Commit graph

225 commits

Author SHA1 Message Date
Zhenwen Dai
7f39070f07 Merge branch 'params' of github.com:SheffieldML/GPy into params 2014-02-21 17:56:43 +00:00
Zhenwen Dai
99c6a2095f adapt the new interface of the variational posterior distribution. 2014-02-21 17:56:37 +00:00
Max Zwiessele
659643038f parameterized now supports deleting of parameters 2014-02-21 17:53:44 +00:00
Max Zwiessele
8b8ca5544f linear and rbf fix for variational gradients in Z 2014-02-21 12:29:28 +00:00
Max Zwiessele
0c92fca31a linear without caching, derivatives done 2014-02-21 09:14:31 +00:00
Max Zwiessele
52ab456bfe posterior with one covariance per dimension and param gradient fix 2014-02-20 14:09:20 +00:00
Max Zwiessele
c4f6b0dbe7 copy and missing data 2014-02-19 15:32:16 +00:00
James Hensman
0082acad63 minor edits 2014-02-19 10:51:12 +00:00
Max Zwiessele
d0c563ff0a added sparsegp with missing data 2014-02-18 15:52:33 +00:00
James Hensman
619278a1c8 missing sys 2014-02-17 11:06:08 +00:00
Max Zwiessele
345e5b3e7c some updates for params changes and likelihood fixes 2014-02-14 15:05:38 +00:00
Max Zwiessele
f21e8e2394 roundtrip error fixed for likelihood tests 2014-02-14 13:43:43 +00:00
James Hensman
3375606039 lots of F-ordering nonsense. Seems to work though 2014-02-14 11:38:34 +00:00
James Hensman
80a734e153 variouschanges 2014-02-13 13:53:49 +00:00
James Hensman
83d3389644 an ugly hack to work around the 'stickiness' of ObservableArray. TODO: remove this hack 2014-02-13 11:34:23 +00:00
James Hensman
6d30a11ae8 Merge branch 'params' of github.com:SheffieldML/GPy into params 2014-02-13 08:53:19 +00:00
James Hensman
3c68a9688d general bugfixing 2014-02-13 08:53:14 +00:00
Alan Saul
46ce76dee8 Fixed bernoulli likelihood divide by 0 and log of 0 2014-02-12 16:48:57 +00:00
James Hensman
87acda0034 here's fitc 2014-02-12 12:29:52 +00:00
James Hensman
d95da876e0 Merge branch 'params' of github.com:SheffieldML/GPy into params 2014-02-12 11:13:29 +00:00
James Hensman
34dec0ade3 derivatives working in DTC 2014-02-12 11:13:24 +00:00
Alan Saul
b16ea0e1cd Fixed merge conflict 2014-02-12 10:52:12 +00:00
James Hensman
32deb2085d Merge branch 'params' of github.com:SheffieldML/GPy into params 2014-02-11 20:05:41 +00:00
James Hensman
54b7581c39 changes to DTC 2014-02-11 20:05:36 +00:00
Max Zwiessele
f2e04f138d Merge branch 'params' of github.com:SheffieldML/GPy into params 2014-02-11 16:54:43 +00:00
Max Zwiessele
9d262936c6 array core and bgplvm working > changes due to __i<op>__ will now be reported 2014-02-11 16:54:33 +00:00
James Hensman
1b29c0939e renaming dtc again 2014-02-11 16:44:59 +00:00
James Hensman
48c68640ef renaming dtc 2014-02-11 16:30:12 +00:00
James Hensman
448197b026 Merge branch 'params' of github.com:SheffieldML/GPy into params 2014-02-11 16:27:21 +00:00
James Hensman
4d1feb9d9d first draft of DTC 2014-02-11 16:27:08 +00:00
Alan Saul
f7cdf598a6 Merge branch 'params' of github.com:SheffieldML/GPy into params 2014-02-11 15:23:49 +00:00
Max Zwiessele
4cfc13d5fc kernel adding now takes over constraints 2014-02-11 15:23:49 +00:00
Max Zwiessele
b7312a1b99 Merge branch 'params' of github.com:SheffieldML/GPy into params 2014-02-11 14:44:22 +00:00
Max Zwiessele
79aca59a37 parameter handling with default constraints 2014-02-11 14:44:15 +00:00
Alan Saul
835934ff51 Merge branch 'params' of github.com:SheffieldML/GPy into params 2014-02-11 14:06:45 +00:00
Alan Saul
c76f1a4d6d Minor reorganising 2014-02-11 14:06:42 +00:00
James Hensman
103aaebb29 dumb merge conflict in a comment 2014-02-11 13:44:18 +00:00
Max Zwiessele
b944427733 rename dK_dtheta > gradients_X 2014-02-11 12:20:57 +00:00
Alan Saul
b72d319909 Merged parameterized fixing 2014-02-10 15:42:47 +00:00
Alan Saul
6cbf810856 Laplace now appears to be grad checking again 2014-02-10 15:40:06 +00:00
Max Zwiessele
6a068775f5 Merge branch 'params' of github.com:SheffieldML/GPy into params
Conflicts:
	GPy/core/parameterization/param.py
	GPy/core/parameterization/parameter_core.py
	GPy/core/parameterization/parameterized.py
2014-02-10 15:21:09 +00:00
Max Zwiessele
e0c68d5eb3 _highest_parent_ now follows the tree, dK_dX > gradient_X, added update_grads_variational to linear, bgplvm for new framework 2014-02-10 15:12:49 +00:00
Alan Saul
d2a0e4a265 Stupid error, needed to actually USE the gradients in student t... Looks like s2 of rasm's may have an extra -? dW_df == -d2logpdf_df not just d2logpdf_df? 2014-02-10 12:54:49 +00:00
Alan Saul
fc44478ed2 Have most of the likelihood testing working, laplace likelihood parameters need fixing, some of the signs are wrong I believe 2014-02-10 12:29:09 +00:00
Alan Saul
0f263d2ff2 Have most of the likelihood testing working, laplace likelihood parameters need fixing, some of the signs are wrong I believe 2014-02-10 12:28:24 +00:00
Alan Saul
186feb45a1 Fixed likelihood tests for new parameters structure 2014-02-07 15:16:52 +00:00
James Hensman
7d2295d854 some messing with fitc 2014-02-06 18:17:04 +00:00
James Hensman
75241ecf89 predictino working nicely for laplace 2014-02-05 17:52:17 +00:00
James Hensman
80629e00b6 fiddling with plotting 2014-02-05 17:12:52 +00:00
James Hensman
432d9668a6 merged simple conflict£ 2014-02-05 16:26:10 +00:00