cf5d2ad | Antoine CARME | 11 May 2022, 11:59:50 UTC | prepare 4.0-rc1 release. | 11 May 2022, 11:59:50 UTC |
aeec0c1 | CARME Antoine | 09 May 2022, 11:55:58 UTC | Update TODO.md | 09 May 2022, 11:55:58 UTC |
aa073f0 | CARME Antoine | 09 May 2022, 11:53:26 UTC | Update TODO.md | 09 May 2022, 11:53:26 UTC |
bf4b8c6 | Antoine CARME | 09 May 2022, 11:40:24 UTC | Updated this file | 09 May 2022, 11:40:24 UTC |
aba4220 | Antoine CARME | 09 May 2022, 11:25:20 UTC | Renamed this file | 09 May 2022, 11:25:20 UTC |
9f252c4 | Antoine CARME | 09 May 2022, 11:24:25 UTC | Renamed this file | 09 May 2022, 11:24:25 UTC |
12b0a76 | CARME Antoine | 09 May 2022, 11:21:57 UTC | Update TODO | 09 May 2022, 11:21:57 UTC |
22b2c18 | Antoine CARME | 09 May 2022, 09:56:01 UTC | Document the Algorithmic Details of PyAF #35. Added more detail about signal transformation (min-max scaling). | 09 May 2022, 09:56:01 UTC |
c282cdb | Antoine CARME | 09 May 2022, 09:17:25 UTC | Document the Algorithmic Details of PyAF #35. Added more detail about seasonal components and AR feature selection. | 09 May 2022, 09:17:25 UTC |
0d44c09 | Antoine CARME | 09 May 2022, 09:04:13 UTC | Document the Algorithmic Details of PyAF #35. Added more detail about seasonal components and AR feature selection. | 09 May 2022, 09:04:13 UTC |
a2b3240 | Antoine CARME | 09 May 2022, 08:04:09 UTC | Document the Algorithmic Details of PyAF #35. Added more detail about BoxCox Transformation, MovingMedian and MovingAverage Trends | 09 May 2022, 08:04:09 UTC |
bcb880b | Antoine CARME | 05 May 2022, 13:10:40 UTC | Document the Algorithmic Details of PyAF #35. Added some notes about PyAF explainability and forecast outputs. | 05 May 2022, 13:10:40 UTC |
2e7c5c2 | Antoine CARME | 05 May 2022, 13:00:29 UTC | Document the Algorithmic Details of PyAF #35. Added some notes about PyAF explainability and forecast outputs. | 05 May 2022, 13:00:29 UTC |
0755445 | Antoine Carme | 04 May 2022, 16:43:41 UTC | Document the Algorithmic Details of PyAF #35. Added some notes about PyAF reproducibility. | 04 May 2022, 16:43:41 UTC |
5f586e1 | Antoine Carme | 02 May 2022, 08:44:24 UTC | Added a missing MASE reference | 02 May 2022, 08:44:24 UTC |
2537dd2 | Antoine Carme | 02 May 2022, 08:43:31 UTC | Document the Algorithmic Details of PyAF #35. Updated this document (WIP) | 02 May 2022, 08:43:31 UTC |
632a64e | Antoine Carme | 02 May 2022, 08:29:47 UTC | Document the Algorithmic Details of PyAF #35. Updated this document (WIP) | 02 May 2022, 08:29:47 UTC |
a2961e0 | Antoine Carme | 28 April 2022, 20:09:40 UTC | Updated these logs | 28 April 2022, 20:09:40 UTC |
a6e18d0 | Antoine Carme | 28 April 2022, 19:46:36 UTC | Made these tests lighter. slow mode is too heavy. | 28 April 2022, 19:46:36 UTC |
dff19ee | Antoine Carme | 28 April 2022, 12:05:32 UTC | Updated these logs | 28 April 2022, 12:05:32 UTC |
83190bb | Antoine Carme | 28 April 2022, 11:00:08 UTC | Document the Algorithmic Details of PyAF #35. Updated this document (WIP) | 28 April 2022, 11:00:08 UTC |
4e05000 | Antoine Carme | 28 April 2022, 10:47:11 UTC | Document the Algorithmic Details of PyAF #35. Updated this document (WIP) | 28 April 2022, 10:47:11 UTC |
7753725 | Antoine Carme | 28 April 2022, 10:38:04 UTC | Document the Algorithmic Details of PyAF #35. Updated this document (WIP) | 28 April 2022, 10:38:04 UTC |
9fe1b49 | Antoine Carme | 28 April 2022, 10:21:45 UTC | Document the Algorithmic Details of PyAF #35. Updated this document (WIP) | 28 April 2022, 10:21:45 UTC |
b614bfe | Antoine Carme | 28 April 2022, 09:47:25 UTC | Use more jobs in circleci | 28 April 2022, 09:47:25 UTC |
defe798 | Antoine Carme | 28 April 2022, 08:23:08 UTC | Investigate PyTorch-based LSTM and MLP models #199. Attempt to speedup pytorch. switch to float32 | 28 April 2022, 08:23:08 UTC |
967785c | Antoine Carme | 27 April 2022, 23:11:55 UTC | Removed some unnnecessary warning | 27 April 2022, 23:11:55 UTC |
9ac512e | Antoine Carme | 27 April 2022, 23:11:30 UTC | Updated build tests. | 27 April 2022, 23:11:30 UTC |
ded8960 | Antoine Carme | 27 April 2022, 23:07:54 UTC | Less blabla. Updated these logs. | 27 April 2022, 23:07:54 UTC |
4754a90 | Antoine Carme | 27 April 2022, 22:09:44 UTC | Reduce Blabla. Display at most 10 skipped models in debug mode. | 27 April 2022, 22:09:44 UTC |
7bbea33 | Antoine Carme | 27 April 2022, 22:09:11 UTC | Reduce Blabla. Display 10 best models perfs only in debug mode. | 27 April 2022, 22:09:11 UTC |
750cd4c | Antoine Carme | 27 April 2022, 19:46:08 UTC | Updated circle-ci | 27 April 2022, 19:46:08 UTC |
8477e55 | Antoine Carme | 27 April 2022, 19:03:52 UTC | Less Blabla. Avoid broadcasting. | 27 April 2022, 19:03:52 UTC |
5e89c71 | Antoine Carme | 27 April 2022, 19:03:07 UTC | Release Process : Profiling and Warning Hunts #195. Attempt to avoid a warning in pearsonr computations | 27 April 2022, 19:03:07 UTC |
254f40e | Antoine Carme | 27 April 2022, 19:01:54 UTC | Improved perfs display in debug mode. Use json format | 27 April 2022, 19:01:54 UTC |
59e8c79 | Antoine Carme | 27 April 2022, 17:04:20 UTC | Investigate PyTorch-based LSTM and MLP models #199. Updated circle-ci | 27 April 2022, 17:04:20 UTC |
f534f19 | Antoine Carme | 27 April 2022, 17:00:16 UTC | Investigate PyTorch-based LSTM and MLP models #199. Parallelized some jobs | 27 April 2022, 17:00:16 UTC |
5def266 | Antoine Carme | 27 April 2022, 15:57:46 UTC | Investigate PyTorch-based LSTM and MLP models #199. Keep tensorflow away from testsgit push | 27 April 2022, 15:57:46 UTC |
7526487 | Antoine Carme | 27 April 2022, 15:04:51 UTC | Investigate PyTorch-based LSTM and MLP models #199. Added package dependencies. | 27 April 2022, 15:04:51 UTC |
81df748 | Antoine Carme | 27 April 2022, 15:02:38 UTC | Investigate PyTorch-based LSTM and MLP models #199. Added package dependencies. | 27 April 2022, 15:02:38 UTC |
2c5fc3f | Antoine Carme | 26 April 2022, 17:21:38 UTC | Rounding effects (4 for perfs , 6 for model weights). Updatd these logs | 26 April 2022, 17:21:38 UTC |
ebeeab7 | Antoine Carme | 26 April 2022, 16:33:04 UTC | More rounding (4 for perfs , 6 for model weights). | 26 April 2022, 16:33:04 UTC |
9b442d6 | Antoine Carme | 26 April 2022, 16:32:40 UTC | Rounding effects (4 for perfs , 6 for model weights). Updatd these logs | 26 April 2022, 16:32:40 UTC |
27a3ed0 | Antoine Carme | 25 April 2022, 17:04:24 UTC | Rounding effects (4 for perfs , 6 for model weights). Updatd these logs | 25 April 2022, 17:04:24 UTC |
b896826 | Antoine Carme | 25 April 2022, 16:44:34 UTC | Store test logs as circle-ci artifacts | 25 April 2022, 16:44:34 UTC |
2847add | Antoine Carme | 25 April 2022, 16:07:07 UTC | Store test logs as circle-ci artifacts | 25 April 2022, 16:07:07 UTC |
18f0293 | Antoine Carme | 25 April 2022, 16:06:00 UTC | Store test logs as circle-ci artifacts | 25 April 2022, 16:06:00 UTC |
902cf6d | Antoine Carme | 25 April 2022, 16:04:26 UTC | Store test logs as circle-ci artifacts | 25 April 2022, 16:04:26 UTC |
c922829 | Antoine Carme | 25 April 2022, 14:49:44 UTC | Investigate PyTorch-based LSTM and MLP models #199. Enforce reproducibility of pytorch and keras models. Added more reporting about model architectures. | 25 April 2022, 14:49:44 UTC |
c745b53 | Antoine Carme | 24 April 2022, 18:14:29 UTC | Added a benchmark of keras and pytorch on a slow pyaf model with 45 threads. | 24 April 2022, 18:14:29 UTC |
9f7caf5 | Antoine Carme | 22 April 2022, 19:28:32 UTC | Added a workaround for an issue with pytorch multiprocessing. | 22 April 2022, 19:28:32 UTC |
a4e255f | Antoine Carme | 22 April 2022, 18:10:15 UTC | Removed an old comment | 22 April 2022, 18:10:15 UTC |
17cb2eb | Antoine Carme | 22 April 2022, 15:44:57 UTC | Round model data to 6 digits. | 22 April 2022, 15:44:57 UTC |
4b54362 | Antoine Carme | 22 April 2022, 15:44:35 UTC | Round performance criteria to 4 digits. | 22 April 2022, 15:44:35 UTC |
f4dd9d9 | Antoine Carme | 22 April 2022, 14:43:42 UTC | Use pytlint. Remove unnecessary imports. | 22 April 2022, 14:43:42 UTC |
da2f519 | Antoine Carme | 21 April 2022, 20:51:53 UTC | Investigate PyTorch-based LSTM and MLP models #199. Added EarlyStopping for PyTorch and Keras. Updatd models complexity. | 21 April 2022, 20:51:53 UTC |
b6b0e07 | Antoine Carme | 21 April 2022, 19:54:27 UTC | Investigate PyTorch-based LSTM and MLP models #199. Less blabla | 21 April 2022, 19:54:27 UTC |
a4c3ebf | Antoine Carme | 21 April 2022, 18:44:26 UTC | Investigate PyTorch-based LSTM and MLP models #199. Avoid importing pytorch and other external modules when the underlying models are not activated. | 21 April 2022, 18:44:26 UTC |
c877999 | Antoine Carme | 21 April 2022, 18:43:16 UTC | Investigate PyTorch-based LSTM and MLP models #199. Corrected dropout layer. Less blabla | 21 April 2022, 18:43:16 UTC |
b54fc4c | Antoine Carme | 21 April 2022, 15:50:10 UTC | Investigate PyTorch-based LSTM and MLP models #199. Improved alignment between keras and pytorch implementations. Similar options. A lot of cleanup. | 21 April 2022, 15:50:10 UTC |
320b290 | Antoine Carme | 19 April 2022, 18:07:53 UTC | Investigate PyTorch-based LSTM and MLP models #199. Updated keras models (WIP). | 19 April 2022, 18:07:53 UTC |
8aee5c3 | Antoine Carme | 19 April 2022, 18:07:18 UTC | Investigate PyTorch-based LSTM and MLP models #199. Updated this script. | 19 April 2022, 18:07:18 UTC |
144bdc2 | Antoine Carme | 19 April 2022, 17:10:52 UTC | Investigate PyTorch-based LSTM and MLP models #199. Updated tests Makefile. | 19 April 2022, 17:10:52 UTC |
7dc3e38 | Antoine Carme | 19 April 2022, 17:07:07 UTC | Investigate PyTorch-based LSTM and MLP models #199. Pytorch impact on model generation. | 19 April 2022, 17:07:07 UTC |
2d8810a | Antoine Carme | 19 April 2022, 17:06:00 UTC | Investigate PyTorch-based LSTM and MLP models #199. Pytorch impact on model options. | 19 April 2022, 17:06:00 UTC |
c914849 | Antoine Carme | 19 April 2022, 17:05:24 UTC | Investigate PyTorch-based LSTM and MLP models #199. Added first models using pytorch. | 19 April 2022, 17:05:24 UTC |
4698440 | Antoine Carme | 19 April 2022, 17:04:34 UTC | Investigate PyTorch-based LSTM and MLP models #199. Added some tests. | 19 April 2022, 17:04:34 UTC |
6c9a97b | CARME Antoine | 19 April 2022, 15:47:50 UTC | Merge pull request #200 from antoinecarme/Exogenous_Variables_Scale_issue_198 Exogenous variables scale issue 198 | 19 April 2022, 15:47:50 UTC |
5353e0c | Antoine Carme | 17 April 2022, 15:28:41 UTC | Updated these logs. | 17 April 2022, 15:28:41 UTC |
09efc13 | Antoine Carme | 17 April 2022, 00:22:32 UTC | Make the Modeling Scale when the Number of Exogenous Variables Increases #198. Added an exog preselection system. WIP. | 17 April 2022, 00:22:32 UTC |
1ed046a | Antoine Carme | 17 April 2022, 00:20:57 UTC | Make the Modeling Scale when the Number of Exogenous Variables Increases #198. Some refactoring. | 17 April 2022, 00:20:57 UTC |
4a71893 | Antoine Carme | 17 April 2022, 00:20:08 UTC | Make the Modeling Scale when the Number of Exogenous Variables Increases #198. Improved time series generation with large number of exog vars. | 17 April 2022, 00:20:08 UTC |
dcb7c2e | Antoine Carme | 17 April 2022, 00:19:35 UTC | Make the Modeling Scale when the Number of Exogenous Variables Increases #198. Improved time series generation with large number of exog vars. | 17 April 2022, 00:19:35 UTC |
ab11a6f | Antoine Carme | 16 April 2022, 23:44:25 UTC | Avoid a warning and an unnecessary import on exit. | 16 April 2022, 23:45:31 UTC |
fbb4c4d | Antoine Carme | 16 April 2022, 11:55:35 UTC | Release Process : Profiling and Warning Hunts #195. Avoid a warning on similar column names | 16 April 2022, 11:55:35 UTC |
386daa5 | Antoine Carme | 16 April 2022, 01:50:27 UTC | Release Process : Profiling and Warning Hunts #195. Avoid a warning on similar column names when merging forecast frames in hierarchical models. | 16 April 2022, 01:50:27 UTC |
c785c6f | Antoine Carme | 16 April 2022, 01:05:29 UTC | Release Process : Profiling and Warning Hunts #195. Avoid a warning on similar column names when merging forecast frames in hierarchical models. | 16 April 2022, 01:05:29 UTC |
e7a6360 | Antoine Carme | 16 April 2022, 00:16:23 UTC | Release Process : Profiling and Warning Hunts #195. Avoid a warning on similar column names when merging forecast frames in hierarchical models. | 16 April 2022, 00:16:23 UTC |
6625537 | Antoine Carme | 15 April 2022, 19:40:53 UTC | Release Process : Review Existing Docs #196. Checked and Refreshed all notebooks. | 15 April 2022, 19:40:53 UTC |
4399136 | CARME Antoine | 15 April 2022, 18:55:59 UTC | Update README.md | 15 April 2022, 18:55:59 UTC |
0ebdf7e | Antoine Carme | 15 April 2022, 18:49:03 UTC | Crash when training using example code #156. Adapted demo script. | 15 April 2022, 18:49:03 UTC |
22bfb32 | Antoine Carme | 15 April 2022, 18:48:57 UTC | Crash when training using example code #156. Adapted demo script. | 15 April 2022, 18:48:57 UTC |
e209e56 | Antoine Carme | 15 April 2022, 18:47:38 UTC | Refactored used module version methods. For debugging purposes. | 15 April 2022, 18:47:38 UTC |
e43ab06 | Antoine Carme | 15 April 2022, 18:47:34 UTC | Refactored used module version methods. For debugging purposes. | 15 April 2022, 18:47:34 UTC |
f7ecbc8 | Antoine Carme | 15 April 2022, 05:00:47 UTC | Corrected handling constant signals transfromations. | 15 April 2022, 05:00:47 UTC |
7fc7646 | Antoine Carme | 15 April 2022, 04:00:57 UTC | Attempt to avoid Ciricle-CI timeouts. Long lasting tests will fail. Again. Again. | 15 April 2022, 04:00:57 UTC |
4d27b88 | Antoine Carme | 15 April 2022, 03:46:36 UTC | Attempt to avoid Ciricle-CI timeouts. Long lasting tests will fail. Again. | 15 April 2022, 03:46:36 UTC |
a411f69 | Antoine Carme | 15 April 2022, 02:46:30 UTC | Attempt to avoid Ciricle-CI timeouts. Long lasting tests will fail. | 15 April 2022, 02:46:30 UTC |
7912ddb | Antoine Carme | 15 April 2022, 01:58:17 UTC | Merge branch 'LSTMX_MLPX' | 15 April 2022, 01:58:17 UTC |
d36cf67 | Antoine Carme | 15 April 2022, 00:22:26 UTC | Exogenous data are not used in MLP and LSTM-based models #197. Avoid CROSTONX. Added these logs | 15 April 2022, 00:22:26 UTC |
62d0469 | Antoine Carme | 15 April 2022, 00:20:42 UTC | Exogenous data are not used in MLP and LSTM-based models #197. Avoid CROSTONX. Updated these logs | 15 April 2022, 00:20:42 UTC |
18d1be1 | Antoine Carme | 14 April 2022, 23:40:03 UTC | Exogenous data are not used in MLP and LSTM-based models #197. Avoid CROSTONX. | 14 April 2022, 23:40:03 UTC |
b318890 | Antoine Carme | 14 April 2022, 21:02:37 UTC | Exogenous data are not used in MLP and LSTM-based models #197. Updated these logs. | 14 April 2022, 21:02:37 UTC |
46881b6 | Antoine Carme | 14 April 2022, 21:02:12 UTC | Exogenous data are not used in MLP and LSTM-based models #197. Added LSTMX and MLPX models. | 14 April 2022, 21:02:12 UTC |
a88756f | Antoine Carme | 14 April 2022, 19:52:47 UTC | Exogenous data are not used in MLP and LSTM-based models #197. LSTMX and MLPX models have more inputs, fixed. | 14 April 2022, 19:52:47 UTC |
713b561 | Antoine Carme | 14 April 2022, 19:51:52 UTC | Exogenous data are not used in MLP and LSTM-based models #197. Added new options for LSTMX and MLPX models. Some refactoring | 14 April 2022, 19:51:52 UTC |
27e7694 | Antoine Carme | 14 April 2022, 19:51:22 UTC | Exogenous data are not used in MLP and LSTM-based models #197. Added new options for LSTMX and MLPX models. | 14 April 2022, 19:51:22 UTC |
0d4a6d7 | Antoine Carme | 14 April 2022, 19:49:50 UTC | Exogenous data are not used in MLP and LSTM-based models #197. Added some tests. | 14 April 2022, 19:49:50 UTC |
5ef8bde | Antoine Carme | 09 April 2022, 18:50:45 UTC | Release Process : Profiling and Warning Hunts #195. Removed an unnecessary step. Handle some extreme cases. | 09 April 2022, 18:50:45 UTC |
ded8a84 | Antoine Carme | 09 April 2022, 18:49:54 UTC | Release Process : Profiling and Warning Hunts #195. Handle some extreme cases. | 09 April 2022, 18:49:54 UTC |