1da8382 | James Hensman | 29 July 2016, 19:38:40 UTC | Merge branch 'master' into active_dims_pickle_bug | 29 July 2016, 19:38:40 UTC |
bd0b423 | James Hensman | 29 July 2016, 19:27:15 UTC | make kernels with active_dims pickleable | 29 July 2016, 19:27:15 UTC |
e155d76 | James Hensman | 29 July 2016, 09:25:43 UTC | Track logprobs (#136) * added code for returning log-densities from HMC * added test code for log-prob tracking * adding access to logprobs from model class | 29 July 2016, 09:25:43 UTC |
6fa46bb | James Hensman | 28 July 2016, 16:30:51 UTC | enables pickling with fixed parameters (#141) | 28 July 2016, 16:30:51 UTC |
e0719ed | James Hensman | 28 July 2016, 16:12:19 UTC | some small edits to keep GPflow working with python 3 (#137) * python3.5 compat in test_likelihoods * py35 compatibility in autoflow | 28 July 2016, 16:12:19 UTC |
b4a8d1b | James Hensman | 27 July 2016, 14:40:51 UTC | made transforms all pickleable (#134) | 27 July 2016, 14:40:51 UTC |
9497d34 | James Hensman | 24 July 2016, 12:05:57 UTC | A dummy commit to trigger codecov.io | 24 July 2016, 12:05:57 UTC |
ddd0c3f | Alexander G. de G. Matthews | 13 July 2016, 08:04:12 UTC | Merge pull request #130 from GPflow/autokill_autoflow Autokill autoflow | 13 July 2016, 08:04:12 UTC |
985946b | James Hensman | 12 July 2016, 21:52:16 UTC | extending test to cover simplified autoflow killing | 12 July 2016, 21:52:16 UTC |
486775c | James Hensman | 12 July 2016, 21:45:01 UTC | some simplification of kill_autoflow | 12 July 2016, 21:45:01 UTC |
874957c | Alexander G. de G. Matthews | 11 July 2016, 11:01:08 UTC | Merge pull request #89 from GPflow/get_param_index a simple function to work out where a parameter is on the free_state | 11 July 2016, 11:01:08 UTC |
ce66d24 | Alexander G. de G. Matthews | 11 July 2016, 10:39:30 UTC | Merge branch 'master' into get_param_index | 11 July 2016, 10:39:30 UTC |
28fb48a | Alexander G. de G. Matthews | 11 July 2016, 10:38:58 UTC | Merge pull request #128 from GPflow/jitter_for_samples adding slight jitter to cholesky when computing samples | 11 July 2016, 10:38:58 UTC |
896ed83 | James Hensman | 11 July 2016, 08:36:00 UTC | Merge branch 'master' into get_param_index | 11 July 2016, 08:36:00 UTC |
31b48a1 | Alexander G. de G. Matthews | 11 July 2016, 08:27:39 UTC | Merge branch 'master' into jitter_for_samples | 11 July 2016, 08:27:39 UTC |
02286ca | Alexander G. de G. Matthews | 11 July 2016, 08:27:20 UTC | Merge pull request #127 from GPflow/softplus using the built-in softplus function for simplicity | 11 July 2016, 08:27:20 UTC |
30c2190 | James Hensman | 10 July 2016, 20:37:01 UTC | adding slight jitter to cholesky when computing samples | 10 July 2016, 20:37:01 UTC |
7172bd3 | James Hensman | 10 July 2016, 07:01:58 UTC | using the built-in softplu function for simplicity | 10 July 2016, 07:01:58 UTC |
691b20b | James Hensman | 09 July 2016, 13:25:40 UTC | minor edits to tests | 09 July 2016, 13:25:40 UTC |
3235b0d | James Hensman | 09 July 2016, 11:59:18 UTC | changes to regression ipynb | 09 July 2016, 11:59:18 UTC |
5ae3f60 | James Hensman | 09 July 2016, 11:57:35 UTC | minor edits in param.py - get_sample_dict became get_samples_df (since it retuns a datafram - set_parameter_dict no longer 'pops', resulting in an empty dict | 09 July 2016, 11:57:35 UTC |
1388d9f | James Hensman | 09 July 2016, 11:31:29 UTC | Merge branch 'master' into get_param_index | 09 July 2016, 11:31:29 UTC |
dc5ac86 | James Hensman | 09 July 2016, 10:54:46 UTC | Correction to VGP from @fujiisoup (#126) Correction to VGP from @fujiisoup. See also #125 | 09 July 2016, 10:54:46 UTC |
e0b2217 | Mark van der Wilk | 09 July 2016, 09:09:55 UTC | Merge pull request #123 from GPflow/lognormal Lognormal | 09 July 2016, 09:09:55 UTC |
73488dc | Mark van der Wilk | 09 July 2016, 08:59:24 UTC | Merge branch 'master' into lognormal | 09 July 2016, 08:59:24 UTC |
9d005bc | James Hensman | 08 July 2016, 15:18:59 UTC | Merge pull request #122 from GPflow/svi-example Extended svi_test.ipynb to show how to use the stochastic optimizer | 08 July 2016, 15:18:59 UTC |
f959f29 | James Hensman | 08 July 2016, 12:00:29 UTC | improvements to svi notebook | 08 July 2016, 12:00:29 UTC |
3722e10 | James Hensman | 08 July 2016, 11:28:26 UTC | correction to lognormal test | 08 July 2016, 11:28:26 UTC |
4f21e67 | James Hensman | 08 July 2016, 11:27:14 UTC | adding a lognormal prior | 08 July 2016, 11:27:14 UTC |
5faf578 | Mark van der Wilk | 08 July 2016, 10:54:18 UTC | Extended svi_test.ipynb to show how to use the stochastic optimizer. | 08 July 2016, 10:54:18 UTC |
b331313 | Mark van der Wilk | 08 July 2016, 10:02:43 UTC | Merge branch 'master' of github.com:GPflow/GPflow into get_param_index | 08 July 2016, 10:02:43 UTC |
62972a2 | Mark van der Wilk | 07 July 2016, 11:49:10 UTC | Fixed tests and small bug. | 07 July 2016, 11:49:10 UTC |
57876ab | Alexander G. de G. Matthews | 07 July 2016, 06:52:26 UTC | Merge pull request #120 from GPflow/lower_bound_transforms improvements to transforms | 07 July 2016, 06:52:26 UTC |
3f71351 | James Hensman | 06 July 2016, 17:33:24 UTC | improvements to transforms | 06 July 2016, 17:33:24 UTC |
dea3e94 | Mark van der Wilk | 06 July 2016, 12:55:46 UTC | get_samples_dict now returns a pandas DataFrame object. | 06 July 2016, 12:55:46 UTC |
6253621 | Mark van der Wilk | 06 July 2016, 11:11:16 UTC | Merge branch 'master' of github.com:GPflow/GPflow into get_param_index | 06 July 2016, 11:11:16 UTC |
de38ef7 | Alexander G. de G. Matthews | 01 July 2016, 12:31:40 UTC | Merge pull request #108 from GPflow/Placeholder Placeholder | 01 July 2016, 12:31:40 UTC |
ab157e8 | James Hensman | 30 June 2016, 09:27:22 UTC | testing samples_dict with fixed param | 30 June 2016, 09:27:22 UTC |
e8d7432 | James Hensman | 30 June 2016, 09:24:52 UTC | a simple test for get_samples_dict | 30 June 2016, 09:24:52 UTC |
45bcb9c | James Hensman | 30 June 2016, 09:13:42 UTC | improvements to the regression notebook | 30 June 2016, 09:13:42 UTC |
8a58c57 | James Hensman | 30 June 2016, 08:53:56 UTC | a new function to make an array of samples into a dictionary of the same form as get_param_dict | 30 June 2016, 08:53:56 UTC |
5be1aa6 | James Hensman | 30 June 2016, 08:31:26 UTC | improved docstring | 30 June 2016, 08:31:26 UTC |
6a0dc6e | James Hensman | 30 June 2016, 08:27:26 UTC | trivial merge | 30 June 2016, 08:27:26 UTC |
deaab00 | James Hensman | 29 June 2016, 19:53:31 UTC | some improvements to update_data notebook | 29 June 2016, 19:53:31 UTC |
e425bbc | James Hensman | 29 June 2016, 19:20:47 UTC | removing more notebooks | 29 June 2016, 19:20:47 UTC |
d2c16d9 | James Hensman | 29 June 2016, 19:15:38 UTC | removing ipynb_checkpoints | 29 June 2016, 19:15:38 UTC |
ca32ff2 | James Hensman | 29 June 2016, 17:06:04 UTC | small improvements to test_param | 29 June 2016, 17:06:04 UTC |
e8b04bc | James Hensman | 29 June 2016, 16:11:29 UTC | trivial merge | 29 June 2016, 16:11:29 UTC |
804435b | Alexander G. de G. Matthews | 29 June 2016, 15:24:25 UTC | Merge pull request #112 from GPflow/to_dict `to_dict` | 29 June 2016, 15:24:25 UTC |
86a4ac5 | James Hensman | 29 June 2016, 14:12:23 UTC | Merge branch 'master' into to_dict | 29 June 2016, 14:12:23 UTC |
70ca464 | James Hensman | 29 June 2016, 14:11:40 UTC | Merge pull request #114 from GPflow/PabloLeon-readmeTypos Update README.md. Thanks @PabloLeon | 29 June 2016, 14:11:40 UTC |
8a32f66 | PabloLeon | 29 June 2016, 13:55:34 UTC | Update README.md | 29 June 2016, 13:55:34 UTC |
1a5fba1 | James Hensman | 29 June 2016, 13:50:05 UTC | improving tests for models that should recompile | 29 June 2016, 13:50:05 UTC |
c140477 | James Hensman | 29 June 2016, 13:45:31 UTC | correcting no-cover statement | 29 June 2016, 13:45:31 UTC |
4505df2 | James Hensman | 29 June 2016, 13:28:33 UTC | removing exectuable bit for test_data_object | 29 June 2016, 13:28:33 UTC |
c050115 | James Hensman | 29 June 2016, 12:51:49 UTC | removing debug statements | 29 June 2016, 12:51:49 UTC |
0e1aa36 | James Hensman | 29 June 2016, 12:08:16 UTC | removing remnanat of calc_feed_dict | 29 June 2016, 12:48:27 UTC |
1e6d90c | Alexander G. de G. Matthews | 29 June 2016, 12:27:17 UTC | Merge branch 'master' into to_dict | 29 June 2016, 12:27:17 UTC |
5c1a44a | Mark van der Wilk | 29 June 2016, 12:18:44 UTC | Merge pull request #113 from GPflow/kern_slicing Kern slicing | 29 June 2016, 12:18:44 UTC |
4bd0764 | James Hensman | 29 June 2016, 10:57:02 UTC | lots of new simple tests for DAtaHolder | 29 June 2016, 10:57:02 UTC |
c5a1306 | James Hensman | 29 June 2016, 10:37:39 UTC | Updating RELEASE, _version, README incrementing version to 0.2.0 adding @fujiisoup to README detailing new features in RELEASE | 29 June 2016, 10:37:39 UTC |
6bb7e15 | James Hensman | 29 June 2016, 10:29:55 UTC | minor edits to docstrings | 29 June 2016, 10:29:55 UTC |
dfed3ca | James Hensman | 29 June 2016, 10:08:51 UTC | Merge remote-tracking branch 'origin' into Placeholder | 29 June 2016, 10:08:51 UTC |
8e008d7 | James Hensman | 29 June 2016, 09:45:53 UTC | adding a string function for the DataHolder Class2 | 29 June 2016, 09:45:53 UTC |
8ed17d0 | James Hensman | 29 June 2016, 09:40:33 UTC | silly bug fix | 29 June 2016, 09:40:33 UTC |
7905bb6 | James Hensman | 29 June 2016, 08:36:52 UTC | Changing data now only recompiles if necessary An option in the dataholder class lets it choose what to do when the shape of data are changed. So now we can do X = np.random.randn(10, 1) Y = np.sin(X) m = GPflow.models.GPR(X, Y, GPflow.kernels.RBF(1)) m.optimize() X_new = np.random.randn(12, 1) Y_new = np.sin(X_new) m.X = X_new m.Y = Y_new m.optimize() # no recompile neded. | 29 June 2016, 08:36:52 UTC |
61df1f1 | James Hensman | 28 June 2016, 19:00:54 UTC | some pep8 changes in test_kerns.py | 28 June 2016, 19:00:54 UTC |
4d628f1 | James Hensman | 28 June 2016, 18:53:15 UTC | Small improvements to kernel slicing Kernels with active dimensions now select columns of data using tensorflow's gather, rather that a packed series of scalar indexes. old: tf.pack([X[:,i] for in in active_dims]) new: tf.transpose(tf.gather(tf.transpose(X), active_dims)) the latter is preferred because the graph is much smaller, especially for a large nuber of active dimensions. In addition, the kernel tests have been tidied and extended to cover more of kernels.py. | 28 June 2016, 18:53:15 UTC |
6d99d4c | Mark van der Wilk | 28 June 2016, 18:03:47 UTC | Fixed test. Also works with python3 now. | 28 June 2016, 18:03:47 UTC |
4148cd9 | Mark van der Wilk | 28 June 2016, 17:33:23 UTC | Merge branch 'master' of github.com:GPflow/GPflow into to_dict | 28 June 2016, 17:33:23 UTC |
1cb047c | Alexis Boukouvalas | 28 June 2016, 12:28:50 UTC | Merge pull request #109 from GPflow/multiclass Multiclass | 28 June 2016, 12:28:50 UTC |
abf3015 | alexggmatthews | 28 June 2016, 12:04:17 UTC | Merge branch 'multiclass' of https://github.com/GPflow/GPflow into multiclass | 28 June 2016, 12:04:17 UTC |
e615c3c | alexggmatthews | 28 June 2016, 12:04:09 UTC | Correcting comment. | 28 June 2016, 12:04:09 UTC |
18bddea | Alexis Boukouvalas | 28 June 2016, 10:51:22 UTC | fixed typos Some small typos fixed | 28 June 2016, 10:51:22 UTC |
568ff1a | alexggmatthews | 27 June 2016, 15:11:34 UTC | Building up multiclass demo. | 27 June 2016, 15:11:34 UTC |
1acef30 | alexggmatthews | 27 June 2016, 13:43:33 UTC | Building up test. | 27 June 2016, 13:43:33 UTC |
aeb4de9 | James Hensman | 27 June 2016, 13:34:41 UTC | docstring improvements | 27 June 2016, 13:34:41 UTC |
7f4fae9 | alexggmatthews | 27 June 2016, 13:31:16 UTC | Building up likelihood tests. | 27 June 2016, 13:31:16 UTC |
a7c6473 | James Hensman | 27 June 2016, 13:23:08 UTC | changed the order of the classe and improved docstrings | 27 June 2016, 13:23:08 UTC |
6516d0e | alexggmatthews | 27 June 2016, 13:21:42 UTC | Building up multiclass tests. | 27 June 2016, 13:21:42 UTC |
7a54ddf | alexggmatthews | 27 June 2016, 13:14:26 UTC | Correcting bug. | 27 June 2016, 13:14:26 UTC |
b91294e | alexggmatthews | 27 June 2016, 13:02:21 UTC | Correcting comment. | 27 June 2016, 13:02:21 UTC |
881f905 | alexggmatthews | 27 June 2016, 12:59:10 UTC | Removing partial softmax functionality because it is confusing. | 27 June 2016, 12:59:10 UTC |
194aaa9 | alexggmatthews | 27 June 2016, 12:07:00 UTC | Building up test likelihoods. | 27 June 2016, 12:07:00 UTC |
a208969 | alexggmatthews | 27 June 2016, 12:00:08 UTC | Tidying up likelihood test. | 27 June 2016, 12:00:08 UTC |
eac7334 | alexggmatthews | 27 June 2016, 11:17:34 UTC | Changes to likelihood tests. | 27 June 2016, 11:17:34 UTC |
050bf0d | alexggmatthews | 27 June 2016, 11:16:48 UTC | Correcting an added bug and reimplimenting clipping. | 27 June 2016, 11:16:48 UTC |
1eb9707 | alexggmatthews | 27 June 2016, 11:03:16 UTC | Some additional clipping. | 27 June 2016, 11:03:16 UTC |
f677790 | fujiisoup | 27 June 2016, 10:20:38 UTC | Test added for different size. | 27 June 2016, 10:20:38 UTC |
a4d2d04 | fujiisoup | 27 June 2016, 10:02:28 UTC | Some documentations prepared. | 27 June 2016, 10:02:28 UTC |
2c3e80e | alexggmatthews | 27 June 2016, 09:31:34 UTC | Removing clipping. | 27 June 2016, 09:31:34 UTC |
b808f87 | alexggmatthews | 27 June 2016, 08:45:23 UTC | Removing Multiclass from non-applicable test. | 27 June 2016, 08:45:23 UTC |
8860da0 | alexggmatthews | 27 June 2016, 08:25:19 UTC | Removing Multiclass from non-applicable test. | 27 June 2016, 08:25:19 UTC |
56437b1 | James Hensman | 26 June 2016, 15:54:19 UTC | FIX: correct call to rng.randint | 26 June 2016, 15:54:19 UTC |
511fe85 | Keisuke Fujii | 26 June 2016, 11:23:55 UTC | Merge pull request #110 from GPflow/DataObject Data object | 26 June 2016, 11:23:55 UTC |
9cd3695 | James Hensman | 25 June 2016, 13:23:18 UTC | removing unised update_data functions | 25 June 2016, 13:23:18 UTC |
08f15e6 | James Hensman | 25 June 2016, 13:13:54 UTC | two minor edits to tests to work with new feed_dict structure | 25 June 2016, 13:13:54 UTC |
a61e3f0 | James Hensman | 25 June 2016, 12:41:06 UTC | improved picklability of DataHolder | 25 June 2016, 12:41:06 UTC |
6e34865 | James Hensman | 25 June 2016, 12:38:15 UTC | enabled minibatching in the svgp | 25 June 2016, 12:38:15 UTC |
491b288 | James Hensman | 25 June 2016, 10:15:51 UTC | adding compile functions for shape-changes For VGP and GPMC, when the data shape changes, the parameter shap must too. This is don at compile time. | 25 June 2016, 10:15:51 UTC |