f66b45f | James Hensman | 13 April 2016, 22:07:31 UTC | Merge branch 'optimizer_tolerance' of github.com:GPflow/GPflow into optimizer_tolerance | 13 April 2016, 22:07:31 UTC |
b8e3d9d | James Hensman | 13 April 2016, 22:06:51 UTC | adding pass-through of tolerance to svgp.py | 13 April 2016, 22:06:51 UTC |
8390cff | alexggmatthews | 13 April 2016, 18:41:26 UTC | Merge branch 'fitc' into optimizer_tolerance | 13 April 2016, 18:41:26 UTC |
c9df777 | alexggmatthews | 13 April 2016, 17:21:38 UTC | Removing unnecessary compile. | 13 April 2016, 17:21:38 UTC |
e517970 | alexggmatthews | 13 April 2016, 17:14:01 UTC | Removing unnecessary compile. | 13 April 2016, 17:14:01 UTC |
9b35ef7 | James Hensman | 13 April 2016, 16:51:38 UTC | bugfix | 13 April 2016, 16:51:38 UTC |
be2edbd | James Hensman | 13 April 2016, 16:50:48 UTC | added tolerance parameter pass-through for scipy optimizers | 13 April 2016, 16:50:48 UTC |
9ea0562 | alexggmatthews | 13 April 2016, 14:13:22 UTC | Building up Snelson experiments. | 13 April 2016, 14:13:22 UTC |
0f2d9c9 | alexggmatthews | 13 April 2016, 13:41:10 UTC | Adding full covariance predictions to FITC. | 13 April 2016, 13:41:10 UTC |
880164a | alexggmatthews | 13 April 2016, 12:35:19 UTC | Building up FITC. | 13 April 2016, 12:35:19 UTC |
7314785 | alexggmatthews | 13 April 2016, 12:27:06 UTC | Renaming snelson demo. | 13 April 2016, 12:27:06 UTC |
69894ba | alexggmatthews | 13 April 2016, 12:26:03 UTC | Snelson demo of FITC vs VFE. | 13 April 2016, 12:26:03 UTC |
2b6d8ba | James Hensman | 13 April 2016, 11:22:49 UTC | simplification in FITC predict | 13 April 2016, 11:22:49 UTC |
5aeaa60 | James Hensman | 13 April 2016, 11:19:25 UTC | edited predicitons for FITC, added to SanityCheck notebook | 13 April 2016, 11:19:25 UTC |
8d9d459 | James Hensman | 13 April 2016, 11:03:03 UTC | FITC: tidied code and fixed sign problem in mahalanobis distance | 13 April 2016, 11:03:03 UTC |
dfcd9af | James Hensman | 13 April 2016, 10:25:48 UTC | merge --theirs | 13 April 2016, 10:25:48 UTC |
df931a9 | alexggmatthews | 13 April 2016, 10:16:16 UTC | FITC method for GPflow | 13 April 2016, 10:16:16 UTC |
d7a5de2 | Alexander G. de G. Matthews | 12 April 2016, 17:00:49 UTC | Merge pull request #35 from GPflow/white_kernels_in_test FIX: redundant white kernels in test (fixes #34) | 12 April 2016, 17:00:49 UTC |
4502659 | James Hensman | 12 April 2016, 15:41:46 UTC | removing white kernels from sanity check notebook | 12 April 2016, 15:41:46 UTC |
853a8c9 | James Hensman | 12 April 2016, 15:39:53 UTC | removing unnecessary white kernels from method equivalence test | 12 April 2016, 15:39:53 UTC |
1eeebe1 | alexggmatthews | 12 April 2016, 14:00:06 UTC | Removing extraneous compile. | 12 April 2016, 14:00:06 UTC |
fab4556 | alexggmatthews | 12 April 2016, 13:51:17 UTC | Removing extraneous optimize. | 12 April 2016, 13:51:17 UTC |
bfef92c | alexggmatthews | 12 April 2016, 13:38:19 UTC | Removing unnecessary compiles. | 12 April 2016, 13:38:19 UTC |
9063c7e | alexggmatthews | 11 April 2016, 09:51:56 UTC | Adding comment about numerical inversion. | 11 April 2016, 09:51:56 UTC |
cb856cb | James Hensman | 09 April 2016, 14:18:10 UTC | removeing suspect test | 09 April 2016, 14:18:10 UTC |
19dc98a | James Hensman | 08 April 2016, 13:06:36 UTC | attempting travis without nose | 08 April 2016, 13:06:36 UTC |
9692ef3 | James Hensman | 08 April 2016, 09:19:51 UTC | allowing logging output in travis.yml | 08 April 2016, 09:19:51 UTC |
a19481f | James Hensman | 08 April 2016, 08:16:30 UTC | tring to speed up method_equivalence tests | 08 April 2016, 08:16:30 UTC |
659249e | James Hensman | 06 April 2016, 16:43:07 UTC | adding timing to travis | 06 April 2016, 16:43:07 UTC |
cd4e93a | James Hensman | 06 April 2016, 16:41:24 UTC | Merge branch 'master' of github.com:GPflow/GPflow | 06 April 2016, 16:41:24 UTC |
cee4dd1 | alexggmatthews | 01 April 2016, 16:16:38 UTC | Merge branch 'master' of https://github.com/GPflow/GPflow | 01 April 2016, 16:16:38 UTC |
96dfb24 | alexggmatthews | 01 April 2016, 16:16:26 UTC | Adding travis wait. | 01 April 2016, 16:16:26 UTC |
430c6eb | James Hensman | 30 March 2016, 16:02:48 UTC | Fixed typos in GPR docstring | 30 March 2016, 16:02:48 UTC |
c935db2 | Alexander G. de G. Matthews | 30 March 2016, 13:51:02 UTC | Merge pull request #30 from GPflow/table-of-methods-for-README table of methods for README | 30 March 2016, 13:51:02 UTC |
cf254b8 | Alexander G. de G. Matthews | 30 March 2016, 13:50:26 UTC | Merge pull request #27 from GPflow/full_cov_predict Full covariance matrix in predictions | 30 March 2016, 13:50:26 UTC |
22edbc4 | alexggmatthews | 30 March 2016, 13:48:49 UTC | Correcting typo in comment. | 30 March 2016, 13:48:49 UTC |
efc9529 | James Hensman | 30 March 2016, 13:45:00 UTC | table of methods for README I think the attached table makes a nice addition to the README. | 30 March 2016, 13:45:00 UTC |
005acd0 | alexggmatthews | 30 March 2016, 11:49:31 UTC | First pass at FITC approximation to the marginal likelihood. | 30 March 2016, 11:49:31 UTC |
0cd507a | alexggmatthews | 30 March 2016, 10:49:31 UTC | Moving code to sgpr.py | 30 March 2016, 10:49:31 UTC |
0069b03 | alexggmatthews | 30 March 2016, 10:41:20 UTC | Updating comments. | 30 March 2016, 10:41:20 UTC |
d80c4ed | alexggmatthews | 30 March 2016, 10:36:28 UTC | Beginnings of gprfitc. | 30 March 2016, 10:36:28 UTC |
c5ea098 | Alexander G. de G. Matthews | 29 March 2016, 18:24:21 UTC | Merge pull request #28 from GPflow/test_improvements Test improvements | 29 March 2016, 18:24:21 UTC |
5bdd8ef | James Hensman | 29 March 2016, 15:13:25 UTC | KeyboardInterrupt tests KeyboardInterrupt tests now cover both scipy optimization and tf optimization. | 29 March 2016, 15:13:25 UTC |
3e4ed1b | James Hensman | 29 March 2016, 14:05:42 UTC | New test in model.py This test makes sure that when a KeyboardInterrupt occurs during optimization, the state of the model is captured. | 29 March 2016, 14:05:42 UTC |
69dc6f6 | James Hensman | 29 March 2016, 11:39:54 UTC | Improved likelihood tests. Likelihoods are now tested with non-standard link functions (like f^2, instead of exp(f)), and variational_expectations with zero variance is tested to be equivalent to a simple call to logp. | 29 March 2016, 11:39:54 UTC |
1dd6b52 | James Hensman | 29 March 2016, 11:05:49 UTC | testing slices in assymetric kerel operations: K(X, Z) | 29 March 2016, 11:05:49 UTC |
e053c94 | James Hensman | 29 March 2016, 10:52:42 UTC | testsing diag of prod kernel | 29 March 2016, 10:52:42 UTC |
9c29e24 | James Hensman | 29 March 2016, 10:51:37 UTC | now testing diagonal kernel functions | 29 March 2016, 10:51:37 UTC |
766c3c0 | Alexander G. de G. Matthews | 29 March 2016, 08:49:33 UTC | Merge pull request #26 from GPflow/mean_fn_tests adding tests for mean functions | 29 March 2016, 08:49:33 UTC |
bf0beb4 | James Hensman | 27 March 2016, 20:02:14 UTC | small decrease in required accuracy in gradients | 27 March 2016, 20:02:14 UTC |
6084265 | James Hensman | 27 March 2016, 18:55:01 UTC | more edits to RELEASE.md | 27 March 2016, 18:55:01 UTC |
aafa969 | James Hensman | 27 March 2016, 18:52:46 UTC | silincing LBFGS-B during tests | 27 March 2016, 18:52:46 UTC |
71eafc8 | James Hensman | 27 March 2016, 18:50:43 UTC | FIX: missing conditional import in sgpmc | 27 March 2016, 18:50:43 UTC |
48f4cff | James Hensman | 27 March 2016, 18:26:37 UTC | bumped the version number | 27 March 2016, 18:26:37 UTC |
29d2b8c | James Hensman | 27 March 2016, 18:22:20 UTC | ensuring compatibility of deprecated conditional functions | 27 March 2016, 18:22:20 UTC |
b51a605 | James Hensman | 27 March 2016, 18:12:55 UTC | ensuring full_cov sizes are consistent in gpmc, sgpmc | 27 March 2016, 18:12:55 UTC |
2546cd8 | James Hensman | 27 March 2016, 17:53:20 UTC | refactored conditionals.py | 27 March 2016, 17:53:20 UTC |
6a6e906 | James Hensman | 27 March 2016, 12:42:36 UTC | simplified loop in conditionals | 27 March 2016, 12:43:57 UTC |
d1053f1 | James Hensman | 27 March 2016, 11:56:28 UTC | adding tests for full_cov | 27 March 2016, 11:56:28 UTC |
288501b | James Hensman | 27 March 2016, 11:35:42 UTC | Merge branch 'master' into full_cov_predict | 27 March 2016, 11:35:42 UTC |
323eb7a | James Hensman | 27 March 2016, 10:49:26 UTC | adding tests for mean functions | 27 March 2016, 10:49:26 UTC |
d3fc6ed | Alexander G. de G. Matthews | 24 March 2016, 17:16:37 UTC | Merge pull request #24 from GPflow/paramlist_bugfix made tf_mode work for paramlist, added appropriate test | 24 March 2016, 17:16:37 UTC |
487713c | James Hensman | 24 March 2016, 16:13:57 UTC | Merge branch 'master' of github.com:GPflow/GPflow | 24 March 2016, 16:13:57 UTC |
401e72f | James Hensman | 24 March 2016, 15:02:07 UTC | minor fixes in SGPR notes | 24 March 2016, 16:13:35 UTC |
a4e9179 | James Hensman | 18 March 2016, 16:18:58 UTC | improved testing of paramlist | 24 March 2016, 16:13:35 UTC |
291d567 | James Hensman | 18 March 2016, 15:46:42 UTC | adding tests for parameter list | 24 March 2016, 16:13:35 UTC |
823c36c | James Hensman | 17 February 2016, 22:25:17 UTC | minor edits | 24 March 2016, 16:13:35 UTC |
87009e0 | James Hensman | 17 February 2016, 22:07:17 UTC | fist attempt at a paramlist object | 24 March 2016, 16:13:34 UTC |
800f684 | James Hensman | 24 March 2016, 16:06:38 UTC | made tf_mode work for paramlist, added appropriate test | 24 March 2016, 16:06:38 UTC |
33f922a | James Hensman | 24 March 2016, 15:07:59 UTC | Update README to include note on sgpr. | 24 March 2016, 15:07:59 UTC |
36f24dd | James Hensman | 24 March 2016, 15:02:07 UTC | minor fixes in SGPR notes | 24 March 2016, 15:02:07 UTC |
e3d2ca2 | Alexander G. de G. Matthews | 24 March 2016, 14:18:35 UTC | Merge pull request #17 from GPflow/paramlist ParamList | 24 March 2016, 14:18:35 UTC |
bea7bb6 | alexggmatthews | 24 March 2016, 14:10:29 UTC | Small correction to README file. | 24 March 2016, 14:10:29 UTC |
cdd06fa | alexggmatthews | 24 March 2016, 14:07:17 UTC | Correcting numbers for references. | 24 March 2016, 14:07:17 UTC |
76d16c5 | Alexander G. de G. Matthews | 24 March 2016, 13:37:06 UTC | Merge pull request #20 from GPflow/sgpr Sgpr | 24 March 2016, 13:37:06 UTC |
a2aae70 | alexggmatthews | 24 March 2016, 13:35:40 UTC | Adding citation to Titsias 2009. | 24 March 2016, 13:35:40 UTC |
d6e0cdc | James Hensman | 24 March 2016, 11:35:39 UTC | extra notes for sgpr prediction | 24 March 2016, 11:35:39 UTC |
bce3ea5 | James Hensman | 24 March 2016, 11:08:56 UTC | Adding SGPR notebook on bound derivation | 24 March 2016, 11:08:56 UTC |
32380d1 | James Hensman | 24 March 2016, 11:08:32 UTC | tidien sgpr code | 24 March 2016, 11:08:32 UTC |
e2a9ff7 | James Hensman | 22 March 2016, 18:57:36 UTC | bugfix | 22 March 2016, 18:57:36 UTC |
f898633 | James Hensman | 22 March 2016, 16:22:42 UTC | merging | 22 March 2016, 16:22:42 UTC |
bf192ae | James Hensman | 22 March 2016, 16:18:47 UTC | adding full_covariance predictions to the sgpr method | 22 March 2016, 16:18:47 UTC |
cae49e8 | alexggmatthews | 22 March 2016, 12:46:43 UTC | Correcting minor typo in README.md | 22 March 2016, 12:46:43 UTC |
a599476 | James Hensman | 22 March 2016, 10:27:54 UTC | Merge branch 'master' into full_cov_predict | 22 March 2016, 10:27:54 UTC |
3cdc945 | James Hensman | 22 March 2016, 10:27:09 UTC | master README now points at master branch travis/codecov | 22 March 2016, 10:27:09 UTC |
b662c35 | James Hensman | 22 March 2016, 10:25:19 UTC | assorted bugfixes and tidying | 22 March 2016, 10:25:19 UTC |
3b04a0e | James Hensman | 22 March 2016, 10:24:52 UTC | added GPmodel functions: predict_f_full_cov and predict_f_samples | 22 March 2016, 10:24:52 UTC |
3bf18f0 | alexggmatthews | 22 March 2016, 10:09:22 UTC | Merge branch 'master' into paramlist | 22 March 2016, 10:09:22 UTC |
270668f | alexggmatthews | 22 March 2016, 10:08:09 UTC | Merge branch 'master' into sgpr | 22 March 2016, 10:08:09 UTC |
7f32f21 | Alexander G. de G. Matthews | 22 March 2016, 09:59:42 UTC | Merge pull request #18 from GPflow/travis Continuous integration | 22 March 2016, 09:59:42 UTC |
1d192ca | James Hensman | 21 March 2016, 17:45:35 UTC | added full_cov option to most methods | 21 March 2016, 17:45:35 UTC |
e18fc5e | James Hensman | 21 March 2016, 08:39:31 UTC | Improving sgpr docstring | 21 March 2016, 08:39:31 UTC |
f5cb9b5 | James Hensman | 20 March 2016, 18:03:38 UTC | added tests to ensure methods are equivalent when appropriate | 21 March 2016, 08:32:16 UTC |
96b8d06 | James Hensman | 20 March 2016, 14:32:29 UTC | added sgpr to standard methods tests | 21 March 2016, 08:32:16 UTC |
0417a22 | James Hensman | 20 March 2016, 14:25:26 UTC | small bugfix in prediction | 21 March 2016, 08:32:16 UTC |
20109ef | James Hensman | 20 March 2016, 13:58:17 UTC | first crack at prediction code | 21 March 2016, 08:32:16 UTC |
038561b | James Hensman | 08 March 2016, 15:30:01 UTC | Adding SGPR class | 21 March 2016, 08:32:02 UTC |
04172a9 | James Hensman | 20 March 2016, 19:36:35 UTC | improves test_hmc code | 21 March 2016, 08:01:41 UTC |
942d6fc | James Hensman | 20 March 2016, 09:20:38 UTC | turned off logging during tests | 21 March 2016, 08:01:41 UTC |
d1a155f | James Hensman | 19 March 2016, 20:16:47 UTC | tidied travis.yml | 21 March 2016, 08:01:41 UTC |