1.],
 37 |             [-6., -4., -2., -2., -1., -1.,  1., -0.],
 38 |             [-7., -5., -3., -1., -2., -2., -0., -0.]]
 39 |         >> print(''.join(s1a), ''.join(s2a))
 40 |             'G-ATTACA', 'GCAT-GCU'
 41 | 
 42 |     """
 43 |     value, matrix = dp(s1, s2,
 44 |                        _needleman_wunsch_fn, border=_needleman_wunsch_border,
 45 |                        penalty=0, window=window, max_dist=max_dist,
 46 |                        max_step=max_step, max_length_diff=max_length_diff, psi=psi)
 47 |     matrix = -matrix
 48 |     return value, matrix
 49 | 
 50 | 
 51 | def _needleman_wunsch_fn(v1, v2):
 52 |     """Needleman-Wunsch
 53 | 
 54 |     Match: +1 -> -1
 55 |     Mismatch or Indel: −1 -> +1
 56 | 
 57 |     The values are reversed because our general dynamic programming algorithm
 58 |     selects the minimal value instead of the maximal value.
 59 |     """
 60 |     d_indel = 1  # gap / indel
 61 |     if v1 == v2:
 62 |         d = -1  # match
 63 |     else:
 64 |         d = 1  # mismatch
 65 |     return d, d_indel
 66 | 
 67 | 
 68 | def _needleman_wunsch_border(ri, ci):
 69 |     if ri == 0:
 70 |         return ci
 71 |     if ci == 0:
 72 |         return ri
 73 |     return 0
 74 | 
 75 | 
 76 | def best_alignment(paths, s1=None, s2=None, gap="-", order=None):
 77 |     """Compute the optimal alignment from the nxm paths matrix.
 78 | 
 79 |     :param paths: Paths matrix (e.g. from needleman_wunsch)
 80 |     :param s1: First sequence, if given the aligned sequence will be created
 81 |     :param s2: Second sequence, if given the aligned sequence will be created
 82 |     :param gap: Gap symbol that is inserted into s1 and s2 to align the sequences
 83 |     :param order: Array with order of comparisons (there might be multiple optimal paths)
 84 |         The default order is 0,1,2: (-1,-1), (-1,-0), (-0,-1)
 85 |         For example, 1,0,2 is (-1,-0), (-1,-1), (-0,-1)
 86 |         There might be more optimal paths than covered by these orderings. For example,
 87 |         when using combinations of these orderings in different parts of the matrix.
 88 |     """
 89 |     i, j = int(paths.shape[0] - 1), int(paths.shape[1] - 1)
 90 |     p = []
 91 |     if paths[i, j] != -1:
 92 |         p.append((i - 1, j - 1))
 93 |     ops = [(-1,-1), (-1,-0), (-0,-1)]
 94 |     if order is None:
 95 |         order = [0, 1, 2]
 96 |     while i > 0 and j > 0:
 97 |         prev_vals = [paths[i + ops[orderi][0], j + ops[orderi][1]] for orderi in order]
 98 |         # c = np.argmax([paths[i - 1, j - 1], paths[i - 1, j], paths[i, j - 1]])
 99 |         c = int(np.argmax(prev_vals))
100 |         opi, opj = ops[order[c]]
101 |         i, j = i + opi, j + opj
102 |         if paths[i, j] != -1:
103 |             p.append((i - 1, j - 1))
104 |     p.pop()
105 |     p.reverse()
106 |     if s1 is not None:
107 |         s1a = []
108 |         s1ip = -1
109 |         for s1i, _ in p:
110 |             if s1i == s1ip + 1:
111 |                 s1a.append(s1[s1i])
112 |             else:
113 |                 s1a.append(gap)
114 |             s1ip = s1i
115 |     else:
116 |         s1a = None
117 |     if s2 is not None:
118 |         s2a = []
119 |         s2ip = -1
120 |         for _, s2i in p:
121 |             if s2i == s2ip + 1:
122 |                 s2a.append(s2[s2i])
123 |             else:
124 |                 s2a.append(gap)
125 |             s2ip = s2i
126 |     else:
127 |         s2a = None
128 | 
129 |     return p, s1a, s2a
130 | 


--------------------------------------------------------------------------------
/modern_time_series_analysis/ModernTimeSeriesAnalysis/MachineLearning/dtaidistance/clustering.py:
--------------------------------------------------------------------------------
  1 | # -*- coding: UTF-8 -*-
  2 | """
  3 | dtaidistance.clustering
  4 | ~~~~~~~~~~~~~~~~~~~~~~~
  5 | 
  6 | Time series clustering.
  7 | 
  8 | :author: Wannes Meert
  9 | :copyright: Copyright 2017 KU Leuven, DTAI Research Group.
 10 | :license: Apache License, Version 2.0, see LICENSE for details.
 11 | 
 12 | """
 13 | import logging
 14 | from pathlib import Path
 15 | from collections import deque
 16 | import numpy as np
 17 | 
 18 | from .util import SeriesContainer
 19 | 
 20 | try:
 21 |     from tqdm import tqdm
 22 | except ImportError:
 23 |     tqdm = None
 24 | 
 25 | 
 26 | logger = logging.getLogger("be.kuleuven.dtai.distance")
 27 | 
 28 | 
 29 | class Hierarchical:
 30 |     """Hierarchical clustering.
 31 | 
 32 |     Note: This method first computes the entire distance matrix. This is not ideal for extremely large
 33 |     data sets.
 34 | 
 35 |     :param dists_fun: Function to compute pairwise distance matrix between set of series.
 36 |     :param dists_options: Arguments to pass to dists_fun.
 37 |     :param max_dist: Do not merge or cluster series that are further apart than this.
 38 |     :param merge_hook: Function that is called when two series are clustered.
 39 |         The function definition is `def merge_hook(from_idx, to_idx, distance)`, where idx is the index of the series.
 40 |     :param order_hook: Function that is called to decide on the next idx out of all shortest distances
 41 |     :param show_progress: Use a tqdm progress bar
 42 |     """
 43 | 
 44 |     def __init__(self, dists_fun, dists_options, max_dist=np.inf,
 45 |                  merge_hook=None, order_hook=None, show_progress=True):
 46 |         self.dists_fun = dists_fun
 47 |         self.dists_options = dists_options
 48 |         self.max_dist = max_dist
 49 |         self.merge_hook = merge_hook
 50 |         self.order_hook = order_hook
 51 |         self.show_progress = show_progress
 52 | 
 53 |     def fit(self, series):
 54 |         """Merge sequences.
 55 | 
 56 |         :param series: Iterator over series.
 57 |         :return: Dictionary with as keys the prototype indicices and as values all the indicides of the series in
 58 |             that cluster.
 59 |         """
 60 |         nb_series = len(series)
 61 |         cluster_idx = dict()
 62 |         dists = self.dists_fun(series, **self.dists_options)
 63 |         min_value = np.min(dists)
 64 |         min_idxs = np.argwhere(dists == min_value)
 65 |         if self.order_hook:
 66 |             min_idx = self.order_hook(min_idxs)
 67 |         else:
 68 |             min_idx = min_idxs[0, :]
 69 |         deleted = set()
 70 |         cnt_merge = 0
 71 |         logger.debug('Merging patterns')
 72 |         if self.show_progress and tqdm:
 73 |             pbar = tqdm(total=dists.shape[0])
 74 |         else:
 75 |             pbar = None
 76 |         # Hierarchical clustering (distance to prototype)
 77 |         while min_value <= self.max_dist:
 78 |             cnt_merge += 1
 79 |             i1, i2 = int(min_idx[0]), int(min_idx[1])
 80 |             if self.merge_hook:
 81 |                 result = self.merge_hook(i2, i1, min_value)
 82 |                 if result:
 83 |                     i1, i2 = result
 84 |             logger.debug("Merge {} <- {} ({:.3f})".format(i1, i2, min_value))
 85 |             if i1 not in cluster_idx:
 86 |                 cluster_idx[i1] = {i1}
 87 |             if i2 in cluster_idx:
 88 |                 cluster_idx[i1].update(cluster_idx[i2])
 89 |                 del cluster_idx[i2]
 90 |             else:
 91 |                 cluster_idx[i1].add(i2)
 92 |             # if recompute:
 93 |             #     for r in range(i1):
 94 |             #         if r not in deleted and abs(len(cur_seqs[r]) - len(cur_seqs[i1])) <= max_length_diff:
 95 |             #             dists[r, i1] = self.dist(cur_seqs[r], cur_seqs[i1], **dist_opts)
 96 |             #     for c in range(i1+1, len(cur_seqs)):
 97 |             #         if c not in deleted and abs(len(cur_seqs[i1]) - len(cur_seqs[c])) <= max_length_diff:
 98 |             #             dists[i1, c] = self.dist(cur_seqs[i1], cur_seqs[c], **dist_opts)
 99 |             for r in range(i2):
100 |                 dists[r, i2] = np.inf
101 |             for c in range(i2 + 1, len(series)):
102 |                 dists[i2, c] = np.inf
103 |             deleted.add(i2)
104 |             if len(deleted) == nb_series - 1:
105 |                 break
106 |             if pbar:
107 |                 pbar.update(1)
108 |             # min_idx = np.unravel_index(np.argmin(dists), dists.shape)
109 |             # min_value = dists[min_idx]
110 |             min_value = np.min(dists)
111 |             # if np.isinf(min_value):
112 |             #     break
113 |             min_idxs = np.argwhere(dists == min_value)
114 |             if self.order_hook:
115 |                 min_idx = self.order_hook(min_idxs)
116 |             else:
117 |                 min_idx = min_idxs[0, :]
118 |         if pbar:
119 |             pbar.update(dists.shape[0] - cnt_merge)
120 | 
121 |         prototypes = []
122 |         for i in range(len(series)):
123 |             if i not in deleted:
124 |                 prototypes.append(i)
125 |                 if i not in cluster_idx:
126 |                     cluster_idx[i] = set(i)
127 |         return cluster_idx
128 | 
129 | 
130 | class BaseTree:
131 |     """Base Tree abstract class.
132 | 
133 |     Returns a datastructure compatible with the Scipy clustering methods:
134 | 
135 |     https://docs.scipy.org/doc/scipy/reference/generated/scipy.cluster.hierarchy.linkage.html
136 | 
137 |     A (n-1) by 4 matrix Z is returned. At the i-th iteration, clusters with indices Z[i, 0] and Z[i, 1] are
138 |     combined to form cluster n + i. A cluster with an index less than n corresponds to one of the original
139 |     observations. The distance between clusters Z[i, 0] and Z[i, 1] is given by Z[i, 2]. The fourth value
140 |     Z[i, 3] represents the number of original observations in the newly formed cluster.
141 |     """
142 | 
143 |     def __init__(self, **kwargs):
144 |         self.linkage = None
145 |         self.series = None
146 |         self._series_y = None
147 |         self.ts_height_factor = None
148 | 
149 |     @property
150 |     def maxnode(self):
151 |         return len(self.series) - 1 + len(self.linkage)
152 | 
153 |     def get_linkage(self, node):
154 |         if node < len(self.series):
155 |             return None
156 |         idx = int(node - len(self.series))
157 |         return self.linkage[idx]
158 | 
159 |     def plot(self, filename=None, axes=None, ts_height=10,
160 |              bottom_margin=2, top_margin=2, ts_left_margin=0, ts_sample_length=1,
161 |              tr_label_margin=3, tr_left_margin=2, ts_label_margin=0,
162 |              show_ts_label=None, show_tr_label=None,
163 |              cmap='viridis_r', ts_color=None):
164 |         """Plot the hierarchy and time series.
165 | 
166 |         :param filename: If a filename is passed, the image is written to this file.
167 |         :param axes: If a axes array is passed the image is added to this figure.
168 |             Expects axes[0] and axes[1] to be present.
169 |         :param ts_height: Height of a time series
170 |         :param bottom_margin: Margin on bottom
171 |         :param top_margin: Margin on top
172 |         :param ts_left_margin: Margin on left of time series image
173 |         :param ts_sample_length: Space between two points in the time series
174 |         :param tr_label_margin: Margin between tree split and label
175 |         :param tr_left_margin: Left margin for tree
176 |         :param ts_label_margin: Margin between start of series and label
177 |         :param show_ts_label: Show label indices. Boolean, callable or subscriptable object.
178 |             If it is a callable object, the index of the time series will be given and the
179 |             return string will be printed.
180 |         :param show_tr_label: Show tree distances. Boolean, callable or subscriptable object.
181 |             If it is a callable object, the index of the time series will be given and the
182 |             return string will be printed.
183 |         :param cmap: Matplotlib colormap name
184 |         :param ts_color: function that takes the index and returns a color
185 |             (compatible with the matplotlib.color color argument)
186 |         """
187 |         # print('linkage')
188 |         # for l in self.linkage:
189 |         #     print(l)
190 |         from matplotlib import pyplot as plt
191 |         from matplotlib.lines import Line2D
192 |         import matplotlib.colors as colors
193 |         import matplotlib.cm as cmx
194 | 
195 |         if show_ts_label is True:
196 |             show_ts_label = lambda idx: str(int(idx))
197 |         elif show_ts_label is False or show_ts_label is None:
198 |             show_ts_label = lambda idx: ""
199 |         elif callable(show_ts_label):
200 |             pass
201 |         elif hasattr(show_ts_label, "__getitem__"):
202 |             show_ts_label_prev = show_ts_label
203 |             show_ts_label = lambda idx: show_ts_label_prev[idx]
204 |         else:
205 |             raise AttributeError("Unknown type for show_ts_label, expecting boolean, subscriptable or callable, "
206 |                                  "got {}".format(type(show_ts_label)))
207 |         if show_tr_label is True:
208 |             show_tr_label = lambda dist: "{:.2f}".format(dist)
209 |         elif show_tr_label is False or show_tr_label is None:
210 |             show_tr_label = lambda dist: ""
211 |         elif callable(show_tr_label):
212 |             pass
213 |         elif hasattr(show_tr_label, "__getitem__"):
214 |             show_tr_label_prev = show_tr_label
215 |             show_tr_label = lambda idx: show_tr_label_prev[idx]
216 |         else:
217 |             raise AttributeError("Unknown type for show_ts_label, expecting boolean, subscriptable or callable, "
218 |                                  "got {}".format(type(show_ts_label)))
219 | 
220 |         self._series_y = [0] * len(self.series)
221 | 
222 |         max_dist = 0
223 |         for _, _, d, _ in self.linkage:
224 |             if not np.isinf(d):
225 |                 max_dist = max(max_dist, d)
226 | 
227 |         node_props = dict()
228 | 
229 |         max_y = self.series.get_max_y()
230 |         self.ts_height_factor = (ts_height / max_y) * 0.9
231 | 
232 |         def count(node, height):
233 |             # print('count({},{})'.format(node, height))
234 |             maxheight = None
235 |             maxcumdist = None
236 |             curdepth = None
237 |             cnt = 0
238 |             left_cnt = None
239 |             right_cnt = None
240 |             if node < len(self.series):
241 |                 # Leaf
242 |                 cnt += 1
243 |                 maxheight = height
244 |                 maxcumdist = 0
245 |                 curdepth = 0
246 |                 left_cnt = 0
247 |                 right_cnt = 0
248 |             else:
249 |                 # Inner node
250 |                 child_left, child_right, dist, cnt2 = self.get_linkage(int(node))
251 |                 child_left, child_right, cnt2 = int(child_left), int(child_right), int(cnt2)
252 |                 if child_left == child_right:
253 |                     raise Exception("Row in linkage contains same node as left and right child: {}-{}".
254 |                                     format(child_left, child_right))
255 |                 if np.isinf(dist):
256 |                     dist = 1.5*max_dist
257 |                 # Left
258 |                 nc, nmh, ncd, nmd = count(child_left, height + 1)
259 |                 cnt += nc
260 |                 maxheight = nmh
261 |                 maxcumdist = nmd + dist
262 |                 curdepth = ncd + 1
263 |                 left_cnt = nc
264 |                 # Right
265 |                 nc, nmh, ncd, nmd = count(child_right, height + 1)
266 |                 cnt += nc
267 |                 maxheight = max(maxheight, nmh)
268 |                 maxcumdist = max(maxcumdist, nmd + dist)
269 |                 curdepth = max(curdepth, ncd + 1)
270 |                 right_cnt = nc
271 |                 # if cnt != cnt2:
272 |                 #     raise Exception("Count in linkage not correct")
273 |             # print('c', node, c)
274 |             node_props[int(node)] = (cnt, curdepth, left_cnt, right_cnt, maxcumdist)
275 |             # print('count({},{}) = {}, {}, {}, {}'.format(node, height, cnt, maxheight, curdepth, maxcumdist))
276 |             return cnt, maxheight, curdepth, maxcumdist
277 | 
278 |         cnt, maxheight, curdepth, maxcumdist = count(self.maxnode, 0)
279 |         # for node, props in node_props.items():
280 |         #     print("{:<3}: {}".format(node, props))
281 | 
282 |         if axes is None:
283 |             fig, ax = plt.subplots(nrows=1, ncols=2, frameon=False)
284 |         else:
285 |             fig, ax = None, axes
286 |         ax[0].set_axis_off()
287 |         # ax[0].set_xlim(left=0, right=curdept)
288 |         ax[0].set_xlim(left=0, right=tr_left_margin + maxcumdist + 0.05)
289 |         ax[0].set_ylim(bottom=0, top=bottom_margin + ts_height * len(self.series) + top_margin)
290 |         # ax[0].plot([0,1],[1,2])
291 |         # ax[0].add_line(Line2D((0,1),(2,2), lw=2, color='black', axes=ax[0]))
292 | 
293 |         ax[1].set_axis_off()
294 |         ax[1].set_xlim(left=0, right=ts_left_margin + ts_sample_length * len(self.series[0]))
295 |         ax[1].set_ylim(bottom=0, top=bottom_margin + ts_height * len(self.series) + top_margin)
296 | 
297 |         if type(cmap) == str:
298 |             cmap = plt.get_cmap(cmap)
299 |         else:
300 |             pass
301 |         line_colors = cmx.ScalarMappable(norm=colors.Normalize(vmin=0, vmax=max_dist), cmap=cmap)
302 | 
303 |         cnt_ts = 0
304 | 
305 |         def plot_i(node, depth, cnt_ts, prev_lcnt, ax, left):
306 |             # print('plot_i', node, depth, cnt_ts, prev_lcnt)
307 |             pcnt, pdepth, plcnt, prcnt, pcdist = node_props[node]
308 |             # px = maxheight - pdepth
309 |             px = tr_left_margin + maxcumdist - pcdist
310 |             py = prev_lcnt * ts_height
311 |             if node < len(self.series):
312 |                 # Plot series
313 |                 # print('plot series y={}'.format(ts_bottom_margin + ts_height * cnt_ts + self.ts_height_factor))
314 |                 self._series_y[int(node)] = bottom_margin + ts_height * cnt_ts
315 |                 serie = self.series[int(node)]
316 |                 ax[1].text(ts_left_margin + ts_label_margin,
317 |                            bottom_margin + ts_height * cnt_ts + ts_height / 2,
318 |                            show_ts_label(int(node)), ha='left', va='center')
319 |                 if ts_color:
320 |                     curcolor = ts_color(int(node))
321 |                 else:
322 |                     curcolor = None
323 |                 ax[1].plot(ts_left_margin + ts_sample_length * np.arange(len(serie)),
324 |                            bottom_margin + ts_height * cnt_ts + self.ts_height_factor * serie,
325 |                            color=curcolor)
326 |                 cnt_ts += 1
327 | 
328 |             else:
329 |                 child_left, child_right, dist, _ = self.get_linkage(node)
330 |                 color = line_colors.to_rgba(dist)
331 |                 ax[0].text(px + tr_label_margin, py,
332 |                            show_tr_label(dist), ha='left', va='center', color=color)
333 | 
334 |                 # Left
335 |                 ccnt, cdepth, clcntl, crcntl, clcdist = node_props[child_left]
336 |                 # print('left', ccnt, cdepth, clcntl, crcntl)
337 |                 # cx = maxheight - cdepth
338 |                 cx = tr_left_margin + maxcumdist - clcdist
339 |                 cy = (prev_lcnt - crcntl) * ts_height
340 |                 if py == cy:
341 |                     cy -= 1 / 2 * ts_height
342 |                 # print('plot line', (px, cx), (py, cy))
343 |                 # ax[0].add_line(Line2D((px, cx), (py, cy), lw=2, color='black', axes=ax[0]))
344 |                 ax[0].add_line(Line2D((px, px), (py, cy), lw=1, color=color, axes=ax[0]))
345 |                 ax[0].add_line(Line2D((px, cx), (cy, cy), lw=1, color=color, axes=ax[0]))
346 |                 cnt_ts = plot_i(child_left, depth + 1, cnt_ts, prev_lcnt - crcntl, ax, True)
347 | 
348 |                 # Right
349 |                 ccnt, cdepth, clcntr, crcntr, crcdist = node_props[child_right]
350 |                 # print('right', ccnt, cdepth, clcntr, crcntr)
351 |                 # cx = maxheight - cdepth
352 |                 cx = tr_left_margin + maxcumdist - crcdist
353 |                 cy = (prev_lcnt + clcntr) * ts_height
354 |                 if py == cy:
355 |                     cy += 1 / 2 * ts_height
356 |                 # print('plot line', (px, cx), (py, cy))
357 |                 # ax[0].add_line(Line2D((px, cx), (py, cy), lw=2, color='black', axes=ax[0]))
358 |                 ax[0].add_line(Line2D((px, px), (py, cy), lw=1, color=color, axes=ax[0]))
359 |                 ax[0].add_line(Line2D((px, cx), (cy, cy), lw=1, color=color, axes=ax[0]))
360 |                 cnt_ts = plot_i(child_right, depth + 1, cnt_ts, prev_lcnt + clcntr, ax, False)
361 |             return cnt_ts
362 | 
363 |         plot_i(self.maxnode, 0, 0, node_props[self.maxnode][2], ax, True)
364 | 
365 |         if filename:
366 |             if isinstance(filename, Path):
367 |                 filename = str(filename)
368 |             plt.savefig(filename, bbox_inches='tight', pad_inches=0)
369 |             plt.close()
370 |             fig, ax = None, None
371 | 
372 |         return fig, ax
373 | 
374 |     def to_dot(self):
375 |         child_left, child_right, dist, cnt = self.get_linkage(self.maxnode)
376 |         node_deque = deque([(self.maxnode, child_left), (self.maxnode, child_right)])
377 |         # print(node_deque)
378 |         s = ["digraph tree {"]
379 |         while len(node_deque) > 0:
380 |             from_node, to_node = node_deque.popleft()
381 |             s.append("  {} -> {};".format(from_node, to_node))
382 |             if to_node >= len(self.series):
383 |                 child_left, child_right, dist, cnt = self.get_linkage(to_node)
384 |                 node_deque.append((to_node, child_left))
385 |                 node_deque.append((to_node, child_right))
386 |             # print(node_deque)
387 |         s.append("}")
388 |         return "\n".join(s)
389 | 
390 | 
391 | class HierarchicalTree(BaseTree):
392 |     """Wrapper to keep track of the full tree that represents the hierarchical clustering.
393 | 
394 |     :param model: Clustering object. For example of class :class:`Hierarchical`.
395 |         If no model is given, the arguments are identical to those of class :class:`Hierarchical`.
396 |     """
397 | 
398 |     def __init__(self, model=None, **kwargs):
399 |         if model is None:
400 |             self._model = Hierarchical(**kwargs)
401 |         else:
402 |             self._model = model
403 |         super().__init__(**kwargs)
404 |         self._model.max_dist = np.inf
405 | 
406 |     def fit(self, series, *args, **kwargs):
407 |         self.series = SeriesContainer.wrap(series)
408 |         self.linkage = []
409 |         new_nodes = {i: i for i in range(len(series))}
410 |         if self._model.merge_hook:
411 |             old_merge_hook = self._model.merge_hook
412 |         else:
413 |             old_merge_hook = None
414 | 
415 |         def merge_hook(from_idx, to_idx, distance):
416 |             # print('merge_hook', from_idx, to_idx)
417 |             new_idx = len(self.series) + len(self.linkage)
418 |             # print('adding to linkage: ', new_nodes[from_idx], new_nodes[to_idx], distance, 0)
419 |             if new_nodes[from_idx] is None:
420 |                 raise Exception('Trying to merge series that is already merged')
421 |             self.linkage.append((new_nodes[from_idx], new_nodes[to_idx], distance, 0))
422 |             new_nodes[to_idx] = new_idx
423 |             new_nodes[from_idx] = None
424 |             if old_merge_hook:
425 |                 old_merge_hook(from_idx, to_idx, distance)
426 | 
427 |         self._model.merge_hook = merge_hook
428 | 
429 |         result = self._model.fit(series, *args, **kwargs)
430 |         self._model.merge_hook = old_merge_hook
431 |         return result
432 | 
433 | 
434 | class LinkageTree(BaseTree):
435 |     """Hierarchical clustering using the Scipy linkage function.
436 | 
437 |     This is the same but faster algorithm as available in Hierarchical (~10 times faster). But with less
438 |     options to steer the clustering (e.g. no possibility to give weights). It still computes the entire
439 |     distance matrix first and is thus not ideal for extremely large data sets.
440 |     """
441 | 
442 |     def __init__(self, dists_fun, dists_options, method='complete'):
443 |         """
444 | 
445 |         :param dists_fun: Distance funcion, e.g. dtw.distance
446 |         :param dists_options: Options passed to dists_fun
447 |         :param method: Linkage method (see scipy.cluster.hierarchy.linkage)
448 |         """
449 |         super().__init__()
450 |         self.dists_fun = dists_fun
451 |         self.dists_options = dists_options
452 |         self.method = method
453 | 
454 |     def fit(self, series):
455 |         self.series = SeriesContainer.wrap(series)
456 |         try:
457 |             from scipy.cluster.hierarchy import linkage
458 |         except ImportError:
459 |             logger.error("The LinkageTree class requires the scipy package to be installed.")
460 |             self.linkage = None
461 |             linkage = None
462 |             return
463 |         dists = self.dists_fun(self.series, **self.dists_options)
464 |         dists_cond = np.zeros(self._size_cond(len(series)))
465 |         idx = 0
466 |         for r in range(len(series) - 1):
467 |             dists_cond[idx:idx + len(series) - r - 1] = dists[r, r + 1:]
468 |             idx += len(series) - r - 1
469 | 
470 |         self.linkage = linkage(dists_cond, method=self.method, metric='euclidean')
471 | 
472 |     def _size_cond(self, size):
473 |         n = int(size)
474 |         return int((n * (n - 1)) / 2)
475 | 
476 | 
477 | class Hooks:
478 |     @staticmethod
479 |     def create_weighthook(weights, series):
480 |         def newhook(i1, i2, dist):
481 |             w1 = weights[i1]
482 |             w2 = weights[i2]
483 |             p1 = series[i1]
484 |             p2 = series[i2]
485 |             if w1 < w2 or (w1 == w2 and len(p1) > len(p2)):
486 |                 i1, i2 = i2, i1
487 |             weights[i1] = w1 + w2
488 |             return i1, i2
489 |         return newhook
490 | 
491 |     @staticmethod
492 |     def create_orderhook(weights):
493 |         def newhook(idxs):
494 |             min_idx = -1
495 |             max_weight = -1
496 |             for r, c in [idxs[ii, :] for ii in range(idxs.shape[0])]:
497 |                 total = weights[r] + weights[c]
498 |                 if total > max_weight:
499 |                     max_weight = total
500 |                     min_idx = (r, c)
501 |             return min_idx
502 |         return newhook
503 | 


--------------------------------------------------------------------------------
/modern_time_series_analysis/ModernTimeSeriesAnalysis/MachineLearning/dtaidistance/dp.py:
--------------------------------------------------------------------------------
  1 | # -*- coding: UTF-8 -*-
  2 | """
  3 | dtaidistance.dp
  4 | ~~~~~~~~~~~~~~~
  5 | 
  6 | Generic Dynamic Programming functions
  7 | 
  8 | :author: Wannes Meert
  9 | :copyright: Copyright 2017-2018 KU Leuven, DTAI Research Group.
 10 | :license: Apache License, Version 2.0, see LICENSE for details.
 11 | 
 12 | """
 13 | import logging
 14 | import numpy as np
 15 | 
 16 | 
 17 | logger = logging.getLogger("be.kuleuven.dtai.distance")
 18 | 
 19 | 
 20 | def dp(s1, s2, fn, border=None, window=None, max_dist=None,
 21 |        max_step=None, max_length_diff=None, penalty=None, psi=None):
 22 |     """
 23 |     Generic dynamic programming.
 24 | 
 25 |     This function does not optimize storage when a window size is given (e.g. in contrast with
 26 |     the fast DTW functions).
 27 | 
 28 |     :param s1: First sequence
 29 |     :param s2: Second sequence
 30 |     :param fn: Function to compare two items from both sequences
 31 |     :param border: Callable object to fill in the initial borders (border(row_idx, col_idx).
 32 |     :param window: see :meth:`distance`
 33 |     :param max_dist: see :meth:`distance`
 34 |     :param max_step: see :meth:`distance`
 35 |     :param max_length_diff: see :meth:`distance`
 36 |     :param penalty: see :meth:`distance`
 37 |     :param psi: see :meth:`distance`
 38 |     :returns: (DTW distance, DTW matrix)
 39 |     """
 40 |     r, c = len(s1), len(s2)
 41 |     if max_length_diff is not None and abs(r - c) > max_length_diff:
 42 |         return np.inf
 43 |     if window is None:
 44 |         window = max(r, c)
 45 |     if not max_step:
 46 |         max_step = np.inf
 47 |     if not max_dist:
 48 |         max_dist = np.inf
 49 |     if not penalty:
 50 |         penalty = 0
 51 |     if psi is None:
 52 |         psi = 0
 53 |     dtw = np.full((r + 1, c + 1), np.inf)
 54 |     if border:
 55 |         for ci in range(c + 1):
 56 |             dtw[0, ci] = border(0, ci)
 57 |         for ri in range(1, r + 1):
 58 |             dtw[ri, 0] = border(ri, 0)
 59 |     for i in range(psi + 1):
 60 |         dtw[0, i] = 0
 61 |         dtw[i, 0] = 0
 62 |     last_under_max_dist = 0
 63 |     i0 = 1
 64 |     i1 = 0
 65 |     for i in range(r):
 66 |         if last_under_max_dist == -1:
 67 |             prev_last_under_max_dist = np.inf
 68 |         else:
 69 |             prev_last_under_max_dist = last_under_max_dist
 70 |         last_under_max_dist = -1
 71 |         i0 = i
 72 |         i1 = i + 1
 73 |         for j in range(max(0, i - max(0, r - c) - window + 1), min(c, i + max(0, c - r) + window)):
 74 |             d, d_indel = fn(s1[i], s2[j])
 75 |             if max_step is not None:
 76 |                 if d > max_step:
 77 |                     d = np.inf
 78 |                 if d_indel > max_step:
 79 |                     d_indel = np.inf
 80 |                 if d > max_step and d_indel > max_step:
 81 |                     continue
 82 |             # print(f"[{i1},{j+1}] -> [{s1[i]},{s2[j]}] -> {d},{d_indel}")
 83 |             dtw[i1, j + 1] = min(d + dtw[i0, j],
 84 |                                  d_indel + dtw[i0, j + 1] + penalty,
 85 |                                  d_indel + dtw[i1, j] + penalty)
 86 |             if max_dist is not None:
 87 |                 if dtw[i1, j + 1] <= max_dist:
 88 |                     last_under_max_dist = j
 89 |                 else:
 90 |                     dtw[i1, j + 1] = np.inf
 91 |                     if prev_last_under_max_dist < j + 1:
 92 |                         break
 93 |         if max_dist is not None and last_under_max_dist == -1:
 94 |             return np.inf, dtw
 95 |     if psi == 0:
 96 |         d = dtw[i1, min(c, c + window - 1)]
 97 |     else:
 98 |         ir = i1
 99 |         ic = min(c, c + window - 1)
100 |         vr = dtw[ir-psi:ir+1, ic]
101 |         vc = dtw[ir, ic-psi:ic+1]
102 |         mir = np.argmin(vr)
103 |         mic = np.argmin(vc)
104 |         if vr[mir] < vc[mic]:
105 |             dtw[ir-psi+mir+1:ir+1, ic] = -1
106 |             d = vr[mir]
107 |         else:
108 |             dtw[ir, ic - psi + mic + 1:ic+1] = -1
109 |             d = vc[mic]
110 |     return d, dtw
111 | 


--------------------------------------------------------------------------------
/modern_time_series_analysis/ModernTimeSeriesAnalysis/MachineLearning/dtaidistance/dtw_ndim.py:
--------------------------------------------------------------------------------
  1 | # -*- coding: UTF-8 -*-
  2 | """
  3 | dtaidistance.dtw_ndim
  4 | ~~~~~~~~~~~~~~~~~~~~~
  5 | 
  6 | Dynamic Time Warping (DTW) for N-dimensional series.
  7 | 
  8 | :author: Wannes Meert
  9 | :copyright: Copyright 2017-2018 KU Leuven, DTAI Research Group.
 10 | :license: Apache License, Version 2.0, see LICENSE for details.
 11 | 
 12 | """
 13 | import os
 14 | import logging
 15 | import math
 16 | import numpy as np
 17 | 
 18 | logger = logging.getLogger("be.kuleuven.dtai.distance")
 19 | dtaidistance_dir = os.path.join(os.path.abspath(os.path.dirname(__file__)), os.pardir)
 20 | 
 21 | try:
 22 |     from . import dtw_c
 23 | except ImportError:
 24 |     # logger.info('C library not available')
 25 |     dtw_c = None
 26 | 
 27 | try:
 28 |     from tqdm import tqdm
 29 | except ImportError:
 30 |     logger.info('tqdm library not available')
 31 |     tqdm = None
 32 | 
 33 | 
 34 | def distance(s1, s2, window=None, max_dist=None,
 35 |              max_step=None, max_length_diff=None, penalty=None, psi=None,
 36 |              use_c=False):
 37 |     """Dynamic Time Warping using multidimensional sequences.
 38 | 
 39 |     cost = EuclideanDistance(s1[i], s2[j])
 40 | 
 41 |     See :py:meth:`dtaidistance.dtw.distance` for parameters.
 42 |     """
 43 |     if use_c:
 44 |         logger.error("No C version implemented (yet)")
 45 |         return
 46 |     r, c = len(s1), len(s2)
 47 |     if max_length_diff is not None and abs(r - c) > max_length_diff:
 48 |         return np.inf
 49 |     if window is None:
 50 |         window = max(r, c)
 51 |     if not max_step:
 52 |         max_step = np.inf
 53 |     else:
 54 |         max_step *= max_step
 55 |     if not max_dist:
 56 |         max_dist = np.inf
 57 |     else:
 58 |         max_dist *= max_dist
 59 |     if not penalty:
 60 |         penalty = 0
 61 |     else:
 62 |         penalty *= penalty
 63 |     if psi is None:
 64 |         psi = 0
 65 |     length = min(c + 1, abs(r - c) + 2 * (window - 1) + 1 + 1 + 1)
 66 |     # print("length (py) = {}".format(length))
 67 |     dtw = np.full((2, length), np.inf)
 68 |     # dtw[0, 0] = 0
 69 |     for i in range(psi + 1):
 70 |         dtw[0, i] = 0
 71 |     last_under_max_dist = 0
 72 |     skip = 0
 73 |     i0 = 1
 74 |     i1 = 0
 75 |     psi_shortest = np.inf
 76 |     for i in range(r):
 77 |         # print("i={}".format(i))
 78 |         # print(dtw)
 79 |         if last_under_max_dist == -1:
 80 |             prev_last_under_max_dist = np.inf
 81 |         else:
 82 |             prev_last_under_max_dist = last_under_max_dist
 83 |         last_under_max_dist = -1
 84 |         skipp = skip
 85 |         skip = max(0, i - max(0, r - c) - window + 1)
 86 |         i0 = 1 - i0
 87 |         i1 = 1 - i1
 88 |         dtw[i1, :] = np.inf
 89 |         j_start = max(0, i - max(0, r - c) - window + 1)
 90 |         j_end = min(c, i + max(0, c - r) + window)
 91 |         if dtw.shape[1] == c + 1:
 92 |             skip = 0
 93 |         if psi != 0 and j_start == 0 and i < psi:
 94 |             dtw[i1, 0] = 0
 95 |         for j in range(j_start, j_end):
 96 |             d = np.sum((s1[i] - s2[j]) ** 2)
 97 |             if d > max_step:
 98 |                 continue
 99 |             assert j + 1 - skip >= 0
100 |             assert j - skipp >= 0
101 |             assert j + 1 - skipp >= 0
102 |             assert j - skip >= 0
103 |             dtw[i1, j + 1 - skip] = d + min(dtw[i0, j - skipp],
104 |                                             dtw[i0, j + 1 - skipp] + penalty,
105 |                                             dtw[i1, j - skip] + penalty)
106 |             # print('({},{}), ({},{}), ({},{})'.format(i0, j - skipp, i0, j + 1 - skipp, i1, j - skip))
107 |             # print('{}, {}, {}'.format(dtw[i0, j - skipp], dtw[i0, j + 1 - skipp], dtw[i1, j - skip]))
108 |             # print('i={}, j={}, d={}, skip={}, skipp={}'.format(i,j,d,skip,skipp))
109 |             # print(dtw)
110 |             if dtw[i1, j + 1 - skip] <= max_dist:
111 |                 last_under_max_dist = j
112 |             else:
113 |                 # print('above max_dist', dtw[i1, j + 1 - skip], i1, j + 1 - skip)
114 |                 dtw[i1, j + 1 - skip] = np.inf
115 |                 if prev_last_under_max_dist + 1 - skipp < j + 1 - skip:
116 |                     # print("break")
117 |                     break
118 |         if last_under_max_dist == -1:
119 |             # print('early stop')
120 |             # print(dtw)
121 |             return np.inf
122 |         if psi != 0 and j_end == len(s2) and len(s1) - 1 - i <= psi:
123 |             psi_shortest = min(psi_shortest, dtw[i1, length - 1])
124 |     if psi == 0:
125 |         d = math.sqrt(dtw[i1, min(c, c + window - 1) - skip])
126 |     else:
127 |         ic = min(c, c + window - 1) - skip
128 |         vc = dtw[i1, ic - psi:ic + 1]
129 |         d = min(np.min(vc), psi_shortest)
130 |         d = math.sqrt(d)
131 |     return d
132 | 
133 | 
134 | def _distance_with_params(t):
135 |     return distance(t[0], t[1], **t[2])
136 | 
137 | 
138 | def warping_paths(s1, s2, window=None, max_dist=None,
139 |                   max_step=None, max_length_diff=None, penalty=None, psi=None,):
140 |     """
141 |     Dynamic Time Warping (keep full matrix) using multidimensional sequences.
142 | 
143 |     cost = EuclideanDistance(s1[i], s2[j])
144 | 
145 |     See :py:meth:`dtaidistance.dtw.warping_paths` for parameters.
146 |     """
147 |     r, c = len(s1), len(s2)
148 |     if max_length_diff is not None and abs(r - c) > max_length_diff:
149 |         return np.inf
150 |     if window is None:
151 |         window = max(r, c)
152 |     if not max_step:
153 |         max_step = np.inf
154 |     else:
155 |         max_step *= max_step
156 |     if not max_dist:
157 |         max_dist = np.inf
158 |     else:
159 |         max_dist *= max_dist
160 |     if not penalty:
161 |         penalty = 0
162 |     else:
163 |         penalty *= penalty
164 |     if psi is None:
165 |         psi = 0
166 |     dtw = np.full((r + 1, c + 1), np.inf)
167 |     # dtw[0, 0] = 0
168 |     for i in range(psi + 1):
169 |         dtw[0, i] = 0
170 |         dtw[i, 0] = 0
171 |     last_under_max_dist = 0
172 |     i0 = 1
173 |     i1 = 0
174 |     for i in range(r):
175 |         if last_under_max_dist == -1:
176 |             prev_last_under_max_dist = np.inf
177 |         else:
178 |             prev_last_under_max_dist = last_under_max_dist
179 |         last_under_max_dist = -1
180 |         i0 = i
181 |         i1 = i + 1
182 |         # print('i =', i, 'skip =',skip, 'skipp =', skipp)
183 |         # jmin = max(0, i - max(0, r - c) - window + 1)
184 |         # jmax = min(c, i + max(0, c - r) + window)
185 |         # print(i,jmin,jmax)
186 |         # x = dtw[i, jmin-skipp:jmax-skipp]
187 |         # y = dtw[i, jmin+1-skipp:jmax+1-skipp]
188 |         # print(x,y,dtw[i+1, jmin+1-skip:jmax+1-skip])
189 |         # dtw[i+1, jmin+1-skip:jmax+1-skip] = np.minimum(x,
190 |         #                                                y)
191 |         for j in range(max(0, i - max(0, r - c) - window + 1), min(c, i + max(0, c - r) + window)):
192 |             # print('j =', j, 'max=',min(c, c - r + i + window))
193 |             d = np.sum((s1[i] - s2[j]) ** 2)
194 |             if max_step is not None and d > max_step:
195 |                 continue
196 |             # print(i, j + 1 - skip, j - skipp, j + 1 - skipp, j - skip)
197 |             dtw[i1, j + 1] = d + min(dtw[i0, j],
198 |                                      dtw[i0, j + 1] + penalty,
199 |                                      dtw[i1, j] + penalty)
200 |             # dtw[i + 1, j + 1 - skip] = d + min(dtw[i + 1, j + 1 - skip], dtw[i + 1, j - skip])
201 |             if max_dist is not None:
202 |                 if dtw[i1, j + 1] <= max_dist:
203 |                     last_under_max_dist = j
204 |                 else:
205 |                     dtw[i1, j + 1] = np.inf
206 |                     if prev_last_under_max_dist < j + 1:
207 |                         break
208 |         if max_dist is not None and last_under_max_dist == -1:
209 |             # print('early stop')
210 |             # print(dtw)
211 |             return np.inf, dtw
212 |     dtw = np.sqrt(dtw)
213 |     if psi == 0:
214 |         d = dtw[i1, min(c, c + window - 1)]
215 |     else:
216 |         ir = i1
217 |         ic = min(c, c + window - 1)
218 |         vr = dtw[ir-psi:ir+1, ic]
219 |         vc = dtw[ir, ic-psi:ic+1]
220 |         mir = np.argmin(vr)
221 |         mic = np.argmin(vc)
222 |         if vr[mir] < vc[mic]:
223 |             dtw[ir-psi+mir+1:ir+1, ic] = -1
224 |             d = vr[mir]
225 |         else:
226 |             dtw[ir, ic - psi + mic + 1:ic+1] = -1
227 |             d = vc[mic]
228 |     return d, dtw
229 | 
230 | 
231 | def distance_matrix(s, max_dist=None, max_length_diff=None,
232 |                     window=None, max_step=None, penalty=None, psi=None,
233 |                     block=None, parallel=False,
234 |                     use_c=False, show_progress=False):
235 |     """Dynamic Time Warping distance matrix using multidimensional sequences.
236 | 
237 |     cost = EuclideanDistance(s1[i], s2[j])
238 | 
239 |     See :py:meth:`dtaidistance.dtw.distance_matrix` for parameters.
240 |     """
241 |     if parallel and not use_c:
242 |         try:
243 |             import multiprocessing as mp
244 |             logger.info('Using multiprocessing')
245 |         except ImportError:
246 |             parallel = False
247 |             mp = None
248 |     else:
249 |         mp = None
250 |     dist_opts = {
251 |         'max_dist': max_dist,
252 |         'max_step': max_step,
253 |         'window': window,
254 |         'max_length_diff': max_length_diff,
255 |         'penalty': penalty,
256 |         'psi': psi
257 |     }
258 |     dists = None
259 |     if max_length_diff is None:
260 |         max_length_diff = np.inf
261 |     large_value = np.inf
262 |     logger.info('Computing distances')
263 |     if use_c:
264 |         logger.error("No C version available (yet)")
265 |     if not use_c:
266 |         logger.info("Compute distances in Python")
267 |         if isinstance(s, np.ndarray) and len(s.shape) == 2:
268 |             ss = [np.asarray(s[i]).reshape(-1) for i in range(s.shape[0])]
269 |             s = ss
270 |         if parallel:
271 |             logger.info("Use parallel computation")
272 |             dists = np.zeros((len(s), len(s))) + large_value
273 |             if block is None:
274 |                 idxs = np.triu_indices(len(s), k=1)
275 |             else:
276 |                 idxsl_r = []
277 |                 idxsl_c = []
278 |                 for r in range(block[0][0], block[0][1]):
279 |                     for c in range(max(r + 1, block[1][0]), min(len(s), block[1][1])):
280 |                         idxsl_r.append(r)
281 |                         idxsl_c.append(c)
282 |                 idxs = (np.array(idxsl_r), np.array(idxsl_c))
283 |             with mp.Pool() as p:
284 |                 dists[idxs] = p.map(_distance_with_params, [(s[r], s[c], dist_opts) for c, r in zip(*idxs)])
285 |                 # pbar = tqdm(total=int((len(s)*(len(s)-1)/2)))
286 |                 # for r in range(len(s)):
287 |                 #     dists[r,r+1:len(s)] = p.map(distance, [(s[r],s[c], dist_opts) for c in range(r+1,len(cur))])
288 |                 #     pbar.update(len(s) - r - 1)
289 |                 # pbar.close()
290 |         else:
291 |             logger.info("Use serial computation")
292 |             dists = np.zeros((len(s), len(s))) + large_value
293 |             if block is None:
294 |                 it_r = range(len(s))
295 |             else:
296 |                 it_r = range(block[0][0], block[0][1])
297 |             if show_progress:
298 |                 it_r = tqdm(it_r)
299 |             for r in it_r:
300 |                 if block is None:
301 |                     it_c = range(r + 1, len(s))
302 |                 else:
303 |                     it_c = range(max(r + 1, block[1][0]), min(len(s), block[1][1]))
304 |                 for c in it_c:
305 |                     if abs(len(s[r]) - len(s[c])) <= max_length_diff:
306 |                         dists[r, c] = distance(s[r], s[c], **dist_opts)
307 |     return dists
308 | 


--------------------------------------------------------------------------------
/modern_time_series_analysis/ModernTimeSeriesAnalysis/MachineLearning/dtaidistance/dtw_ndim_visualisation.py:
--------------------------------------------------------------------------------
  1 | # -*- coding: UTF-8 -*-
  2 | """
  3 | dtaidistance.dtw_visualisation
  4 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  5 | 
  6 | Dynamic Time Warping (DTW) visualisations.
  7 | 
  8 | :author: Wannes Meert
  9 | :copyright: Copyright 2017 KU Leuven, DTAI Research Group.
 10 | :license: Apache License, Version 2.0, see LICENSE for details.
 11 | 
 12 | """
 13 | import os
 14 | import logging
 15 | import math
 16 | import numpy as np
 17 | 
 18 | from .util import dtaidistance_dir
 19 | 
 20 | logger = logging.getLogger("be.kuleuven.dtai.distance")
 21 | 
 22 | from . import dtw
 23 | try:
 24 |     from . import dtw_c
 25 | except ImportError:
 26 |     # logger.info('C library not available')
 27 |     dtw_c = None
 28 | 
 29 | try:
 30 |     from tqdm import tqdm
 31 | except ImportError:
 32 |     logger.info('tqdm library not available')
 33 |     tqdm = None
 34 | 
 35 | 
 36 | def plot_warping(s1, s2, path, filename=None):
 37 |     """Plot the optimal warping between to sequences.
 38 | 
 39 |     :param s1: From sequence.
 40 |     :param s2: To sequence.
 41 |     :param path: Optimal warping path.
 42 |     :param filename: Filename path (optional).
 43 |     """
 44 |     import matplotlib.pyplot as plt
 45 |     import matplotlib as mpl
 46 |     fig, ax = plt.subplots(nrows=2, ncols=1, sharex=True, sharey=True)
 47 |     ax[0].pcolormesh(np.transpose(s1))
 48 |     ax[1].pcolormesh(np.transpose(s2))
 49 |     transFigure = fig.transFigure.inverted()
 50 |     lines = []
 51 |     line_options = {'linewidth': 2, 'color': 'orange', 'alpha': 0.8}
 52 |     for r_c, c_c in path:
 53 |         if r_c < 0 or c_c < 0:
 54 |             continue
 55 |         coord1 = transFigure.transform(ax[0].transData.transform([r_c+.5, 0]))
 56 |         coord2 = transFigure.transform(ax[1].transData.transform([c_c+.5, 0]))
 57 |         lines.append(mpl.lines.Line2D((coord1[0], coord2[0]), (coord1[1], coord2[1]),
 58 |                                       transform=fig.transFigure, **line_options))
 59 |     fig.lines = lines
 60 |     if filename:
 61 |         plt.savefig(filename)
 62 |         plt.close()
 63 |         fig, ax = None, None
 64 |     return fig, ax
 65 | 
 66 | 
 67 | def plot_warpingpaths(s1, s2, paths, path=None, filename=None, shownumbers=False):
 68 |     """Plot the warping paths matrix.
 69 | 
 70 |     :param s1: Series 1
 71 |     :param s2: Series 2
 72 |     :param paths: Warping paths matrix
 73 |     :param path: Path to draw (typically this is the best path)
 74 |     :param filename: Filename for the image (optional)
 75 |     :param shownumbers: Show distances also as numbers
 76 |     """
 77 |     from matplotlib import pyplot as plt
 78 |     from matplotlib import gridspec
 79 | 
 80 |     fig = plt.figure(figsize=(10, 10), frameon=True)
 81 |     gs = gridspec.GridSpec(2, 2, wspace=1, hspace=1,
 82 |                            left=0, right=1.0, bottom=0, top=1.0,
 83 |                            height_ratios=[1, 6],
 84 |                            width_ratios=[1, 6])
 85 | 
 86 |     if path is None:
 87 |         p = dtw.best_path(paths)
 88 |     else:
 89 |         p = path
 90 | 
 91 |     ax0 = fig.add_subplot(gs[0, 0])
 92 |     ax0.set_axis_off()
 93 |     ax0.text(0, 0, "Dist = {:.4f}".format(paths[p[-1][0], p[-1][1]]))
 94 |     ax0.xaxis.set_major_locator(plt.NullLocator())
 95 |     ax0.yaxis.set_major_locator(plt.NullLocator())
 96 | 
 97 |     # Top time series
 98 |     ax1 = fig.add_subplot(gs[0, 1:])
 99 |     ax1.set_ylim([0, s2.shape[1]])
100 |     ax1.set_axis_off()
101 |     ax1.xaxis.tick_top()
102 |     ax1.pcolormesh(np.transpose(s2))
103 |     ax1.xaxis.set_major_locator(plt.NullLocator())
104 |     ax1.yaxis.set_major_locator(plt.NullLocator())
105 | 
106 |     # Left time series
107 |     ax2 = fig.add_subplot(gs[1:, 0])
108 |     ax2.set_xlim([0, s1.shape[1]])
109 |     ax2.set_axis_off()
110 |     ax2.xaxis.set_major_locator(plt.NullLocator())
111 |     ax2.yaxis.set_major_locator(plt.NullLocator())
112 |     ax2.pcolormesh(np.flipud(s1))
113 | 
114 |     ax3 = fig.add_subplot(gs[1:, 1:])
115 |     ax3.matshow(paths[1:, 1:])
116 |     py, px = zip(*p)
117 |     ax3.plot(px, py, ".-", color="red")
118 |     if shownumbers:
119 |         for r in range(1, paths.shape[0]):
120 |             for c in range(1, paths.shape[1]):
121 |                 ax3.text(c - 1, r - 1, "{:.2f}".format(paths[r, c]))
122 | 
123 |     gs.tight_layout(fig, pad=1.0, h_pad=1.0, w_pad=1.0)
124 | 
125 |     ax = fig.axes
126 | 
127 |     if filename:
128 |         if type(filename) != str:
129 |             filename = str(filename)
130 |         plt.savefig(filename)
131 |         plt.close()
132 |         fig, ax = None, None
133 |     return fig, ax


--------------------------------------------------------------------------------
/modern_time_series_analysis/ModernTimeSeriesAnalysis/MachineLearning/dtaidistance/dtw_visualisation.py:
--------------------------------------------------------------------------------
  1 | # -*- coding: UTF-8 -*-
  2 | """
  3 | dtaidistance.dtw_visualisation
  4 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  5 | 
  6 | Dynamic Time Warping (DTW) visualisations.
  7 | 
  8 | :author: Wannes Meert
  9 | :copyright: Copyright 2017 KU Leuven, DTAI Research Group.
 10 | :license: Apache License, Version 2.0, see LICENSE for details.
 11 | 
 12 | """
 13 | import os
 14 | import logging
 15 | import math
 16 | import numpy as np
 17 | 
 18 | from .util import dtaidistance_dir
 19 | 
 20 | logger = logging.getLogger("be.kuleuven.dtai.distance")
 21 | 
 22 | from . import dtw
 23 | try:
 24 |     from . import dtw_c
 25 | except ImportError:
 26 |     # logger.info('C library not available')
 27 |     dtw_c = None
 28 | 
 29 | try:
 30 |     from tqdm import tqdm
 31 | except ImportError:
 32 |     logger.info('tqdm library not available')
 33 |     tqdm = None
 34 | 
 35 | 
 36 | def plot_warp(from_s, to_s, new_s, path, filename=None):
 37 |     """Plot the warped sequence and its relation to the original sequence
 38 |     and the target sequence.
 39 | 
 40 |     :param from_s: From sequence.
 41 |     :param to_s: To sequence.
 42 |     :param new_s: Warped version of from sequence.
 43 |     :param path: Optimal warping path.
 44 |     :param filename: Filename path (optional).
 45 |     """
 46 |     try:
 47 |         import matplotlib.pyplot as plt
 48 |         import matplotlib as mpl
 49 |     except ImportError:
 50 |         logger.error("The plot_warp function requires the matplotlib package to be installed.")
 51 |         return
 52 |     fig, ax = plt.subplots(nrows=3, ncols=1, sharex=True, sharey=True)
 53 |     ax[0].plot(from_s, label="From")
 54 |     ax[0].legend()
 55 |     ax[1].plot(to_s, label="To")
 56 |     ax[1].legend()
 57 |     transFigure = fig.transFigure.inverted()
 58 |     lines = []
 59 |     line_options = {'linewidth': 0.5, 'color': 'orange', 'alpha': 0.8}
 60 |     for r_c, c_c in path:
 61 |         if r_c < 0 or c_c < 0:
 62 |             continue
 63 |         coord1 = transFigure.transform(ax[0].transData.transform([r_c, from_s[r_c]]))
 64 |         coord2 = transFigure.transform(ax[1].transData.transform([c_c, to_s[c_c]]))
 65 |         lines.append(mpl.lines.Line2D((coord1[0], coord2[0]), (coord1[1], coord2[1]),
 66 |                                       transform=fig.transFigure, **line_options))
 67 |     ax[2].plot(new_s, label="From-warped")
 68 |     ax[2].legend()
 69 |     for i in range(len(to_s)):
 70 |         coord1 = transFigure.transform(ax[1].transData.transform([i, to_s[i]]))
 71 |         coord2 = transFigure.transform(ax[2].transData.transform([i, new_s[i]]))
 72 |         lines.append(mpl.lines.Line2D((coord1[0], coord2[0]), (coord1[1], coord2[1]),
 73 |                                       transform=fig.transFigure, **line_options))
 74 |     fig.lines = lines
 75 |     if filename:
 76 |         plt.savefig(filename)
 77 |         plt.close()
 78 |         fig, ax = None, None
 79 |     return fig, ax
 80 | 
 81 | 
 82 | def plot_warping(s1, s2, path, filename=None):
 83 |     """Plot the optimal warping between to sequences.
 84 | 
 85 |     :param s1: From sequence.
 86 |     :param s2: To sequence.
 87 |     :param path: Optimal warping path.
 88 |     :param filename: Filename path (optional).
 89 |     """
 90 |     import matplotlib.pyplot as plt
 91 |     import matplotlib as mpl
 92 |     fig, ax = plt.subplots(nrows=2, ncols=1, sharex=True, sharey=True)
 93 |     ax[0].plot(s1)
 94 |     ax[1].plot(s2)
 95 |     transFigure = fig.transFigure.inverted()
 96 |     lines = []
 97 |     line_options = {'linewidth': 0.5, 'color': 'orange', 'alpha': 0.8}
 98 |     for r_c, c_c in path:
 99 |         if r_c < 0 or c_c < 0:
100 |             continue
101 |         coord1 = transFigure.transform(ax[0].transData.transform([r_c, s1[r_c]]))
102 |         coord2 = transFigure.transform(ax[1].transData.transform([c_c, s2[c_c]]))
103 |         lines.append(mpl.lines.Line2D((coord1[0], coord2[0]), (coord1[1], coord2[1]),
104 |                                       transform=fig.transFigure, **line_options))
105 |     fig.lines = lines
106 |     if filename:
107 |         plt.savefig(filename)
108 |         plt.close()
109 |         fig, ax = None, None
110 |     return fig, ax
111 | 
112 | 
113 | def plot_warpingpaths(s1, s2, paths, path=None, filename=None, shownumbers=False):
114 |     """Plot the warping paths matrix.
115 | 
116 |     :param s1: Series 1
117 |     :param s2: Series 2
118 |     :param paths: Warping paths matrix
119 |     :param path: Path to draw (typically this is the best path)
120 |     :param filename: Filename for the image (optional)
121 |     :param shownumbers: Show distances also as numbers
122 |     """
123 |     from matplotlib import pyplot as plt
124 |     from matplotlib import gridspec
125 |     from matplotlib.ticker import FuncFormatter
126 | 
127 |     ratio = max(len(s1), len(s2))
128 |     min_y = min(np.min(s1), np.min(s2))
129 |     max_y = max(np.max(s1), np.max(s2))
130 | 
131 |     fig = plt.figure(figsize=(10, 10), frameon=True)
132 |     gs = gridspec.GridSpec(2, 2, wspace=1, hspace=1,
133 |                            left=0, right=1.0, bottom=0, top=1.0,
134 |                            height_ratios=[1, 6],
135 |                            width_ratios=[1, 6])
136 |     max_s2_x = np.max(s2)
137 |     max_s2_y = len(s2)
138 |     max_s1_x = np.max(s1)
139 |     min_s1_x = np.min(s1)
140 |     max_s1_y = len(s1)
141 | 
142 |     if path is None:
143 |         p = dtw.best_path(paths)
144 |     else:
145 |         p = path
146 | 
147 |     def format_fn2_x(tick_val, tick_pos):
148 |         return max_s2_x - tick_val
149 | 
150 |     def format_fn2_y(tick_val, tick_pos):
151 |         return int(max_s2_y - tick_val)
152 | 
153 |     ax0 = fig.add_subplot(gs[0, 0])
154 |     ax0.set_axis_off()
155 |     ax0.text(0, 0, "Dist = {:.4f}".format(paths[p[-1][0], p[-1][1]]))
156 |     ax0.xaxis.set_major_locator(plt.NullLocator())
157 |     ax0.yaxis.set_major_locator(plt.NullLocator())
158 | 
159 |     ax1 = fig.add_subplot(gs[0, 1:])
160 |     ax1.set_ylim([min_y, max_y])
161 |     ax1.set_axis_off()
162 |     ax1.xaxis.tick_top()
163 |     # ax1.set_aspect(0.454)
164 |     ax1.plot(range(len(s2)), s2, ".-")
165 |     ax1.xaxis.set_major_locator(plt.NullLocator())
166 |     ax1.yaxis.set_major_locator(plt.NullLocator())
167 | 
168 |     ax2 = fig.add_subplot(gs[1:, 0])
169 |     ax2.set_xlim([-max_y, -min_y])
170 |     ax2.set_axis_off()
171 |     # ax2.set_aspect(0.8)
172 |     # ax2.xaxis.set_major_formatter(FuncFormatter(format_fn2_x))
173 |     # ax2.yaxis.set_major_formatter(FuncFormatter(format_fn2_y))
174 |     ax2.xaxis.set_major_locator(plt.NullLocator())
175 |     ax2.yaxis.set_major_locator(plt.NullLocator())
176 |     ax2.plot(-s1, range(max_s1_y, 0, -1), ".-")
177 | 
178 |     ax3 = fig.add_subplot(gs[1:, 1:])
179 |     # ax3.set_aspect(1)
180 |     ax3.matshow(paths[1:, 1:])
181 |     # ax3.grid(which='major', color='w', linestyle='-', linewidth=0)
182 |     # ax3.set_axis_off()
183 |     py, px = zip(*p)
184 |     ax3.plot(px, py, ".-", color="red")
185 |     # ax3.xaxis.set_major_locator(plt.NullLocator())
186 |     # ax3.yaxis.set_major_locator(plt.NullLocator())
187 |     if shownumbers:
188 |         for r in range(1, paths.shape[0]):
189 |             for c in range(1, paths.shape[1]):
190 |                 ax3.text(c - 1, r - 1, "{:.2f}".format(paths[r, c]))
191 | 
192 |     gs.tight_layout(fig, pad=1.0, h_pad=1.0, w_pad=1.0)
193 |     # fig.subplots_adjust(hspace=0, wspace=0)
194 | 
195 |     ax = fig.axes
196 | 
197 |     if filename:
198 |         if type(filename) != str:
199 |             filename = str(filename)
200 |         plt.savefig(filename)
201 |         plt.close()
202 |         fig, ax = None, None
203 |     return fig, ax
204 | 
205 | def plot_matrix(distances, filename=None, ax=None, shownumbers=False):
206 |     from matplotlib import pyplot as plt
207 | 
208 |     if ax is None:
209 |         if shownumbers:
210 |             figsize = (15, 15)
211 |         else:
212 |             figsize = None
213 |         fig, ax = plt.subplots(nrows=1, ncols=1, figsize=figsize)
214 |     else:
215 |         fig = None
216 | 
217 |     ax.xaxis.set_ticks_position('top')
218 |     ax.yaxis.set_ticks_position('both')
219 | 
220 |     im = ax.imshow(distances)
221 |     idxs = [str(i) for i in range(len(distances))]
222 |     # Show all ticks
223 |     ax.set_xticks(np.arange(len(idxs)))
224 |     ax.set_xticklabels(idxs)
225 |     ax.set_yticks(np.arange(len(idxs)))
226 |     ax.set_yticklabels(idxs)
227 | 
228 |     ax.set_title("Distances between series", pad=30)
229 | 
230 |     if shownumbers:
231 |         for i in range(len(idxs)):
232 |             for j in range(len(idxs)):
233 |                 if not np.isinf(distances[i, j]):
234 |                     l = "{:.2f}".format(distances[i, j])
235 |                     ax.text(j, i, l, ha="center", va="center", color="w")
236 | 
237 |     if filename:
238 |         if type(filename) != str:
239 |             filename = str(filename)
240 |         plt.savefig(filename)
241 |         plt.close()
242 |         fig, ax = None, None
243 |     return fig, ax
244 | 


--------------------------------------------------------------------------------
/modern_time_series_analysis/ModernTimeSeriesAnalysis/MachineLearning/dtaidistance/util.py:
--------------------------------------------------------------------------------
  1 | # -*- coding: UTF-8 -*-
  2 | """
  3 | dtaidistance.util
  4 | ~~~~~~~~~~~~~~~~~
  5 | 
  6 | Utility functions for DTAIDistance.
  7 | 
  8 | :author: Wannes Meert
  9 | :copyright: Copyright 2017-2018 KU Leuven, DTAI Research Group.
 10 | :license: Apache License, Version 2.0, see LICENSE for details.
 11 | 
 12 | """
 13 | import os
 14 | import sys
 15 | import logging
 16 | from array import array
 17 | from pathlib import Path
 18 | import tempfile
 19 | 
 20 | import numpy as np
 21 | 
 22 | 
 23 | logger = logging.getLogger("be.kuleuven.dtai.distance")
 24 | 
 25 | 
 26 | dtaidistance_dir = os.path.abspath(os.path.dirname(__file__))
 27 | 
 28 | 
 29 | def prepare_directory(directory=None):
 30 |     """Prepare the given directory, create it if necessary.
 31 |     If no directory is given, a new directory will be created in the system's temp directory.
 32 |     """
 33 |     if directory is not None:
 34 |         directory = Path(directory)
 35 |         if not directory.exists():
 36 |             directory.mkdir(parents=True)
 37 |         logger.debug("Using directory: {}".format(directory))
 38 |         return Path(directory)
 39 |     directory = tempfile.mkdtemp(prefix="dtaidistance_")
 40 |     logger.debug("Using directory: {}".format(directory))
 41 |     return Path(directory)
 42 | 
 43 | 
 44 | class SeriesContainer:
 45 |     def __init__(self, series):
 46 |         """Container for a list of series.
 47 | 
 48 |         This wrapper class knows how to deal with multiple types of datastructures to represent
 49 |         a list of sequences:
 50 |         - List[array.array]
 51 |         - List[numpy.array]
 52 |         - List[List]
 53 |         - numpy.array
 54 |         - numpy.matrix
 55 | 
 56 |         When using the C-based extensions, the data is automatically verified and converted.
 57 |         """
 58 |         if isinstance(series, SeriesContainer):
 59 |             self.series = series.series
 60 |         elif isinstance(series, np.ndarray) and len(series.shape) == 2:
 61 |             # A matrix always returns a 2D array, also if you select one row (to be consistent
 62 |             # and always be a matrix datastructure). The methods in this toolbox expect a
 63 |             # 1D array thus we need to convert to a 1D or 2D array.
 64 |             # self.series = [np.asarray(series[i]).reshape(-1) for i in range(series.shape[0])]
 65 |             self.series = np.asarray(series, order='C')
 66 |         elif type(series) == set or type(series) == tuple:
 67 |             self.series = list(series)
 68 |         else:
 69 |             self.series = series
 70 | 
 71 |     def c_data(self):
 72 |         """Return a datastructure that the C-component knows how to handle.
 73 |         The method tries to avoid copying or reallocating memory.
 74 | 
 75 |         :return: Either a list of buffers or a two-dimensional buffer. The
 76 |             buffers are guaranteed to be C-contiguous and can thus be used
 77 |             as regular pointer-based arrays in C.
 78 |         """
 79 |         if type(self.series) == list:
 80 |             for i in range(len(self.series)):
 81 |                 serie = self.series[i]
 82 |                 if isinstance(serie, np.ndarray):
 83 |                     if not serie.flags.c_contiguous:
 84 |                         serie = np.asarray(serie, order='C')
 85 |                         self.series[i] = serie
 86 |                 elif isinstance(serie, array):
 87 |                     pass
 88 |                 else:
 89 |                     raise Exception("Type of series not supported, "
 90 |                                     "expected numpy.array or array.array but got {}".format(type(serie)))
 91 |         elif isinstance(self.series, np.ndarray):
 92 |             if not self.series.flags.c_contiguous:
 93 |                 self.series = self.series.copy(order='C')
 94 |         return self.series
 95 | 
 96 |     def get_max_y(self):
 97 |         max_y = 0
 98 |         if isinstance(self.series, np.ndarray):
 99 |             max_y = max(np.max(self.series), abs(np.min(self.series)))
100 |         else:
101 |             for serie in self.series:
102 |                 max_y = max(max_y, np.max(serie), abs(np.min(serie)))
103 |         return max_y
104 | 
105 |     def __getitem__(self, item):
106 |         return self.series[item]
107 | 
108 |     def __len__(self):
109 |         return len(self.series)
110 | 
111 |     def __str__(self):
112 |         return "SeriesContainer:\n{}".format(self.series)
113 | 
114 |     @staticmethod
115 |     def wrap(series):
116 |         if isinstance(series, SeriesContainer):
117 |             return series
118 |         return SeriesContainer(series)
119 | 
120 | 
121 | def recompile():
122 |     import subprocess as sp
123 |     sp.run([sys.executable, 'setup.py', 'build_ext', '--inplace'], cwd=dtaidistance_dir)
124 | 


--------------------------------------------------------------------------------
/modern_time_series_analysis/ModernTimeSeriesAnalysis/StateSpaceModels/1_Structural_Time_Series_INSTRUCTOR.ipynb:
--------------------------------------------------------------------------------
  1 | {
  2 |  "cells": [
  3 |   {
  4 |    "cell_type": "code",
  5 |    "execution_count": null,
  6 |    "metadata": {},
  7 |    "outputs": [],
  8 |    "source": [
  9 |     "%matplotlib inline\n",
 10 |     "import matplotlib\n",
 11 |     "matplotlib.rcParams['figure.figsize'] = [8, 3]\n",
 12 |     "import matplotlib.pyplot as plt\n",
 13 |     "\n",
 14 |     "import pandas as pd\n",
 15 |     "import numpy as np\n",
 16 |     "import statsmodels.api as sm\n",
 17 |     "import statsmodels\n",
 18 |     "\n",
 19 |     "import scipy\n",
 20 |     "from scipy.stats import pearsonr\n",
 21 |     "\n",
 22 |     "from pandas.plotting import register_matplotlib_converters\n",
 23 |     "register_matplotlib_converters()"
 24 |    ]
 25 |   },
 26 |   {
 27 |    "cell_type": "code",
 28 |    "execution_count": null,
 29 |    "metadata": {},
 30 |    "outputs": [],
 31 |    "source": [
 32 |     "print(matplotlib.__version__)\n",
 33 |     "print(pd.__version__)\n",
 34 |     "print(np.__version__)\n",
 35 |     "print(statsmodels.__version__)\n",
 36 |     "print(scipy.__version__)\n"
 37 |    ]
 38 |   },
 39 |   {
 40 |    "cell_type": "markdown",
 41 |    "metadata": {},
 42 |    "source": [
 43 |     "## Obtain and visualize data"
 44 |    ]
 45 |   },
 46 |   {
 47 |    "cell_type": "code",
 48 |    "execution_count": null,
 49 |    "metadata": {},
 50 |    "outputs": [],
 51 |    "source": [
 52 |     "## data obtained from https://datahub.io/core/global-temp#data\n",
 53 |     "df = pd.read_csv(\"global_temps.csv\")\n",
 54 |     "df.head()"
 55 |    ]
 56 |   },
 57 |   {
 58 |    "cell_type": "code",
 59 |    "execution_count": null,
 60 |    "metadata": {},
 61 |    "outputs": [],
 62 |    "source": [
 63 |     "df.Mean[:100].plot()"
 64 |    ]
 65 |   },
 66 |   {
 67 |    "cell_type": "markdown",
 68 |    "metadata": {},
 69 |    "source": [
 70 |     "## Exercise: what is wrong with the data and plot above? How can we fix this?"
 71 |    ]
 72 |   },
 73 |   {
 74 |    "cell_type": "code",
 75 |    "execution_count": null,
 76 |    "metadata": {},
 77 |    "outputs": [],
 78 |    "source": [
 79 |     " df = df.pivot(index='Date', columns='Source', values='Mean')"
 80 |    ]
 81 |   },
 82 |   {
 83 |    "cell_type": "code",
 84 |    "execution_count": null,
 85 |    "metadata": {},
 86 |    "outputs": [],
 87 |    "source": [
 88 |     "df.head()"
 89 |    ]
 90 |   },
 91 |   {
 92 |    "cell_type": "code",
 93 |    "execution_count": null,
 94 |    "metadata": {},
 95 |    "outputs": [],
 96 |    "source": [
 97 |     "df.GCAG.plot()"
 98 |    ]
 99 |   },
100 |   {
101 |    "cell_type": "code",
102 |    "execution_count": null,
103 |    "metadata": {},
104 |    "outputs": [],
105 |    "source": [
106 |     "type(df.index)"
107 |    ]
108 |   },
109 |   {
110 |    "cell_type": "markdown",
111 |    "metadata": {},
112 |    "source": [
113 |     "## Exercise: how can we make the index more time aware?"
114 |    ]
115 |   },
116 |   {
117 |    "cell_type": "code",
118 |    "execution_count": null,
119 |    "metadata": {},
120 |    "outputs": [],
121 |    "source": [
122 |     "df.index = pd.to_datetime(df.index)"
123 |    ]
124 |   },
125 |   {
126 |    "cell_type": "code",
127 |    "execution_count": null,
128 |    "metadata": {},
129 |    "outputs": [],
130 |    "source": [
131 |     "type(df.index)"
132 |    ]
133 |   },
134 |   {
135 |    "cell_type": "code",
136 |    "execution_count": null,
137 |    "metadata": {},
138 |    "outputs": [],
139 |    "source": [
140 |     "df.GCAG.plot()"
141 |    ]
142 |   },
143 |   {
144 |    "cell_type": "code",
145 |    "execution_count": null,
146 |    "metadata": {},
147 |    "outputs": [],
148 |    "source": [
149 |     "df['1880']"
150 |    ]
151 |   },
152 |   {
153 |    "cell_type": "code",
154 |    "execution_count": null,
155 |    "metadata": {},
156 |    "outputs": [],
157 |    "source": [
158 |     "plt.plot(df['1880':'1950'][['GCAG', 'GISTEMP']])"
159 |    ]
160 |   },
161 |   {
162 |    "cell_type": "code",
163 |    "execution_count": null,
164 |    "metadata": {},
165 |    "outputs": [],
166 |    "source": [
167 |     "plt.plot(df['1950':][['GISTEMP']])"
168 |    ]
169 |   },
170 |   {
171 |    "cell_type": "markdown",
172 |    "metadata": {},
173 |    "source": [
174 |     "## Exercise: How strongly do these measurements correlate contemporaneously? What about with a time lag?"
175 |    ]
176 |   },
177 |   {
178 |    "cell_type": "code",
179 |    "execution_count": null,
180 |    "metadata": {},
181 |    "outputs": [],
182 |    "source": [
183 |     "plt.scatter(df['1880':'1900'][['GCAG']], df['1880':'1900'][['GISTEMP']])"
184 |    ]
185 |   },
186 |   {
187 |    "cell_type": "code",
188 |    "execution_count": null,
189 |    "metadata": {},
190 |    "outputs": [],
191 |    "source": [
192 |     "plt.scatter(df['1880':'1899'][['GCAG']], df['1881':'1900'][['GISTEMP']])"
193 |    ]
194 |   },
195 |   {
196 |    "cell_type": "code",
197 |    "execution_count": null,
198 |    "metadata": {},
199 |    "outputs": [],
200 |    "source": [
201 |     "pearsonr(df['1880':'1899'].GCAG, df['1881':'1900'].GISTEMP)"
202 |    ]
203 |   },
204 |   {
205 |    "cell_type": "code",
206 |    "execution_count": null,
207 |    "metadata": {},
208 |    "outputs": [],
209 |    "source": [
210 |     "df['1880':'1899'][['GCAG']].head()"
211 |    ]
212 |   },
213 |   {
214 |    "cell_type": "code",
215 |    "execution_count": null,
216 |    "metadata": {},
217 |    "outputs": [],
218 |    "source": [
219 |     "df['1881':'1900'][['GISTEMP']].head()"
220 |    ]
221 |   },
222 |   {
223 |    "cell_type": "code",
224 |    "execution_count": null,
225 |    "metadata": {},
226 |    "outputs": [],
227 |    "source": [
228 |     "min(df.index)"
229 |    ]
230 |   },
231 |   {
232 |    "cell_type": "code",
233 |    "execution_count": null,
234 |    "metadata": {},
235 |    "outputs": [],
236 |    "source": [
237 |     "max(df.index)"
238 |    ]
239 |   },
240 |   {
241 |    "cell_type": "markdown",
242 |    "metadata": {},
243 |    "source": [
244 |     "## Unobserved component model"
245 |    ]
246 |   },
247 |   {
248 |    "cell_type": "code",
249 |    "execution_count": null,
250 |    "metadata": {},
251 |    "outputs": [],
252 |    "source": [
253 |     "train = df['1960':]"
254 |    ]
255 |   },
256 |   {
257 |    "cell_type": "markdown",
258 |    "metadata": {},
259 |    "source": [
260 |     "### model parameters"
261 |    ]
262 |   },
263 |   {
264 |    "cell_type": "code",
265 |    "execution_count": null,
266 |    "metadata": {},
267 |    "outputs": [],
268 |    "source": [
269 |     "# smooth trend model without seasonal or cyclical components\n",
270 |     "model = {\n",
271 |     "    'level': 'smooth trend', 'cycle': False, 'seasonal': None, \n",
272 |     "}\n"
273 |    ]
274 |   },
275 |   {
276 |    "cell_type": "markdown",
277 |    "metadata": {},
278 |    "source": [
279 |     "### fitting a model"
280 |    ]
281 |   },
282 |   {
283 |    "cell_type": "code",
284 |    "execution_count": null,
285 |    "metadata": {
286 |     "scrolled": true
287 |    },
288 |    "outputs": [],
289 |    "source": [
290 |     "# https://www.statsmodels.org/dev/generated/statsmodels.tsa.statespace.structural.UnobservedComponents.html\n",
291 |     "gcag_mod = sm.tsa.UnobservedComponents(train['GCAG'], **model)\n",
292 |     "gcag_res = gcag_mod.fit()"
293 |    ]
294 |   },
295 |   {
296 |    "cell_type": "code",
297 |    "execution_count": null,
298 |    "metadata": {},
299 |    "outputs": [],
300 |    "source": [
301 |     "fig = gcag_res.plot_components(legend_loc='lower right', figsize=(15, 9));"
302 |    ]
303 |   },
304 |   {
305 |    "cell_type": "markdown",
306 |    "metadata": {},
307 |    "source": [
308 |     "## Plotting predictions"
309 |    ]
310 |   },
311 |   {
312 |    "cell_type": "code",
313 |    "execution_count": null,
314 |    "metadata": {},
315 |    "outputs": [],
316 |    "source": [
317 |     "# Perform rolling prediction and multistep forecast\n",
318 |     "num_steps = 20\n",
319 |     "predict_res = gcag_res.get_prediction(dynamic=train['GCAG'].shape[0] - num_steps)\n",
320 |     "\n",
321 |     "predict = predict_res.predicted_mean\n",
322 |     "ci = predict_res.conf_int()"
323 |    ]
324 |   },
325 |   {
326 |    "cell_type": "code",
327 |    "execution_count": null,
328 |    "metadata": {},
329 |    "outputs": [],
330 |    "source": [
331 |     "plt.plot(predict)"
332 |    ]
333 |   },
334 |   {
335 |    "cell_type": "code",
336 |    "execution_count": null,
337 |    "metadata": {},
338 |    "outputs": [],
339 |    "source": [
340 |     "plt.scatter(train['GCAG'], predict)"
341 |    ]
342 |   },
343 |   {
344 |    "cell_type": "code",
345 |    "execution_count": null,
346 |    "metadata": {},
347 |    "outputs": [],
348 |    "source": [
349 |     "fig, ax = plt.subplots()\n",
350 |     "# Plot the results\n",
351 |     "ax.plot(train['GCAG'], 'k.', label='Observations');\n",
352 |     "ax.plot(train.index[:-num_steps], predict[:-num_steps], label='One-step-ahead Prediction');\n",
353 |     "\n",
354 |     "ax.plot(train.index[-num_steps:], predict[-num_steps:], 'r', label='Multistep Prediction');\n",
355 |     "ax.plot(train.index[-num_steps:], ci.iloc[-num_steps:], 'k--');\n",
356 |     "\n",
357 |     "# Cleanup the image\n",
358 |     "legend = ax.legend(loc='upper left');"
359 |    ]
360 |   },
361 |   {
362 |    "cell_type": "code",
363 |    "execution_count": null,
364 |    "metadata": {},
365 |    "outputs": [],
366 |    "source": [
367 |     "fig, ax = plt.subplots()\n",
368 |     "# Plot the results\n",
369 |     "ax.plot(train.index[-40:], train['GCAG'][-40:], 'k.', label='Observations');\n",
370 |     "ax.plot(train.index[-40:-num_steps], predict[-40:-num_steps], label='One-step-ahead Prediction');\n",
371 |     "\n",
372 |     "ax.plot(train.index[-num_steps:], predict[-num_steps:], 'r', label='Multistep Prediction');\n",
373 |     "ax.plot(train.index[-num_steps:], ci.iloc[-num_steps:], 'k--');\n",
374 |     "\n",
375 |     "# Cleanup the image\n",
376 |     "legend = ax.legend(loc='upper left');"
377 |    ]
378 |   },
379 |   {
380 |    "cell_type": "markdown",
381 |    "metadata": {},
382 |    "source": [
383 |     "## Exercise: consider adding a seasonal term for 12 periods for the model fit above. Does this improve the fit of the model?"
384 |    ]
385 |   },
386 |   {
387 |    "cell_type": "code",
388 |    "execution_count": null,
389 |    "metadata": {},
390 |    "outputs": [],
391 |    "source": [
392 |     "seasonal_model = {\n",
393 |     "    'level': 'local linear trend',\n",
394 |     "    'seasonal': 12\n",
395 |     "}\n",
396 |     "mod = sm.tsa.UnobservedComponents(train['GCAG'], **seasonal_model)\n",
397 |     "res = mod.fit(method='powell', disp=False)"
398 |    ]
399 |   },
400 |   {
401 |    "cell_type": "code",
402 |    "execution_count": null,
403 |    "metadata": {},
404 |    "outputs": [],
405 |    "source": [
406 |     "fig = res.plot_components(legend_loc='lower right', figsize=(15, 9));"
407 |    ]
408 |   },
409 |   {
410 |    "cell_type": "markdown",
411 |    "metadata": {},
412 |    "source": [
413 |     "## How does this compare to the original model?"
414 |    ]
415 |   },
416 |   {
417 |    "cell_type": "code",
418 |    "execution_count": null,
419 |    "metadata": {},
420 |    "outputs": [],
421 |    "source": [
422 |     "pearsonr(gcag_res.predict(), train['GCAG'])"
423 |    ]
424 |   },
425 |   {
426 |    "cell_type": "code",
427 |    "execution_count": null,
428 |    "metadata": {},
429 |    "outputs": [],
430 |    "source": [
431 |     "np.mean(np.abs(gcag_res.predict() - train['GCAG']))"
432 |    ]
433 |   },
434 |   {
435 |    "cell_type": "code",
436 |    "execution_count": null,
437 |    "metadata": {},
438 |    "outputs": [],
439 |    "source": [
440 |     "np.mean(np.abs(res.predict() - train['GCAG']))"
441 |    ]
442 |   },
443 |   {
444 |    "cell_type": "markdown",
445 |    "metadata": {},
446 |    "source": [
447 |     "## Explore the seasonality more"
448 |    ]
449 |   },
450 |   {
451 |    "cell_type": "code",
452 |    "execution_count": null,
453 |    "metadata": {},
454 |    "outputs": [],
455 |    "source": [
456 |     "seasonal_model = {\n",
457 |     "    'level': 'local level',\n",
458 |     "    'seasonal': 12\n",
459 |     "}\n",
460 |     "llmod = sm.tsa.UnobservedComponents(train['GCAG'], **seasonal_model)\n",
461 |     "ll_level_res = llmod.fit(method='powell', disp=False)"
462 |    ]
463 |   },
464 |   {
465 |    "cell_type": "code",
466 |    "execution_count": null,
467 |    "metadata": {},
468 |    "outputs": [],
469 |    "source": [
470 |     "fig = ll_level_res.plot_components(legend_loc='lower right', figsize=(15, 9));"
471 |    ]
472 |   },
473 |   {
474 |    "cell_type": "code",
475 |    "execution_count": null,
476 |    "metadata": {},
477 |    "outputs": [],
478 |    "source": [
479 |     "np.mean(np.abs(ll_level_res.predict() - train['GCAG']))"
480 |    ]
481 |   },
482 |   {
483 |    "cell_type": "code",
484 |    "execution_count": null,
485 |    "metadata": {},
486 |    "outputs": [],
487 |    "source": [
488 |     "train[:48].GCAG.plot()"
489 |    ]
490 |   },
491 |   {
492 |    "cell_type": "code",
493 |    "execution_count": null,
494 |    "metadata": {},
495 |    "outputs": [],
496 |    "source": []
497 |   },
498 |   {
499 |    "cell_type": "markdown",
500 |    "metadata": {},
501 |    "source": [
502 |     "## Exercise: a common null model for time series is to predict the value at time t-1 for the value at time t. How does such a model compare to the models we fit here?"
503 |    ]
504 |   },
505 |   {
506 |    "cell_type": "markdown",
507 |    "metadata": {},
508 |    "source": [
509 |     "### Consider correlation"
510 |    ]
511 |   },
512 |   {
513 |    "cell_type": "code",
514 |    "execution_count": null,
515 |    "metadata": {},
516 |    "outputs": [],
517 |    "source": [
518 |     "pearsonr(ll_level_res.predict(), train['GCAG'])"
519 |    ]
520 |   },
521 |   {
522 |    "cell_type": "code",
523 |    "execution_count": null,
524 |    "metadata": {},
525 |    "outputs": [],
526 |    "source": [
527 |     "pearsonr(train['GCAG'].iloc[:-1, ], train['GCAG'].iloc[1:, ])"
528 |    ]
529 |   },
530 |   {
531 |    "cell_type": "markdown",
532 |    "metadata": {},
533 |    "source": [
534 |     "### What about mean absolute error?"
535 |    ]
536 |   },
537 |   {
538 |    "cell_type": "code",
539 |    "execution_count": null,
540 |    "metadata": {},
541 |    "outputs": [],
542 |    "source": [
543 |     "np.mean(np.abs(ll_level_res.predict() - train['GCAG']))"
544 |    ]
545 |   },
546 |   {
547 |    "cell_type": "code",
548 |    "execution_count": null,
549 |    "metadata": {},
550 |    "outputs": [],
551 |    "source": [
552 |     "np.mean(np.abs(train['GCAG'].iloc[:-1, ].values, train['GCAG'].iloc[1:, ].values))"
553 |    ]
554 |   },
555 |   {
556 |    "cell_type": "code",
557 |    "execution_count": null,
558 |    "metadata": {},
559 |    "outputs": [],
560 |    "source": []
561 |   }
562 |  ],
563 |  "metadata": {
564 |   "kernelspec": {
565 |    "display_name": "Python 3",
566 |    "language": "python",
567 |    "name": "python3"
568 |   },
569 |   "language_info": {
570 |    "codemirror_mode": {
571 |     "name": "ipython",
572 |     "version": 3
573 |    },
574 |    "file_extension": ".py",
575 |    "mimetype": "text/x-python",
576 |    "name": "python",
577 |    "nbconvert_exporter": "python",
578 |    "pygments_lexer": "ipython3",
579 |    "version": "3.6.8"
580 |   }
581 |  },
582 |  "nbformat": 4,
583 |  "nbformat_minor": 2
584 | }
585 | 


--------------------------------------------------------------------------------
/modern_time_series_analysis/ModernTimeSeriesAnalysis/StateSpaceModels/1_Structural_Time_Series_STUDENT.ipynb:
--------------------------------------------------------------------------------
  1 | {
  2 |  "cells": [
  3 |   {
  4 |    "cell_type": "code",
  5 |    "execution_count": null,
  6 |    "metadata": {},
  7 |    "outputs": [],
  8 |    "source": [
  9 |     "%matplotlib inline\n",
 10 |     "import matplotlib\n",
 11 |     "matplotlib.rcParams['figure.figsize'] = [8, 3]\n",
 12 |     "import matplotlib.pyplot as plt\n",
 13 |     "\n",
 14 |     "import pandas as pd\n",
 15 |     "import numpy as np\n",
 16 |     "import statsmodels.api as sm\n",
 17 |     "import statsmodels\n",
 18 |     "\n",
 19 |     "import scipy\n",
 20 |     "from scipy.stats import pearsonr\n",
 21 |     "\n",
 22 |     "from pandas.plotting import register_matplotlib_converters\n",
 23 |     "register_matplotlib_converters()"
 24 |    ]
 25 |   },
 26 |   {
 27 |    "cell_type": "code",
 28 |    "execution_count": null,
 29 |    "metadata": {},
 30 |    "outputs": [],
 31 |    "source": [
 32 |     "print(matplotlib.__version__)\n",
 33 |     "print(pd.__version__)\n",
 34 |     "print(np.__version__)\n",
 35 |     "print(statsmodels.__version__)\n",
 36 |     "print(scipy.__version__)\n"
 37 |    ]
 38 |   },
 39 |   {
 40 |    "cell_type": "markdown",
 41 |    "metadata": {},
 42 |    "source": [
 43 |     "## Obtain and visualize data"
 44 |    ]
 45 |   },
 46 |   {
 47 |    "cell_type": "code",
 48 |    "execution_count": null,
 49 |    "metadata": {},
 50 |    "outputs": [],
 51 |    "source": [
 52 |     "## data obtained from https://datahub.io/core/global-temp#data\n",
 53 |     "df = pd.read_csv(\"global_temps.csv\")\n",
 54 |     "df.head()"
 55 |    ]
 56 |   },
 57 |   {
 58 |    "cell_type": "code",
 59 |    "execution_count": null,
 60 |    "metadata": {},
 61 |    "outputs": [],
 62 |    "source": [
 63 |     "df.Mean[:100].plot()"
 64 |    ]
 65 |   },
 66 |   {
 67 |    "cell_type": "markdown",
 68 |    "metadata": {},
 69 |    "source": [
 70 |     "## Exercise: what is wrong with the data and plot above? How can we fix this?"
 71 |    ]
 72 |   },
 73 |   {
 74 |    "cell_type": "code",
 75 |    "execution_count": null,
 76 |    "metadata": {},
 77 |    "outputs": [],
 78 |    "source": []
 79 |   },
 80 |   {
 81 |    "cell_type": "markdown",
 82 |    "metadata": {},
 83 |    "source": [
 84 |     "## Exercise: how can we make the index more time aware?"
 85 |    ]
 86 |   },
 87 |   {
 88 |    "cell_type": "code",
 89 |    "execution_count": null,
 90 |    "metadata": {},
 91 |    "outputs": [],
 92 |    "source": []
 93 |   },
 94 |   {
 95 |    "cell_type": "markdown",
 96 |    "metadata": {},
 97 |    "source": [
 98 |     "## Exercise: How strongly do these measurements correlate contemporaneously? What about with a time lag?"
 99 |    ]
100 |   },
101 |   {
102 |    "cell_type": "code",
103 |    "execution_count": null,
104 |    "metadata": {},
105 |    "outputs": [],
106 |    "source": []
107 |   },
108 |   {
109 |    "cell_type": "markdown",
110 |    "metadata": {},
111 |    "source": [
112 |     "## Unobserved component model"
113 |    ]
114 |   },
115 |   {
116 |    "cell_type": "code",
117 |    "execution_count": null,
118 |    "metadata": {},
119 |    "outputs": [],
120 |    "source": [
121 |     "train = df['1960':]"
122 |    ]
123 |   },
124 |   {
125 |    "cell_type": "markdown",
126 |    "metadata": {},
127 |    "source": [
128 |     "### model parameters"
129 |    ]
130 |   },
131 |   {
132 |    "cell_type": "code",
133 |    "execution_count": null,
134 |    "metadata": {},
135 |    "outputs": [],
136 |    "source": [
137 |     "# smooth trend model without seasonal or cyclical components\n",
138 |     "model = {\n",
139 |     "    'level': 'smooth trend', 'cycle': False, 'seasonal': None, \n",
140 |     "}\n"
141 |    ]
142 |   },
143 |   {
144 |    "cell_type": "markdown",
145 |    "metadata": {},
146 |    "source": [
147 |     "### fitting a model"
148 |    ]
149 |   },
150 |   {
151 |    "cell_type": "code",
152 |    "execution_count": null,
153 |    "metadata": {
154 |     "scrolled": true
155 |    },
156 |    "outputs": [],
157 |    "source": [
158 |     "# https://www.statsmodels.org/dev/generated/statsmodels.tsa.statespace.structural.UnobservedComponents.html\n",
159 |     "gcag_mod = sm.tsa.UnobservedComponents(train['GCAG'], **model)\n",
160 |     "gcag_res = gcag_mod.fit()"
161 |    ]
162 |   },
163 |   {
164 |    "cell_type": "code",
165 |    "execution_count": null,
166 |    "metadata": {},
167 |    "outputs": [],
168 |    "source": [
169 |     "fig = gcag_res.plot_components(legend_loc='lower right', figsize=(15, 9));"
170 |    ]
171 |   },
172 |   {
173 |    "cell_type": "markdown",
174 |    "metadata": {},
175 |    "source": [
176 |     "## Plotting predictions"
177 |    ]
178 |   },
179 |   {
180 |    "cell_type": "code",
181 |    "execution_count": null,
182 |    "metadata": {},
183 |    "outputs": [],
184 |    "source": [
185 |     "# Perform rolling prediction and multistep forecast\n",
186 |     "num_steps = 20\n",
187 |     "predict_res = gcag_res.get_prediction(dynamic=train['GCAG'].shape[0] - num_steps)\n",
188 |     "\n",
189 |     "predict = predict_res.predicted_mean\n",
190 |     "ci = predict_res.conf_int()"
191 |    ]
192 |   },
193 |   {
194 |    "cell_type": "code",
195 |    "execution_count": null,
196 |    "metadata": {},
197 |    "outputs": [],
198 |    "source": [
199 |     "plt.plot(predict)"
200 |    ]
201 |   },
202 |   {
203 |    "cell_type": "code",
204 |    "execution_count": null,
205 |    "metadata": {},
206 |    "outputs": [],
207 |    "source": [
208 |     "plt.scatter(train['GCAG'], predict)"
209 |    ]
210 |   },
211 |   {
212 |    "cell_type": "code",
213 |    "execution_count": null,
214 |    "metadata": {},
215 |    "outputs": [],
216 |    "source": [
217 |     "fig, ax = plt.subplots()\n",
218 |     "# Plot the results\n",
219 |     "ax.plot(train['GCAG'], 'k.', label='Observations');\n",
220 |     "ax.plot(train.index[:-num_steps], predict[:-num_steps], label='One-step-ahead Prediction');\n",
221 |     "\n",
222 |     "ax.plot(train.index[-num_steps:], predict[-num_steps:], 'r', label='Multistep Prediction');\n",
223 |     "ax.plot(train.index[-num_steps:], ci.iloc[-num_steps:], 'k--');\n",
224 |     "\n",
225 |     "# Cleanup the image\n",
226 |     "legend = ax.legend(loc='upper left');"
227 |    ]
228 |   },
229 |   {
230 |    "cell_type": "code",
231 |    "execution_count": null,
232 |    "metadata": {},
233 |    "outputs": [],
234 |    "source": [
235 |     "fig, ax = plt.subplots()\n",
236 |     "# Plot the results\n",
237 |     "ax.plot(train.index[-40:], train['GCAG'][-40:], 'k.', label='Observations');\n",
238 |     "ax.plot(train.index[-40:-num_steps], predict[-40:-num_steps], label='One-step-ahead Prediction');\n",
239 |     "\n",
240 |     "ax.plot(train.index[-num_steps:], predict[-num_steps:], 'r', label='Multistep Prediction');\n",
241 |     "ax.plot(train.index[-num_steps:], ci.iloc[-num_steps:], 'k--');\n",
242 |     "\n",
243 |     "# Cleanup the image\n",
244 |     "legend = ax.legend(loc='upper left');"
245 |    ]
246 |   },
247 |   {
248 |    "cell_type": "markdown",
249 |    "metadata": {},
250 |    "source": [
251 |     "## Exercise: consider adding a seasonal term for 12 periods for the model fit above. Does this improve the fit of the model?"
252 |    ]
253 |   },
254 |   {
255 |    "cell_type": "code",
256 |    "execution_count": null,
257 |    "metadata": {},
258 |    "outputs": [],
259 |    "source": []
260 |   },
261 |   {
262 |    "cell_type": "markdown",
263 |    "metadata": {},
264 |    "source": [
265 |     "## How does this compare to the original model?"
266 |    ]
267 |   },
268 |   {
269 |    "cell_type": "code",
270 |    "execution_count": null,
271 |    "metadata": {},
272 |    "outputs": [],
273 |    "source": []
274 |   },
275 |   {
276 |    "cell_type": "markdown",
277 |    "metadata": {},
278 |    "source": [
279 |     "## Let's explore the seasonality more"
280 |    ]
281 |   },
282 |   {
283 |    "cell_type": "code",
284 |    "execution_count": null,
285 |    "metadata": {},
286 |    "outputs": [],
287 |    "source": [
288 |     "seasonal_model = {\n",
289 |     "    'level': 'local level',\n",
290 |     "    'seasonal': 12\n",
291 |     "}\n",
292 |     "llmod = sm.tsa.UnobservedComponents(train['GCAG'], **seasonal_model)\n",
293 |     "ll_level_res = llmod.fit(method='powell', disp=False)"
294 |    ]
295 |   },
296 |   {
297 |    "cell_type": "code",
298 |    "execution_count": null,
299 |    "metadata": {},
300 |    "outputs": [],
301 |    "source": [
302 |     "fig = ll_level_res.plot_components(legend_loc='lower right', figsize=(15, 9));"
303 |    ]
304 |   },
305 |   {
306 |    "cell_type": "code",
307 |    "execution_count": null,
308 |    "metadata": {},
309 |    "outputs": [],
310 |    "source": [
311 |     "np.mean(np.abs(ll_level_res.predict() - train['GCAG']))"
312 |    ]
313 |   },
314 |   {
315 |    "cell_type": "code",
316 |    "execution_count": null,
317 |    "metadata": {},
318 |    "outputs": [],
319 |    "source": [
320 |     "train[:48].GCAG.plot()"
321 |    ]
322 |   },
323 |   {
324 |    "cell_type": "markdown",
325 |    "metadata": {},
326 |    "source": [
327 |     "## Exercise: a common null model for time series is to predict the value at time t-1 for the value at time t. How does such a model compare to the models we fit here?"
328 |    ]
329 |   },
330 |   {
331 |    "cell_type": "markdown",
332 |    "metadata": {},
333 |    "source": [
334 |     "### Consider correlation"
335 |    ]
336 |   },
337 |   {
338 |    "cell_type": "code",
339 |    "execution_count": null,
340 |    "metadata": {},
341 |    "outputs": [],
342 |    "source": []
343 |   },
344 |   {
345 |    "cell_type": "markdown",
346 |    "metadata": {},
347 |    "source": [
348 |     "### What about mean absolute error?"
349 |    ]
350 |   },
351 |   {
352 |    "cell_type": "code",
353 |    "execution_count": null,
354 |    "metadata": {},
355 |    "outputs": [],
356 |    "source": []
357 |   }
358 |  ],
359 |  "metadata": {
360 |   "kernelspec": {
361 |    "display_name": "Python 3",
362 |    "language": "python",
363 |    "name": "python3"
364 |   },
365 |   "language_info": {
366 |    "codemirror_mode": {
367 |     "name": "ipython",
368 |     "version": 3
369 |    },
370 |    "file_extension": ".py",
371 |    "mimetype": "text/x-python",
372 |    "name": "python",
373 |    "nbconvert_exporter": "python",
374 |    "pygments_lexer": "ipython3",
375 |    "version": "3.6.8"
376 |   }
377 |  },
378 |  "nbformat": 4,
379 |  "nbformat_minor": 2
380 | }
381 | 


--------------------------------------------------------------------------------
/modern_time_series_analysis/ModernTimeSeriesAnalysis/StateSpaceModels/Nile.csv:
--------------------------------------------------------------------------------
  1 | "","year","val"
  2 | "1",1871,1120
  3 | "2",1872,1160
  4 | "3",1873,963
  5 | "4",1874,1210
  6 | "5",1875,1160
  7 | "6",1876,1160
  8 | "7",1877,813
  9 | "8",1878,1230
 10 | "9",1879,1370
 11 | "10",1880,1140
 12 | "11",1881,995
 13 | "12",1882,935
 14 | "13",1883,1110
 15 | "14",1884,994
 16 | "15",1885,1020
 17 | "16",1886,960
 18 | "17",1887,1180
 19 | "18",1888,799
 20 | "19",1889,958
 21 | "20",1890,1140
 22 | "21",1891,1100
 23 | "22",1892,1210
 24 | "23",1893,1150
 25 | "24",1894,1250
 26 | "25",1895,1260
 27 | "26",1896,1220
 28 | "27",1897,1030
 29 | "28",1898,1100
 30 | "29",1899,774
 31 | "30",1900,840
 32 | "31",1901,874
 33 | "32",1902,694
 34 | "33",1903,940
 35 | "34",1904,833
 36 | "35",1905,701
 37 | "36",1906,916
 38 | "37",1907,692
 39 | "38",1908,1020
 40 | "39",1909,1050
 41 | "40",1910,969
 42 | "41",1911,831
 43 | "42",1912,726
 44 | "43",1913,456
 45 | "44",1914,824
 46 | "45",1915,702
 47 | "46",1916,1120
 48 | "47",1917,1100
 49 | "48",1918,832
 50 | "49",1919,764
 51 | "50",1920,821
 52 | "51",1921,768
 53 | "52",1922,845
 54 | "53",1923,864
 55 | "54",1924,862
 56 | "55",1925,698
 57 | "56",1926,845
 58 | "57",1927,744
 59 | "58",1928,796
 60 | "59",1929,1040
 61 | "60",1930,759
 62 | "61",1931,781
 63 | "62",1932,865
 64 | "63",1933,845
 65 | "64",1934,944
 66 | "65",1935,984
 67 | "66",1936,897
 68 | "67",1937,822
 69 | "68",1938,1010
 70 | "69",1939,771
 71 | "70",1940,676
 72 | "71",1941,649
 73 | "72",1942,846
 74 | "73",1943,812
 75 | "74",1944,742
 76 | "75",1945,801
 77 | "76",1946,1040
 78 | "77",1947,860
 79 | "78",1948,874
 80 | "79",1949,848
 81 | "80",1950,890
 82 | "81",1951,744
 83 | "82",1952,749
 84 | "83",1953,838
 85 | "84",1954,1050
 86 | "85",1955,918
 87 | "86",1956,986
 88 | "87",1957,797
 89 | "88",1958,923
 90 | "89",1959,975
 91 | "90",1960,815
 92 | "91",1961,1020
 93 | "92",1962,906
 94 | "93",1963,901
 95 | "94",1964,1170
 96 | "95",1965,912
 97 | "96",1966,746
 98 | "97",1967,919
 99 | "98",1968,718
100 | "99",1969,714
101 | "100",1970,740
102 | 


--------------------------------------------------------------------------------
/modern_time_series_analysis/README.md:
--------------------------------------------------------------------------------
 1 | # Modern Time Series Analysis
 2 | 
 3 | ## Install and Setup
 4 | 
 5 | Okay, I wanted to streamline this.  So you can just use this repo:
 6 | 
 7 |     pip install -r requirements.txt
 8 | 
 9 | OR, since this entire tutorial is run from Jupyter notebooks, you probably want to install your requirements there:
10 | 
11 | ```shell
12 | $ which jupyter
13 | /home/my_user_name/stuff/bin/jupyter
14 | $ /home/my_user_name/stuff/bin/pip install -r requirements.txt
15 | 
16 | $ cd /full/path/to/modern_time_series_analysis/
17 | $ jupyter notebook
18 | ```
19 | 
20 | ## Syllabus
21 | 
22 | And all the data and notebooks you need will be [here](https://github.com/theJollySin/scipy_con_2019/tree/master/modern_time_series_analysis/ModernTimeSeriesAnalysis):
23 | 
24 | 1. [Structural Time Series](ModernTimeSeriesAnalysis/StateSpaceModels/1_Structural_Time_Series_INSTRUCTOR.ipynb )
25 | 2. [Gaussian HMM](ModernTimeSeriesAnalysis/StateSpaceModels/2_Gaussian_HMM_INSTRUCTOR.ipynb)
26 | 3. [Trees for Classification and Prediction](ModernTimeSeriesAnalysis/MachineLearning/3_Trees_for_Classification_and_Prediction_INSTRUCTOR.ipynb)
27 | 4. [Clustering](ModernTimeSeriesAnalysis/MachineLearning/4_Clustering_INSTRUCTOR.ipynb)
28 | 5. [Forecasting electricity use with mxnet](ModernTimeSeriesAnalysis/DeepLearning/Electricity/5_Forecasting_electric_use_with_mxnet_INSTRUCTOR.ipynb)
29 | 6. [Stocks](ModernTimeSeriesAnalysis/DeepLearning/Stocks/6_Stocks_INSTRUCTOR.ipynb)
30 | 
31 | 
32 | ## Notes
33 | 
34 | Don't expect anything exciting here.  These are literally just my notes
35 | 
36 | 
37 | ### Structural Time Series
38 | 
39 | * [ARIMA](https://en.wikipedia.org/wiki/Autoregressive_integrated_moving_average) models are old school, but if your machine learning toys don't do better than this, why bother?
40 | * [Kalman Filtering!](https://en.wikipedia.org/wiki/Kalman_filter) - Love me a Kalman Filter
41 | 
42 | 
43 | ### Machine Learning - Time Series Trees for Classification and Prediction
44 | 
45 | * Most machine learning on time series uses tools that weren't designed for time series data. We always have to use other tools and "make 'em work".
46 | * Doctors and nurses to "feature detection" on "time series data" when they look at heart rates from ECGs. Everything old is new again.
47 | * We're going to cover [Random Forests](https://en.wikipedia.org/wiki/Random_forest) and [Gradient Boosted Trees](https://en.wikipedia.org/wiki/Gradient_boosting#Gradient_tree_boosting) with [xgboost](https://xgboost.readthedocs.io/en/latest/).
48 | * So we're going to look at different time series data sets and figure out which ones look the most similar. So, if we had little snippets of time series data, the human eye might be able to find similar sets. But feature detection by human labor is costly and slow. ...It'll probably still be computationally intensive.
49 | * [Cesium](https://github.com/cesium-ml/cesium) is a feature generation library - Mostly just for initial exploration.
50 | * Are we chopping up our time series into blocks because it is cyclical in nature? Or would we do that anyway?
51 |   * We used a sliding window, we DID NOT chop it into blocks.
52 | * We used a machine learning method on a problem that had very little data. This was not a good choice.
53 | * The important lessons here were supposed to be:
54 |   1. Chop up your time series data to get more data.
55 |   2. If you don't have enough data, machine learning is the wrong approach.
56 | 
57 | 
58 | ### Machine Learning - Clustering
59 | 
60 | * [Dynamic Time Warping](https://en.wikipedia.org/wiki/Dynamic_time_warping)
61 | * [Here](https://github.com/wannesm/dtaidistance) is one little library for DTW
62 | 
63 | 
64 | ### Deep Learning - Electric Use
65 | 
66 | * Typically if you want to put time series data into a neural network, you use [RNN](https://en.wikipedia.org/wiki/Recurrent_neural_network).
67 | * Research later: [GRU vs LSTM](https://datascience.stackexchange.com/questions/14581/when-to-use-gru-over-lstm)
68 | * You can also use a [CNN](https://en.wikipedia.org/wiki/Convolutional_neural_network)
69 | * CNNs might be a little better for classification than prediction, compared to RNN
70 | * This example finally uses parallel time series signals to be processed together; a *MUCH* more interesting problem.
71 | 
72 | 
73 | ### Deep Learning - Stocks
74 | 
75 | * Do this on your own later.
76 | 
77 | 


--------------------------------------------------------------------------------
/modern_time_series_analysis/SciPyModernTimeSeries.pdf:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/john-science/scipy_con_2019/7280bc1949f90b151048c0ed127bdd656064c2cb/modern_time_series_analysis/SciPyModernTimeSeries.pdf


--------------------------------------------------------------------------------
/modern_time_series_analysis/requirements.txt:
--------------------------------------------------------------------------------
 1 | absl-py==0.7.1
 2 | astor==0.8.0
 3 | cesium==0.9.9
 4 | certifi==2024.7.4
 5 | chardet==3.0.4
 6 | hmmlearn==0.2.2
 7 | gast==0.2.2
 8 | google-pasta==0.1.7
 9 | graphviz==0.8.4
10 | grpcio==1.53.2
11 | h5py==2.9.0
12 | idna==3.7
13 | joblib>=1.2.0
14 | Keras-Applications==1.0.8
15 | Keras-Preprocessing==1.1.0
16 | Markdown==3.1.1
17 | mxnet==1.9.1
18 | numpy==1.22.0
19 | pandas==0.24.2
20 | patsy==0.5.1
21 | protobuf==3.18.3
22 | python-dateutil==2.8.0
23 | pytz==2019.1
24 | requests==2.32.0
25 | scikit-learn==1.5.0
26 | scipy==1.10.0
27 | six==1.12.0
28 | sklearn==0.0
29 | statsmodels==0.10.0
30 | tensorboard==1.14.0
31 | tensorflow==2.12.1
32 | tensorflow-estimator==1.14.0
33 | termcolor==1.1.0
34 | urllib3==1.26.19
35 | Werkzeug==3.0.6
36 | wrapt==1.11.2
37 | xgboost==0.90
38 | 


--------------------------------------------------------------------------------