https://github.com/antoinecarme/pyaf

sort by:
Revision Author Date Message Commit Date
ef20599 Use Python Logging #10 Update these logs. 08 December 2016, 08:45:09 UTC
3f5707e Enhance artificial dataset tests. #6 Added logging in these tests. 08 December 2016, 08:37:12 UTC
797dba0 Enhance artificial dataset tests. #6 Corrected a bad previous commit. Small signals were not properly modelled. 08 December 2016, 08:35:50 UTC
86882b8 Save plots (for debugging purposes). 07 December 2016, 09:37:27 UTC
310fbb6 Use pyaf.std instaed of pyaf dor logging low-level functions. 07 December 2016, 09:36:01 UTC
e89d98f Use Python Logging #10 Sample config file. 06 December 2016, 10:15:24 UTC
baa7ef6 This test should fail more often (for warnings). 06 December 2016, 10:15:24 UTC
a54d69f Use Python Logging #10 Some test. 06 December 2016, 10:15:24 UTC
42596ef Use Python Logging #10 Adapt existing code. Use logger.info instead of print 06 December 2016, 10:15:24 UTC
d451a4a Use Python Logging #10 Added two loggers : one for generic modeling and one for hierarchical. 06 December 2016, 10:15:24 UTC
e4f1831 Use Python Logging #10 Corrected AR models (when the residue is almost constant). Added a failing test. 06 December 2016, 10:15:24 UTC
bd17967 Delete README.odt Removed some old file. 04 December 2016, 15:02:12 UTC
c4de244 Delete README.html Removed some old file. 04 December 2016, 15:01:54 UTC
b517868 Isolate codegen tests for the moment. 02 December 2016, 09:12:03 UTC
54d9cca Use Python Logging #10 Prototyping notebook 02 December 2016, 09:08:33 UTC
3d46125 Add a Jupyter Notebook to demonstrate the use of Hierarchical Forecasting #9 Updated the notebook : 1. Hierarchy definition and plots. 2. Started the model training. 02 December 2016, 08:09:27 UTC
57af39a Remvoed an redundant member. 02 December 2016, 08:08:48 UTC
2c414b4 Use outputs directory for storing plots produced by these tests. 01 December 2016, 12:45:25 UTC
ffa3f90 Add a Jupyter Notebook to demonstrate the use of Hierarchical Forecasting #9 Added a test script matching the notebook model. 01 December 2016, 08:31:32 UTC
ab6c7c6 Add a Jupyter Notebook to demonstrate the use of Hierarchical Forecasting #9 Add the possibility to plot model hierarchies. 01 December 2016, 08:30:11 UTC
698343b Added a reference for exogenous notebook. 30 November 2016, 12:02:18 UTC
71a3922 Add a Jupyter Notebook to demonstrate the use of Hierarchical Forecasting #9 Data description WIP 30 November 2016, 08:30:12 UTC
8762c0c Perform signal transformatiion in a uniform way #7 prototyping notebook. 30 November 2016, 08:30:12 UTC
142e04c Update README.md 29 November 2016, 13:28:48 UTC
2e4cb33 Update README.md 29 November 2016, 13:27:24 UTC
f5d9c37 Perform signal transformatiion in a uniform way #7 The relative difference transformation is buggy. Disabling for the moment. 23 November 2016, 14:42:16 UTC
eda74f7 Compress a large csv file. 23 November 2016, 14:18:16 UTC
f876d5c Added some failing tests. 23 November 2016, 14:14:53 UTC
a59f3d5 1. Update SQL code generatiion for transrfofmations. 2. Some tests on Oracle database. 23 November 2016, 14:09:41 UTC
b0d52f7 Corrected plots. 22 November 2016, 10:54:50 UTC
59abcbb Enhance artificial dataset tests. #6 Updated this test. 22 November 2016, 07:42:15 UTC
766d909 Disabled idempotency tests by default. 22 November 2016, 07:39:44 UTC
666fb6c Enhance artificial dataset tests. #6 Updated dataset geenration. 22 November 2016, 07:38:58 UTC
86d3cab Perform signal transformation in a uniform way #7 Restructured signal transformations. Seprated trasnaltion/scaling invariance from real transformations. 22 November 2016, 07:36:46 UTC
16f3479 Perform signal transformation in a uniform way #7 Correcetd two bugs around signal transformation. 22 November 2016, 07:35:48 UTC
a3b46d0 Perform signal transformation in a uniform way #7 Removed some unnecessary code ... 22 November 2016, 07:35:00 UTC
798e65f Perform signal transformation in a uniform way #7 Added new options. 22 November 2016, 07:34:17 UTC
cc135aa Perform signal transformation in a uniform way #7 Added new options. 22 November 2016, 07:33:22 UTC
76e70c6 Perform signal transformation in a uniform way #7 Added new tests, 22 November 2016, 07:32:13 UTC
090b426 Perform signal transformation in a uniform way #7 Some tests. 22 November 2016, 07:30:26 UTC
451737b speedup the modleling process removved some unnnecessary frame copies. 21 November 2016, 09:43:01 UTC
017ec1a Speedup the modelling process. 21 November 2016, 09:40:35 UTC
e8f5c54 Enhance artificial dataset tests. #6 Automatically generated tests. 20 November 2016, 17:23:29 UTC
8cd8da1 Enhance artificial dataset tests. #6 sample notebook. 19 November 2016, 10:28:10 UTC
85a73f3 Enhance artificial dataset tests. #6 Artifical daatset generation improvement (WIP) 19 November 2016, 10:23:45 UTC
49729ce Reduce memory footprint #5 Corrrected a typo. 19 November 2016, 10:21:36 UTC
9928946 Reduce memory footprint #5 ARXX models footprint improved. 19 November 2016, 10:20:29 UTC
e545e91 Reduce memory footprint #5 tets notebook 19 November 2016, 10:18:43 UTC
4806071 Reduce memory footprint #5 Exogenous datasets mmeory usage improved. 19 November 2016, 10:16:58 UTC
c177ca9 Added dynamical part of all yahoo finance datasets. 17 November 2016, 11:19:07 UTC
deed6c1 Support Date types #4 Added prper time variable for some benchmarks. 17 November 2016, 11:17:59 UTC
07a404d Avoid some numpy warnings when computing personr for constant predictors. 16 November 2016, 18:59:22 UTC
b74cad3 Added soem mechanism to filter out low quality predictors. 16 November 2016, 18:58:18 UTC
c207a7e Added some exclusion mechanism to filter out low-quality predictors. 16 November 2016, 18:57:15 UTC
6b303bf Support Date types #4 Use original date format in the benchmarks. 16 November 2016, 18:55:24 UTC
81bf5a4 Support Date types #4 Force time values type to be the same (future dates and past dates). 16 November 2016, 10:06:28 UTC
0f704b3 Added new hierarchical tests. 14 November 2016, 19:35:29 UTC
2dfe26e Updated the column names. 14 November 2016, 19:35:05 UTC
03a72b4 Added options for controlling the combination methods. 14 November 2016, 19:34:33 UTC
0187bfc Added hierarchies tests with various combination approaches (BU, TD, MO, OC) 14 November 2016, 19:33:51 UTC
157a294 Added grouped time series tests with various combination approaches (BU, TD, MO, OC) 14 November 2016, 19:33:08 UTC
f7a6333 Added a comment. 14 November 2016, 19:32:13 UTC
921a074 Moved this test in the right test directory. 14 November 2016, 13:16:29 UTC
0bedd00 Unified the hierarchies with grouped signals. Now a grouped signal is only a special kind of hierarchy (generated in a sophisticated way!). WIP for middle out forecasting method. 14 November 2016, 13:03:18 UTC
2b387a5 Added some hierarchy test (AU). 14 November 2016, 13:02:25 UTC
312b1fc Added some test for grouped series. 14 November 2016, 12:41:25 UTC
4220643 Compare the sql execution wit the file-based apply-out. use python object class names to distinguish reliably between trend./cycle types. 13 November 2016, 15:09:52 UTC
c287024 Additional test. 13 November 2016, 15:06:42 UTC
1a35395 Unified signal name between the modeling and code generation. 12 November 2016, 19:31:48 UTC
6175a5f Corrected exogenous variables code generation. 12 November 2016, 18:16:32 UTC
bccfcca Corrected SQL generation for cycles. 11 November 2016, 23:13:40 UTC
8731c42 Example used in the exogenous notebook. 09 November 2016, 22:01:46 UTC
a8d246d Dispaly the 10 most interesting predictors for the ARX componenet. 09 November 2016, 22:00:43 UTC
c55e108 Display the most intersting predictors for the AR component. 09 November 2016, 21:59:53 UTC
cc71cb6 1. renamed encoded exogenous avriables. 2. Correcetd some encoding for future exogenous data. 09 November 2016, 21:59:02 UTC
1303097 Added some doc/notebook describing the use of exogenous variables. 09 November 2016, 21:58:16 UTC
0206103 sample encoding of exogenous dataset (for docs) 09 November 2016, 21:46:27 UTC
26d07f3 Added a dataset for documenting exogenous variables feature. 09 November 2016, 21:43:08 UTC
0a5533b Corrected the NN5 benchmark script. 08 November 2016, 20:17:52 UTC
02a12b4 limit the number of cpus used when running the benchamrks. 08 November 2016, 20:17:52 UTC
d9ab3e4 Update README.md 01 November 2016, 08:52:09 UTC
eff5060 Update README.md 01 November 2016, 08:50:57 UTC
ff41a9b Update README.md 14 October 2016, 17:54:01 UTC
08ed896 Update AUTHORS.rst 14 October 2016, 17:52:35 UTC
37fb666 Corrected links inside this doc. 14 October 2016, 17:51:45 UTC
b6630c6 Increased the length of GOOG signal. 14 October 2016, 11:25:21 UTC
470c943 Update README.md 14 October 2016, 10:49:36 UTC
518484d Update README.md 14 October 2016, 10:49:00 UTC
55fc67d reduced plot size again 14 October 2016, 10:42:49 UTC
cafc931 Redcued the size of the plots. 14 October 2016, 10:41:11 UTC
8b00ddf Some introductory notebook to time series forecasting with PyAF. 14 October 2016, 10:34:41 UTC
995c7c0 Updated these scripts. 13 October 2016, 21:55:01 UTC
0d9dc2e Updated cac40 notebook. 13 October 2016, 18:48:55 UTC
7c198ec Update AUTHORS.rst 13 October 2016, 17:25:26 UTC
89aa3ba updated these notebooks 13 October 2016, 17:13:39 UTC
6db50b5 Corrected imports. 13 October 2016, 17:13:22 UTC
136773f Corrected imports. 13 October 2016, 17:12:53 UTC
11b91f2 Updated this file. 13 October 2016, 17:12:33 UTC
2931205 Update README.md 13 October 2016, 17:10:52 UTC
6ad65f6 Update README.md 13 October 2016, 17:08:49 UTC
back to top