43 commits

Author	SHA1	Message	Date
James Hensman	dc33aa1b8c	used scipy.weave to improve the speed of rbf grads ... for large number of input dimensions with ARD, wqe have approx tenfold speedup.	2013-05-30 09:29:26 +01:00
James Hensman	83510e7515	first attemot at the new constraint framework	2013-04-30 11:25:57 +01:00
James Hensman	9fb090a508	James and Nicolos massive Yak shaving session	2013-04-26 19:32:33 +01:00
James Hensman	dd27b285d6	rbf computation of psi2 now works if there's only one datum	2013-04-15 17:53:26 +01:00
James Hensman	1d094229df	manual merge in plot_latent	2013-04-12 22:44:11 +01:00
James Hensman	a42ec441e4	some simplification of the psi2_statistics in rbf	2013-04-12 15:02:56 +01:00
Nicolo Fusi	a42d84274d	fixed merge conflict on BPGLVM	2013-04-12 13:35:00 +01:00
Nicolo Fusi	c7e8345c96	--march=native was causing problems on the stupid compiler on MacOS	2013-04-12 13:31:45 +01:00
James Hensman	48b0ac6399	some minor improvements in visualize	2013-04-10 20:02:22 +01:00
Nicolo Fusi	7e0e4f63c4	OMP for psi2 computations in RBF	2013-04-10 16:50:02 +01:00
James Hensman	99ca20b77c	improved weaving	2013-04-10 16:12:09 +01:00
James Hensman	4aca883df3	weaved some rbf code	2013-04-10 15:50:31 +01:00
James Hensman	85810030c5	small efficiency changes in rbf	2013-03-27 15:08:25 +00:00
James Hensman	adfa6de1d8	rbf now works in a more memory friendly fashion	2013-03-27 15:03:46 +00:00
Nicolas	3dd62c8251	Few bugs fixed in the documentation	2013-03-11 18:45:04 +00:00
Nicolas	a012cd64a5	Small changes	2013-03-11 17:25:08 +00:00
Alan Saul	12d6f5056b	removed keyname partial	2013-03-11 12:15:59 +00:00
Nicolo Fusi	de665361a7	all kernels working fine with the psi statistics now	2013-02-15 18:08:40 +00:00
Nicolas	666f334a67	small fixes in the kernel documentation	2013-01-31 17:36:01 +00:00
Nicolas	160568c727	Merge branch 'master' of github.com:SheffieldML/GPy	2013-01-31 17:28:49 +00:00
Nicolo Fusi	b25e123113	changed default ARD setting in linear	2013-01-31 17:24:53 +00:00
Nicolas	a510524620	small changes in the way covariance functions handle lengthscale as input	2013-01-31 17:19:15 +00:00
Nicolo Fusi	9c88687634	doc style change	2013-01-31 10:57:43 +00:00
Nicolo Fusi	8a5f075ef0	RBF (both ARD and non-ARD) kernels working nicely with psi statistics	2013-01-30 16:27:45 +00:00
Nicolo Fusi	2493e2d336	integrated sparse GP regression and BGPLVM classes	2013-01-30 15:51:36 +00:00
Nicolo Fusi	8a3e10700d	massive merge of the debug branch	2013-01-30 15:30:54 +00:00
Nicolo Fusi	936d08723e	BGPLVM working with rbf+white	2013-01-29 14:02:41 +00:00
Nicolo Fusi	8b6e244cf1	decent gradients for most parameters	2013-01-28 15:55:40 +00:00
Nicolas	a76c53e699	bug fixed in example (in regression.py)	2013-01-18 17:43:32 +00:00
Nicolas	ffd05027cb	More tests in unit_tests	2013-01-18 16:14:13 +00:00
Nicolas	69743be33e	last ARD flag changes to kernels	2013-01-18 16:03:20 +00:00
Nicolas	eb86182f7d	solved conflicts for rbf kernel	2013-01-18 14:23:57 +00:00
Nicolas	68d7e23648	rbf kernel now has an ARD flag	2013-01-18 13:58:41 +00:00
Neil Lawrence	3c700dfa50	Changed get_param and set_param to _get_params and _set_params	2013-01-18 12:31:37 +00:00
Neil Lawrence	0ee11a2077	Added datasets.py back in and minor changes.	2012-12-05 19:19:15 -08:00
James Hensman	6d0a7afd4c	bias kernel psi stats implemented.	2012-11-30 15:49:20 +00:00
James Hensman	31b7a0520e	sparseGPLVM demo now working	2012-11-30 11:59:22 +00:00
James Hensman	ba84a43ea3	sparse GP regression now working on this branch	2012-11-30 11:47:35 +00:00
James Hensman	78d1abfc22	GP_regression demo working with new style gradients for rbf, linear, white, bias	2012-11-30 11:04:14 +00:00
James Hensman	5f92ff6785	GPLVM demo working	2012-11-30 10:31:02 +00:00
James Hensman	4b4f1da128	very basic GP_regression demo is working	2012-11-30 09:11:46 +00:00
Nicolo Fusi	92adeecb6c	added copyright notice and license at the top	2012-11-29 16:39:20 +00:00
Nicolo Fusi	63321e8409	kerns	2012-11-29 16:31:48 +00:00