45 commits

Author	SHA1	Message	Date
Max Zwiessele	4543fc3480	[link|unlink_parameter] renaming add_parameter to link_parameter	2014-09-08 08:57:28 +01:00
mzwiessele	d000893878	[ard] enhanced ard handling and plotting ... Conflicts: GPy/kern/_src/linear.py GPy/models/ss_gplvm.py	2014-08-25 09:55:58 -07:00
Zhenwen Dai	b9fdbedf20	update additive kernel for SSGPLVM	2014-08-18 18:14:16 +01:00
Zhenwen Dai	785c580032	merge the current devel into psi2	2014-08-11 18:01:23 +01:00
mzwiessele	93b92111f8	[minor] minor changes	2014-08-05 08:28:03 -07:00
Zhenwen Dai	216de32c0c	fix add kernel and VarDTC_minibatch speed tuning	2014-07-01 18:01:33 +01:00
Zhenwen Dai	129985998c	fix ss_mrd and fix white and bias kernel	2014-07-01 13:16:57 +01:00
Max Zwiessele	1ebcee967a	[psi-stats] add kernel was missing a psi zero call	2014-05-20 14:47:57 +01:00
Zhenwen Dai	3ee2dc920c	a little merge	2014-05-15 12:17:21 +01:00
Max Zwiessele	02b5ee1e46	[parameterized] restructered a lot and finalized some stuff	2014-05-15 11:29:20 +01:00
Max Zwiessele	675d498794	[active_dims] all kernels now have int arrays as active_dims	2014-05-09 14:05:22 +01:00
Zhenwen Dai	06336bf0d4	switch psi2 statistics design	2014-05-06 16:30:17 +01:00
mzwiessele	4e793fe3a0	slicing: slicing now thorughs the right error	2014-04-28 15:14:21 +01:00
Max Zwiessele	1dabf67c93	assertion checks for all kernels	2014-03-28 12:02:34 +00:00
Max Zwiessele	d5a8e5dd3a	right active dims when adding kernels	2014-03-27 13:08:54 +00:00
Max Zwiessele	92bfb3784c	ODE_UY gradient checks now kernel unit	2014-03-21 10:55:34 +00:00
Max Zwiessele	bf6cdf5d31	add kernel adding another add kernel	2014-03-21 10:16:49 +00:00
Max Zwiessele	6637eb7ac8	adding kernels flattening and parameters already in hierarchy	2014-03-18 17:41:08 +00:00
Max Zwiessele	19dc7cecf4	slicing finished with independent outputs	2014-03-17 16:22:16 +00:00
Max Zwiessele	62d594d977	slicing now returns the right shape, when computing derivative wrt X or Z	2014-03-17 15:43:09 +00:00
Max Zwiessele	efcce6d0af	active_dims as extra parameter for kernels, it tells which input dimensions to work on	2014-03-14 09:18:08 +00:00
Max Zwiessele	e9c96632ba	product kernel and combination kernel updates	2014-03-13 11:01:48 +00:00
Max Zwiessele	ecccf0cbbf	add kernel has its own gradients update	2014-03-13 09:09:13 +00:00
Max Zwiessele	603b70bb6e	Merge branch 'params' of github.com:SheffieldML/GPy into input_dims	2014-03-12 12:07:15 +00:00
Max Zwiessele	54239555a1	psi_stat slices for kernels	2014-03-12 12:03:37 +00:00
Max Zwiessele	01f5d789c5	automatic slicing	2014-03-11 16:24:09 +00:00
Max Zwiessele	2d8246d33f	Combination Kernel for add and prod	2014-03-11 10:24:15 +00:00
Max Zwiessele	7e9078b0f9	merged params here	2014-03-10 16:01:32 +00:00
Max Zwiessele	dab35dcbb0	merged params here	2014-03-10 16:00:35 +00:00
James Hensman	73cb20afb0	bugfix	2014-03-10 11:19:57 +00:00
James Hensman	603733c6f7	added update_gradints_diag to the add and base kernels	2014-03-10 11:14:19 +00:00
Max Zwiessele	546d5dfff3	all parameters in memory	2014-03-10 08:21:13 +00:00
Max Zwiessele	db5fd17609	slicing support for kernel input dimension	2014-03-07 16:59:41 +00:00
James Hensman	9867330861	moved plot functionality from add to kern	2014-02-26 14:30:28 +00:00
Max Zwiessele	d29fa56af2	merged variational posterior changes	2014-02-26 08:30:27 +00:00
Max Zwiessele	51dca0fcbc	ard plotting	2014-02-26 08:22:38 +00:00
James Hensman	80acca640f	messing with kernels	2014-02-25 17:15:38 +00:00
James Hensman	4eac0bd6db	started sorting out some tests	2014-02-25 08:58:51 +00:00
James Hensman	da4686dd3c	renaming: posterior_variationa -> variational_posterior	2014-02-24 19:31:13 +00:00
Max Zwiessele	b200b9fa90	input_sensitivity and ard plotting	2014-02-24 14:47:43 +00:00
James Hensman	87ce8fea0b	weird Max related stuff is happening	2014-02-20 14:24:41 +00:00
James Hensman	d636c8c30c	everything is broken	2014-02-20 14:04:16 +00:00
James Hensman	de51ad638a	prod now seems to work for sparse	2014-02-19 22:23:07 +00:00
James Hensman	92d71384b7	deleted kernpart, prod and add seem to work okay.	2014-02-19 17:37:18 +00:00
James Hensman	20f02a80b4	rbf and white seem to work	2014-02-19 15:00:48 +00:00