2d29f76 | Sam | 19 December 2018, 09:45:26 UTC | Update README.md | 19 December 2018, 09:45:26 UTC |
7dcc897 | Sam | 19 December 2018, 09:36:36 UTC | Temporary fix on the predictN bug. Currently the predict() function will call a deepcopy to the dlm class itself and the copy is stored as a member variable with the dlm object. As a result, when the predict() is called multiple times, the dlm object is copied recursively and the memory/complxity is exploded in the end. The temporary solution is to void the _predictModel variable before making the deepcopy. The current predict() complxity is O(n). TESTED=""" >>> from pydlm import dlm,trend >>> dlm1 = dlm(range(1000)) + trend(1) >>> dlm1.fit() >>> for i in range(1000): >>> ... dlm1.predict() """ | 19 December 2018, 09:36:36 UTC |
e0287d4 | Sam | 14 December 2018, 10:22:54 UTC | Use matrix config to support python 3.7 | 14 December 2018, 10:22:54 UTC |
b651932 | Sam | 14 December 2018, 10:13:20 UTC | Set zip_safe=False | 14 December 2018, 10:13:20 UTC |
75bb84b | Sam | 14 December 2018, 09:58:43 UTC | Give 3.7 a try | 14 December 2018, 09:58:43 UTC |
6b7d678 | Sam | 14 December 2018, 09:40:30 UTC | Give 3.7 a try. | 14 December 2018, 09:40:30 UTC |
c996663 | David Tolpin | 14 December 2018, 09:39:35 UTC | unittest is part of standard python 2 and 3 and cannot be installed from pypi (#30) | 14 December 2018, 09:39:35 UTC |
1d31a60 | Sam | 14 December 2018, 09:30:52 UTC | Drop 3.7 to first pass the tests. Might need to remove tests requirement? | 14 December 2018, 09:30:52 UTC |
d0ae158 | Sam | 14 December 2018, 09:13:21 UTC | Drop python 3.4 as well... | 14 December 2018, 09:13:21 UTC |
a21aa15 | Sam | 14 December 2018, 09:12:27 UTC | Dropping python 2.7 in the test. Adding python 3.7. | 14 December 2018, 09:12:27 UTC |
71a52f0 | Sam | 14 December 2018, 08:50:35 UTC | Separate _dlm.py into multiple modules each with detailed functionality. Easier to read and maintain. | 14 December 2018, 08:50:35 UTC |
35561cf | Sam | 03 July 2018, 11:06:13 UTC | Update README.md | 03 July 2018, 11:06:13 UTC |
3a52d06 | Sam | 03 July 2018, 10:34:15 UTC | Update README.md | 03 July 2018, 10:34:15 UTC |
a2ea2ea | Sam | 18 May 2018, 07:40:03 UTC | Update README.md | 18 May 2018, 07:40:03 UTC |
aa039f8 | David Tolpin | 18 May 2018, 07:25:00 UTC | passing printInfo to builder so that output does not pollute the terminal when switched off (#27) | 18 May 2018, 07:25:00 UTC |
d35f8bb | Sam | 25 April 2018, 07:09:08 UTC | Add functions to build dlm from exported model. | 25 April 2018, 07:09:08 UTC |
b9aa643 | Sam | 24 April 2018, 08:46:17 UTC | Remove the _clean() function as now predictN won't modify the model object and _clean() is useless. | 24 April 2018, 08:46:17 UTC |
6e5aa6f | Sam | 25 March 2018, 23:35:38 UTC | Add a method to allow builder to be initialized from an exported builder. | 25 March 2018, 23:35:38 UTC |
fb8a53d | Sam | 25 March 2018, 23:11:32 UTC | Add function to export dlm builder. We only support exporting builder for models without dynamic components. | 25 March 2018, 23:11:32 UTC |
67e7bf5 | Sam | 25 February 2018, 09:57:33 UTC | upadte comments on autoReg#2 | 25 February 2018, 09:57:33 UTC |
868d6ee | Sam | 25 February 2018, 09:49:18 UTC | Add link to auto regressor | 25 February 2018, 09:49:18 UTC |
5f660a2 | Sam | 25 February 2018, 09:46:13 UTC | Update date comment on auto regressor component. | 25 February 2018, 09:46:13 UTC |
4ea2755 | Sam | 25 February 2018, 09:29:17 UTC | Merge branch 'master' of github.com:wwrechard/pydlm | 25 February 2018, 09:29:17 UTC |
aac8682 | Sam | 25 February 2018, 09:28:56 UTC | Fixing python 3 range issue. (Hopefully) | 25 February 2018, 09:28:56 UTC |
9da2e47 | Sam | 25 February 2018, 09:19:46 UTC | Update README.md | 25 February 2018, 09:19:46 UTC |
31f7369 | Sam | 25 February 2018, 09:03:45 UTC | remove all the tests that depend on the data argument in autoReg. | 25 February 2018, 09:03:45 UTC |
2be6c03 | Sam | 25 February 2018, 08:57:33 UTC | Refactor auto regressor to fetch data directly from dlm main class instead of user input. This helps simplifying the predict function. It does not need to manually construct the auto regressor features when predict into the future. With the new auto regressor structure, predict function just need to send the reference to the raw data and the predicted data and autoReg will construct the feature vector itself. | 25 February 2018, 08:57:33 UTC |
754735e | Sam | 14 February 2018, 08:26:35 UTC | Update README.md | 14 February 2018, 08:26:35 UTC |
9c828ba | Sam | 11 February 2018, 09:35:38 UTC | Update setup.py | 11 February 2018, 09:35:38 UTC |
c9f7736 | Sam | 11 February 2018, 09:35:21 UTC | Update README.md | 11 February 2018, 09:35:21 UTC |
83b8569 | Sam | 11 February 2018, 09:34:53 UTC | Update CHANGELOG.md | 11 February 2018, 09:34:53 UTC |
78e8a21 | Sam | 11 February 2018, 09:33:53 UTC | Update README.md | 11 February 2018, 09:33:53 UTC |
947261d | Sam | 11 February 2018, 09:29:04 UTC | Update the version number. | 11 February 2018, 09:29:04 UTC |
c073dde | Sam | 11 February 2018, 09:23:14 UTC | Merge branch 'master' of github.com:wwrechard/pydlm | 11 February 2018, 09:23:14 UTC |
d68ecaf | Sam | 11 February 2018, 09:22:52 UTC | Fix the overloading problem for noisePrior(). | 11 February 2018, 09:22:52 UTC |
d134357 | DeliciousHair | 28 November 2017, 07:08:01 UTC | Fix typo 'filtering' and add variable options in dlm instantiation. * modified: pydlm/dlm.py fixed annoying typo fitering --> filtering * modified: pydlm/dlm.py modified: pydlm/func/_dlm.py * modified: pydlm/func/_dlm.py * modified: pydlm/func/_dlm.py | 28 November 2017, 07:08:01 UTC |
e9e6c69 | Sam | 28 November 2017, 05:26:56 UTC | Update README.md | 28 November 2017, 05:26:56 UTC |
86c1a53 | Sam | 28 November 2017, 05:15:56 UTC | Change the dlm to use copied model for prediction, so that the status of main model won't be affected. | 28 November 2017, 05:15:56 UTC |
f479ff5 | Sam | 15 November 2017, 09:02:36 UTC | Change the auto noise initializer to take care of the missing data. | 15 November 2017, 09:02:36 UTC |
a06975d | Sam | 16 October 2017, 10:07:21 UTC | Update README.md | 16 October 2017, 10:07:21 UTC |
ed0bb21 | Sam | 16 October 2017, 09:52:58 UTC | Add an autoNoise initialization to the model. | 16 October 2017, 09:52:58 UTC |
60a5a31 | Sam | 16 October 2017, 09:28:33 UTC | Change the default prior variance to 100. | 16 October 2017, 09:28:33 UTC |
bd88a2b | Sam | 28 September 2017, 07:35:09 UTC | Update README.md | 28 September 2017, 07:35:09 UTC |
796c3d7 | Sam | 28 September 2017, 07:31:26 UTC | Update README.md | 28 September 2017, 07:31:26 UTC |
a61401f | Sam | 28 September 2017, 07:26:35 UTC | Clean the log info that is used for debugging | 28 September 2017, 07:26:35 UTC |
0c71c14 | Sam | 28 September 2017, 07:24:30 UTC | Bug fix on continuePrediction (currentDate ---> startDate) to make prediction work for autoReg. | 28 September 2017, 07:24:30 UTC |
813db24 | Sam | 28 September 2017, 06:41:32 UTC | Merge branch 'master' of github.com:wwrechard/pydlm | 28 September 2017, 06:41:32 UTC |
d5e88c1 | Sam | 28 September 2017, 06:40:59 UTC | Fix a bug that continuePredict fails when AutoReg has degree > 1 | 28 September 2017, 06:40:59 UTC |
c768d86 | Sam | 27 September 2017, 08:40:06 UTC | Update README.md | 27 September 2017, 08:40:06 UTC |
0cb3f42 | Sam | 22 September 2017, 09:27:09 UTC | confirm the tests have covered the tune. Add list() wrapper to map() for fixing python 3 error. | 22 September 2017, 09:27:09 UTC |
cf24f3a | Sam | 22 September 2017, 09:20:49 UTC | Merge branch 'master' of github.com:wwrechard/pydlm | 22 September 2017, 09:20:49 UTC |
9971846 | Sam | 22 September 2017, 09:19:12 UTC | Add tests for runing plot() and tune() to confirm the coverage. (Expect to see travis failure for python3). | 22 September 2017, 09:19:12 UTC |
46a0ce6 | Sam | 30 August 2017, 02:10:34 UTC | Update README.md | 30 August 2017, 02:10:34 UTC |
cf7e988 | Sam | 30 August 2017, 02:10:22 UTC | Update README.md | 30 August 2017, 02:10:22 UTC |
5aca832 | Sam | 30 August 2017, 02:09:25 UTC | Update README | 30 August 2017, 02:09:25 UTC |
e0d3aa4 | Sam | 30 August 2017, 02:06:52 UTC | Released 0.1.1.9 and added to Changelogs. | 30 August 2017, 02:06:52 UTC |
995e64a | Sam | 04 August 2017, 07:21:31 UTC | Update README.md | 04 August 2017, 07:21:31 UTC |
93f93ba | Sam | 31 July 2017, 05:23:09 UTC | correct document. | 31 July 2017, 05:23:09 UTC |
74a6aed | Sam | 31 July 2017, 00:46:20 UTC | Merge branch 'master' of github.com:wwrechard/pydlm | 31 July 2017, 00:46:20 UTC |
c605299 | Sam | 31 July 2017, 00:46:06 UTC | Add the example to the user manual. | 31 July 2017, 00:46:06 UTC |
67fd0c0 | Sam | 28 July 2017, 01:39:08 UTC | Update README.md | 28 July 2017, 01:39:08 UTC |
cc47378 | Sam | 27 July 2017, 10:12:24 UTC | Merge branch 'master' of github.com:wwrechard/pydlm | 27 July 2017, 10:12:24 UTC |
a46a045 | Sam | 27 July 2017, 10:11:51 UTC | Complete the description of example code. | 27 July 2017, 10:11:51 UTC |
c5ba823 | Sam | 27 July 2017, 10:03:13 UTC | Update README.md | 27 July 2017, 10:03:13 UTC |
3085c31 | Sam | 27 July 2017, 09:59:48 UTC | Update README.md | 27 July 2017, 09:59:48 UTC |
b2f35ce | Sam | 27 July 2017, 09:56:47 UTC | Update README.md | 27 July 2017, 09:56:47 UTC |
9cb47f0 | Sam | 27 July 2017, 09:52:36 UTC | Update README.md | 27 July 2017, 09:52:36 UTC |
72e80ff | Sam | 27 July 2017, 09:49:51 UTC | Update README.md | 27 July 2017, 09:49:51 UTC |
be4cf2e | Sam | 27 July 2017, 09:49:20 UTC | Update README.md | 27 July 2017, 09:49:20 UTC |
9848c62 | Sam | 27 July 2017, 09:47:57 UTC | Update README.md | 27 July 2017, 09:47:57 UTC |
fdac95e | Sam | 27 July 2017, 09:44:29 UTC | Update README.md | 27 July 2017, 09:44:29 UTC |
4064c76 | Sam | 27 July 2017, 09:43:14 UTC | remove changelog to CHANGELOG.md | 27 July 2017, 09:43:14 UTC |
11635ce | Sam | 27 July 2017, 09:41:52 UTC | remove changelog~ | 27 July 2017, 09:41:52 UTC |
cfbdf47 | Sam | 27 July 2017, 09:40:42 UTC | Add change log. | 27 July 2017, 09:40:42 UTC |
e665936 | Sam | 27 July 2017, 09:40:02 UTC | Update README.md | 27 July 2017, 09:40:02 UTC |
9e87916 | Sam | 27 July 2017, 09:27:48 UTC | Update README.md | 27 July 2017, 09:27:48 UTC |
98dc0be | Sam | 27 July 2017, 08:44:18 UTC | Add more images for the google science example. | 27 July 2017, 08:44:18 UTC |
463d7eb | Sam | 27 July 2017, 07:45:43 UTC | Add images for Google data science example | 27 July 2017, 07:45:43 UTC |
bb7c41e | Sam | 26 July 2017, 09:52:08 UTC | Update README.md | 26 July 2017, 09:52:08 UTC |
eeb44bc | Sam | 26 July 2017, 06:39:40 UTC | Complete description | 26 July 2017, 06:39:40 UTC |
c2d7c54 | Sam | 26 July 2017, 06:37:07 UTC | Complete the example using Google data post. | 26 July 2017, 06:37:07 UTC |
aeed16b | Sam | 25 July 2017, 10:11:04 UTC | Fix bug in testDlm. The predictN now output python list of float instead of list of numpy matrix. | 25 July 2017, 10:11:04 UTC |
e0b93fc | Sam | 25 July 2017, 10:09:04 UTC | reduce the default prior variance from 1e7 to 100 for seasonality component. | 25 July 2017, 10:09:04 UTC |
87e1e23 | Sam | 25 July 2017, 09:36:19 UTC | Add description to dlm.plotPredictN | 25 July 2017, 09:36:19 UTC |
1e0d52b | Sam | 25 July 2017, 09:31:30 UTC | Add a plot function to directly plot N-day prediction with the same input as predictN() | 25 July 2017, 09:31:30 UTC |
8530a6a | Sam | 25 July 2017, 08:47:00 UTC | Ingore .csv~ tmp files. | 25 July 2017, 08:47:00 UTC |
2784479 | Sam | 25 July 2017, 08:45:10 UTC | Add example using Google data science blog post. | 25 July 2017, 08:45:10 UTC |
9762858 | Sam | 20 June 2017, 07:39:08 UTC | Update README.md | 20 June 2017, 07:39:08 UTC |
7347112 | Sam | 20 June 2017, 07:31:24 UTC | Update README.md | 20 June 2017, 07:31:24 UTC |
cd1f81a | Sam | 20 June 2017, 07:18:47 UTC | Merge branch 'master' of github.com:wwrechard/pydlm | 20 June 2017, 07:18:47 UTC |
22de129 | Sam | 20 June 2017, 06:57:29 UTC | The .getMSE() method does not take care of missing data. Fix the bug when None exists in the data. | 20 June 2017, 06:57:29 UTC |
bf3e648 | Sam | 19 June 2017, 01:25:13 UTC | Create LICENSE (#12) | 19 June 2017, 01:25:13 UTC |
77eba88 | Sam | 15 June 2017, 07:57:43 UTC | update the class description. | 15 June 2017, 07:57:43 UTC |
f184105 | Sam | 15 June 2017, 07:52:21 UTC | Update the trend degree to match the actual degree of a polynomial, i.e., degree=0 indicats a constant, degree=1 indicates a linear trend and so on. | 15 June 2017, 07:52:21 UTC |
2cc1f20 | Sam | 15 June 2017, 06:53:08 UTC | Create README.md | 15 June 2017, 06:53:08 UTC |
c840843 | Sam | 15 June 2017, 06:51:46 UTC | Temporarily disable python 3.7 check due to numpy Bug issue: https://github.com/python/cpython/pull/1236 | 15 June 2017, 06:51:46 UTC |
128441b | Sam | 15 June 2017, 05:03:55 UTC | Update README.md | 15 June 2017, 05:03:55 UTC |
910de99 | Sam | 17 May 2017, 10:47:41 UTC | Fix indent | 17 May 2017, 10:47:41 UTC |
e4cf329 | Sam | 17 May 2017, 10:46:09 UTC | Merge branch 'master' of github.com:wwrechard/pydlm | 17 May 2017, 10:46:09 UTC |
9040ecc | Sam | 17 May 2017, 10:45:41 UTC | Update README.md | 17 May 2017, 10:45:41 UTC |