138 commits

Author	SHA1	Message	Date
Max Zwiessele	32491abaff	gradchecker now with fixed inputs	2014-02-10 16:42:00 +00:00
Max Zwiessele	9eef4ebded	fixes added for gradchecking	2014-02-10 16:08:11 +00:00
Max Zwiessele	826d2f04ff	checkgrad now global and callable from any parameter	2014-02-10 16:01:55 +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
James Hensman	6346af8764	assorted fixes	2014-02-06 09:28:27 +00:00
James Hensman	f653bc430e	an afternoon's work on the laplace approximation	2014-02-05 16:23:35 +00:00
James Hensman	d074280213	removed some superfluous things from the model class	2014-01-27 13:04:23 +00:00
Max Zwiessele	cec766b723	getstate > _getstate and setstate > _setstate	2014-01-24 15:48:23 +00:00
Max Zwiessele	e128059377	new gradient handling way nicer	2014-01-24 15:07:28 +00:00
James Hensman	c1a416debc	more gradient based tomfoolery	2014-01-24 14:15:32 +00:00
James Hensman	e94b16e0b3	exact gaussian inference now accepts a kern and X, not K	2014-01-24 11:03:17 +00:00
Max Zwiessele	0733886ba0	all parameterization stuff now in seperate module -> GPy.core.parameterization	2013-12-16 13:45:24 +00:00
Max Zwiessele	6c60eef4bd	biiig changes in tieing, and printing -> hirarchy now always shown	2013-11-11 17:46:33 +00:00
Max Zwiessele	765ab41045	bgplvm integrated	2013-11-07 08:52:33 +00:00
Max Zwiessele	8c02e4af36	added BGPLVM in parameterized	2013-11-06 15:15:13 +00:00
Max Zwiessele	3316d29341	added gplbm and sparse gp to new parameterized structure	2013-11-06 11:40:54 +00:00
Max Zwiessele	6feb5dd2f1	adjusted parameters to report their changes	2013-11-03 13:58:15 +00:00
Max Zwiessele	d3721b76a8	changing all parameterized objects to be compatible with the new parameterization	2013-10-25 15:29:04 +01:00
Max Zwiessele	27724fd456	GPRegression working, gradients still todo	2013-10-22 16:16:54 +01:00
Max Zwiessele	9a70f27010	first adjustments to the model and gps -> updates and gradient transforms	2013-10-22 13:41:01 +01:00
Max Zwiessele	055eb3b70b	starting to sort out likelihoods WARNING not working	2013-10-17 14:39:44 +01:00
Max Zwiessele	3a97eaf684	warning messages optional for re-constraining	2013-10-14 17:59:41 +01:00
Max Zwiessele	2018368d8d	domain and trtansformations namespace prettyfying	2013-10-11 17:26:04 +01:00
James McMurray	aa5ebcbd8f	Fixed more errors in docs 2	2013-09-20 17:46:23 +01:00
Ricardo	c129b98b3b	Merge branch 'devel' of github.com:SheffieldML/GPy into devel ... Conflicts: GPy/examples/classification.py	2013-09-20 15:55:48 +01:00
Ricardo	a51af5b8c4	epsilon and power_ep now are parameters of update_likelihood.	2013-09-20 13:22:38 +01:00
James McMurray	c36a6b341c	Started fixing docs	2013-09-20 11:15:41 +01:00
Ricardo	603aa18482	Merge branch 'devel' of github.com:SheffieldML/GPy into devel ... Conflicts: GPy/examples/regression.py GPy/kern/constructors.py GPy/testing/kernel_tests.py GPy/util/multioutput.py	2013-09-16 12:32:42 +01:00
Ricardo	1931e447f4	print_all function removed, print m works as before.	2013-09-16 12:24:37 +01:00
Neil Lawrence	dffa9541c2	Merged conflict of model.py	2013-09-14 20:05:42 +01:00
Neil Lawrence	00d335444d	Changed kern.py so that X2 is correctly passed as None to the kern parts for dK_dX. Modified several kern parts so that dK_dX is correctly computed (factors of 2 missing). Removed spurious factors of 2 from gplvm, bcgplvm, sparse_gp and fitc code.	2013-09-14 20:02:40 +01:00
Ricardo	f8c9e6b982	Merge branch 'linK_functions2' into devel ... Conflicts: GPy/core/gp.py GPy/core/gp_base.py GPy/core/sparse_gp.py GPy/examples/regression.py GPy/kern/constructors.py GPy/kern/parts/coregionalise.py GPy/models/__init__.py GPy/models/sparse_gp_classification.py GPy/util/__init__.py	2013-09-13 15:57:34 +01:00
James Hensman	0070a73004	Fixed Alans checkgrad bug	2013-09-12 10:20:28 +01:00
Max Zwiessele	57eda0b77c	added print m and print m.all differentiation	2013-09-02 11:53:45 +01:00
Neil Lawrence	c140b2118a	Various doc string edits and consistency changes in comments.	2013-08-19 20:11:25 +02:00
Max Zwiessele	ede5654e3c	input_sensitivity right way	2013-08-02 12:18:43 +01:00
Max Zwiessele	3051c2a7e5	minor/pep8	2013-07-29 15:29:00 +01:00
Ricardo	70c44b2cdd	Multioutput is working	2013-07-18 18:49:26 +01:00
Andreas	f56b66cd7a	input sensitivity for rbf_inv	2013-07-17 21:42:30 +01:00
Max Zwiessele	66a6bde715	more robust gradient clippinggit stat	2013-06-28 10:59:29 +01:00
Max Zwiessele	146dff7cbf	_get/set_params into parameterized	2013-06-26 17:04:10 +01:00
Max Zwiessele	f5effb8cb6	added robust pickling, switches to old behaviour, if get/setstate not implemented	2013-06-26 16:48:48 +01:00
Max Zwiessele	e869fcaf65	pcikling?	2013-06-25 17:43:42 +01:00
Max Zwiessele	5f1445bf04	model robustness greatly improved and avisualize adjustments for plot_latent	2013-06-07 17:11:45 +01:00
Max Zwiessele	73bed5be82	robust failure handling in model objective and gradient	2013-06-07 14:07:21 +01:00
Max Zwiessele	924a319b98	catch linalg errors inside model and more sopihisticated non-pd checks	2013-06-07 13:34:45 +01:00
Ricardo	ab6a87a4d5	pseudo_EM modified	2013-06-05 18:01:28 +01:00
Max Zwiessele	0490861099	kern params adapted: Nparams > num_params and fixes of input_dim	2013-06-05 16:14:30 +01:00
Max Zwiessele	db78b233b8	Nparams > num_params and Nparam_tranformed > num_params_transformed	2013-06-05 15:29:18 +01:00
James Hensman	56a4bc4e21	an assortment of fixes	2013-06-05 14:52:37 +01:00