216 commits

Author	SHA1	Message	Date
Mike Croucher	560950466d	itertools fixes from 2to3	2015-03-01 09:24:22 +00:00
Mike Croucher	ebc0b6e1a5	Resolved merge conflict	2015-02-26 14:37:58 +00:00
mzwiessele	f1e7760403	[verbose optimization] added automatic detection of ipython notebook support, this is experimental	2015-02-26 13:52:20 +00:00
Mike Croucher	c4fb58176d	Exception fixes for Python 3 compat	2015-02-26 13:33:39 +00:00
Mike Croucher	1521b3e260	Convert print to function for Python 3 compatibility. This breaks compatibility for versions of Python < 2.6	2015-02-26 08:11:11 +00:00
Mike Croucher	2ca24a88f5	More relative import fixes for Python 3 compatibility	2015-02-26 07:14:40 +00:00
mzwiessele	924899069e	[optimization] experimental auto detect of ipython notebook	2015-02-24 11:56:36 +00:00
Max Zwiessele	57e941140c	[opt] unified printing of status of optimization	2015-01-15 09:02:45 +00:00
Max Zwiessele	53dcd3f9fa	[objective grads] undid the change, as this would lead to dramatic problems in reloading old models using the pickle module	2015-01-13 13:53:28 +00:00
Max Zwiessele	20eff02061	[objective grads] undid the change, as this would lead to dramatic problems in reloading old models using the pickle module	2015-01-13 13:46:49 +00:00
Max Zwiessele	1c6cef44b6	[progress] show progress of optimization using optimize(itpython_notebook=True)	2015-01-13 11:49:40 +00:00
Max Zwiessele	cd8dd9ab98	[optimization prints] unified printouts for optimizers, added ipython_notebook flag for use in ipython notebooks	2015-01-12 15:43:33 +00:00
Max Zwiessele	e03d6d4f71	[model print] updates now shown in print out	2015-01-12 11:55:44 +00:00
Max Zwiessele	dff577d22b	renamed opimizer methods to unobscure gradients and objective	2015-01-12 11:35:57 +00:00
Max Zwiessele	35a33f94e8	[html print] more table based corrections for html printing	2014-12-05 16:23:50 +00:00
Max Zwiessele	0e0220921f	[html repr] included css styling for html print outs	2014-12-05 11:13:18 +00:00
Max Zwiessele	4e91a012e6	[model] update messages a little nicer	2014-11-28 10:11:58 +00:00
James Hensman	9bc65ac814	more ]#copyrighting	2014-11-21 11:40:50 +00:00
Max Zwiessele	3c642a5600	[MRD] updates and nicer plotting	2014-11-12 15:08:38 +00:00
James Hensman	213e0a7559	fixed or zero size models will now not raise an error when trying to optimize	2014-11-06 14:18:57 +00:00
Zhenwen Dai	1dbe3e34b0	update docstring for checkgrad	2014-11-05 16:42:28 +00:00
Zhenwen Dai	902f8f138b	clean up parallel framework	2014-11-05 14:34:03 +00:00
Zhenwen Dai	9febc73be3	model checkgrad enhancement	2014-11-05 11:58:05 +00:00
Max Zwiessele	3358d06e42	Merge branch 'devel' of github.com:SheffieldML/GPy into devel	2014-10-16 12:43:19 +01:00
Max Zwiessele	93ca35b319	[whitespaces]	2014-10-16 12:43:15 +01:00
Neil Lawrence	8fa890639e	Display of models and params for the notebook.	2014-10-16 04:06:03 +01:00
James Hensman	31478d4d59	improved docsting for optimize	2014-09-17 11:22:31 +01:00
Max Zwiessele	4543fc3480	[link|unlink_parameter] renaming add_parameter to link_parameter	2014-09-08 08:57:28 +01:00
mzwiessele	a358c51829	[printing] added model details for printing	2014-09-05 14:21:34 +01:00
Zhenwen Dai	e46f3b342e	remove the print message in model.checkgrad	2014-09-02 12:33:52 +01:00
Zhenwen Dai	808cfb0501	Allow the default constraint of a Param object to be 'fixed'	2014-09-02 11:52:09 +01:00
Zhenwen Dai	1110cc31e6	some progress for parameter tie	2014-08-29 18:52:53 +01:00
mzwiessele	93b92111f8	[minor] minor changes	2014-08-05 08:28:03 -07:00
James Hensman	51e48f7508	fixed a bug in optimize restarts: it now used optimizer_array	2014-07-16 09:40:44 +01:00
mzwiessele	8a8818aa91	yak shaving and whitespaces	2014-06-30 10:26:17 -07:00
mzwiessele	431f67a8e3	[optimize] bugfix	2014-06-30 09:24:26 -07:00
mzwiessele	94bff66c7b	[optmimize] bugfix	2014-06-20 12:10:21 -07:00
mzwiessele	73d9118043	[whitespaces] & [opt] minor fix of optimizer, when Optimizer is provided (set model to self)	2014-06-20 10:28:32 -07:00
mzwiessele	eaa18e3512	[model] optimizer can now be an optimizer instance, instead of a string	2014-06-19 12:07:58 -07:00
mzwiessele	5a2bc4863b	[optimizer] one copy for the optimizer in optimizer_array, use this instead of _set|get_params_transformed	2014-05-22 11:39:04 +01:00
Max Zwiessele	24a34aee95	[model] exactly two parameter copy in memory	2014-05-21 16:33:44 +01:00
Max Zwiessele	58a05f37b7	[latentfunctioninference] superclass LatentFunctionInference added, which contains a call just before and just after optimization	2014-05-15 14:06:00 +01:00
Max Zwiessele	68a8c3a691	gradient can be zero and two parameter cancellation is caught	2014-03-31 14:09:47 +01:00
Max Zwiessele	59ff2c8d00	gradient can be zero and two parameter cancellation is caught	2014-03-31 14:09:06 +01:00
Max Zwiessele	f3b74fa85f	pickling and caching	2014-03-31 12:45:09 +01:00
James Hensman	a74965b038	moved input_sensitivity to the gp class	2014-03-27 10:08:45 +00:00
Max Zwiessele	3db095338d	objective function seperate from calls for optimizer	2014-03-24 13:32:56 +00:00
Max Zwiessele	29ff406c08	objective_function now standalone and only internal robust optimization loop	2014-03-24 12:41:10 +00:00
Max Zwiessele	3e5e3a099e	checkgrad is zero test	2014-03-14 11:32:08 +00:00
James Hensman	da4303f71b	bugfix for grad_dict	2014-03-14 10:29:14 +00:00