From 543a04dc1b3797c9207e5ad35554f41282f668ae Mon Sep 17 00:00:00 2001 From: Max Zwiessele Date: Mon, 3 Oct 2016 12:19:24 +0100 Subject: [PATCH 01/11] chg: fixed naming in variational priors : --- GPy/core/parameterization/variational.py | 17 ++++++++++------- 1 file changed, 10 insertions(+), 7 deletions(-) diff --git a/GPy/core/parameterization/variational.py b/GPy/core/parameterization/variational.py index 84a0b739..1dbd205b 100644 --- a/GPy/core/parameterization/variational.py +++ b/GPy/core/parameterization/variational.py @@ -10,7 +10,7 @@ from .param import Param from paramz.transformations import Logexp, Logistic,__fixed__ class VariationalPrior(Parameterized): - def __init__(self, name='latent space', **kw): + def __init__(self, name='latent prior', **kw): super(VariationalPrior, self).__init__(name=name, **kw) def KL_divergence(self, variational_posterior): @@ -23,6 +23,9 @@ class VariationalPrior(Parameterized): raise NotImplementedError("override this for variational inference of latent space") class NormalPrior(VariationalPrior): + def __init__(self, name='normal_prior', **kw): + super(VariationalPrior, self).__init__(name=name, **kw) + def KL_divergence(self, variational_posterior): var_mean = np.square(variational_posterior.mean).sum() var_S = (variational_posterior.variance - np.log(variational_posterior.variance)).sum() @@ -58,7 +61,7 @@ class SpikeAndSlabPrior(VariationalPrior): pi = self.pi[idx] else: pi = self.pi - + var_mean = np.square(mu)/self.variance var_S = (S/self.variance - np.log(S)) var_gamma = (gamma*np.log(gamma/pi)).sum()+((1-gamma)*np.log((1-gamma)/(1-pi))).sum() @@ -163,12 +166,12 @@ class NormalPosterior(VariationalPosterior): """Compute the KL divergence to another NormalPosterior Object. This only holds, if the two NormalPosterior objects have the same shape, as we do computational tricks for the multivariate normal KL divergence. """ return .5*( - np.sum(self.variance/other.variance) - + ((other.mean-self.mean)**2/other.variance).sum() + np.sum(self.variance/other.variance) + + ((other.mean-self.mean)**2/other.variance).sum() - self.num_data * self.input_dim + np.sum(np.log(other.variance)) - np.sum(np.log(self.variance)) ) - + class SpikeAndSlabPosterior(VariationalPosterior): ''' The SpikeAndSlab distribution for variational approximations. @@ -190,11 +193,11 @@ class SpikeAndSlabPosterior(VariationalPosterior): else: self.gamma = Param("binary_prob",binary_prob,Logistic(1e-10,1.-1e-10)) self.link_parameter(self.gamma) - + def propogate_val(self): if self.group_spike: self.gamma.values[:] = self.gamma_group.values - + def collate_gradient(self): if self.group_spike: self.gamma_group.gradient = self.gamma.gradient.reshape(self.gamma.shape).sum(axis=0) From 73e59571b49a0d31d2b3305ebb784e7b2cf384d6 Mon Sep 17 00:00:00 2001 From: Max Zwiessele Date: Mon, 3 Oct 2016 12:23:44 +0100 Subject: [PATCH 02/11] chg: version update on paramz --- setup.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/setup.py b/setup.py index 43e70815..942cf93c 100644 --- a/setup.py +++ b/setup.py @@ -148,7 +148,7 @@ setup(name = 'GPy', include_package_data = True, py_modules = ['GPy.__init__'], test_suite = 'GPy.testing', - install_requires = ['numpy>=1.7', 'scipy>=0.16', 'six', 'paramz>=0.6.8'], + install_requires = ['numpy>=1.7', 'scipy>=0.16', 'six', 'paramz>=0.6.9'], extras_require = {'docs':['sphinx'], 'optional':['mpi4py', 'ipython>=4.0.0', From 002ee4824af9bdd31a1e16f8767447e5ad2d6b69 Mon Sep 17 00:00:00 2001 From: Max Zwiessele Date: Mon, 3 Oct 2016 12:24:09 +0100 Subject: [PATCH 03/11] =?UTF-8?q?Bump=20version:=201.5.3=20=E2=86=92=201.5?= =?UTF-8?q?.4?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- GPy/__version__.py | 2 +- appveyor.yml | 2 +- setup.cfg | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) diff --git a/GPy/__version__.py b/GPy/__version__.py index a06ff4e0..3a384fb2 100644 --- a/GPy/__version__.py +++ b/GPy/__version__.py @@ -1 +1 @@ -__version__ = "1.5.3" +__version__ = "1.5.4" diff --git a/appveyor.yml b/appveyor.yml index eafd348a..36092f90 100644 --- a/appveyor.yml +++ b/appveyor.yml @@ -3,7 +3,7 @@ environment: secure: 8/ZjXFwtd1S7ixd7PJOpptupKKEDhm2da/q3unabJ00= COVERALLS_REPO_TOKEN: secure: d3Luic/ESkGaWnZrvWZTKrzO+xaVwJWaRCEP0F+K/9DQGPSRZsJ/Du5g3s4XF+tS - gpy_version: 1.5.3 + gpy_version: 1.5.4 matrix: - PYTHON_VERSION: 2.7 MINICONDA: C:\Miniconda-x64 diff --git a/setup.cfg b/setup.cfg index b4028e9f..2ca0f9e6 100644 --- a/setup.cfg +++ b/setup.cfg @@ -1,5 +1,5 @@ [bumpversion] -current_version = 1.5.3 +current_version = 1.5.4 tag = True commit = True From 378ba4956f24466623b02280c2afb4e13153f1f3 Mon Sep 17 00:00:00 2001 From: Max Zwiessele Date: Mon, 3 Oct 2016 12:55:18 +0100 Subject: [PATCH 04/11] chg: pkg: CHANGELOG update --- CHANGELOG.md | 2029 ++++++++++++++++++++++++-------------------------- 1 file changed, 993 insertions(+), 1036 deletions(-) diff --git a/CHANGELOG.md b/CHANGELOG.md index b9c1c51c..4125feb9 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -1,13 +1,63 @@ # Changelog -## v1.5.3 (2016-09-06) +## %%__version__%% (unreleased) + +### New + +* Added deployment pull request instructions for developers. [mzwiessele] + +* Using gitchangelog to keep track of changes and log new features. [mzwiessele] + +### Changes + +* Version update on paramz. [Max Zwiessele] + +* Fixed naming in variational priors : [Max Zwiessele] + +* Changelog update. [mzwiessele] + +### Fix + +* Bug in dataset (in fn download_url) which wrongly interprets the Content-Length meta data, and just takes first character. [Michael T Smith] + +* What's new update fix #425 in changelog. [mzwiessele] ### Other +* Bump version: 1.5.3 → 1.5.4. [Max Zwiessele] + +* Merge pull request #443 from SheffieldML/dataset_download_url_bugfix. [Max Zwiessele] + + fix: Bug in datasets.py + +* Merge branch 'kurtCutajar-devel' into devel. [mzwiessele] + * Bump version: 1.5.2 → 1.5.3. [mzwiessele] * Merge branch 'devel' into kurtCutajar-devel. [mzwiessele] +* Bump version: 1.5.1 → 1.5.2. [mzwiessele] + +* Minor readme changes. [mzwiessele] + +* Bump version: 1.5.0 → 1.5.1. [mzwiessele] + +* Bump version: 1.4.3 → 1.5.0. [mzwiessele] + +* Bump version: 1.4.2 → 1.4.3. [mzwiessele] + +* Bump version: 1.4.1 → 1.4.2. [mzwiessele] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [mzwiessele] + +* [kern] fix #440. [mzwiessele] + +* Bump version: 1.4.0 → 1.4.1. [mzwiessele] + +* [setup] added bumpversion tagging again. [mzwiessele] + +* [mrd] init updates and added tests. [mzwiessele] + * [doc] cleanup. [mzwiessele] * [merge] into new devel. [Max Zwiessele] @@ -40,74 +90,6 @@ * Added core code for GpSSM and GpGrid. [kcutajar] - -## v1.5.2 (2016-09-06) - -### New - -* Added deployment pull request instructions for developers. [mzwiessele] - -### Other - -* Bump version: 1.5.1 → 1.5.2. [mzwiessele] - -* Minor readme changes. [mzwiessele] - - -## v1.5.1 (2016-09-06) - -### Fix - -* What's new update fix #425 in changelog. [mzwiessele] - -### Other - -* Bump version: 1.5.0 → 1.5.1. [mzwiessele] - - -## v1.5.0 (2016-09-06) - -### New - -* Using gitchangelog to keep track of changes and log new features. [mzwiessele] - -### Other - -* Bump version: 1.4.3 → 1.5.0. [mzwiessele] - - -## v1.4.3 (2016-09-06) - -### Changes - -* Changelog update. [mzwiessele] - -### Other - -* Bump version: 1.4.2 → 1.4.3. [mzwiessele] - - -## v1.4.2 (2016-09-06) - -### Other - -* Bump version: 1.4.1 → 1.4.2. [mzwiessele] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [mzwiessele] - -* [kern] fix #440. [mzwiessele] - - -## v1.4.1 (2016-09-06) - -### Other - -* Bump version: 1.4.0 → 1.4.1. [mzwiessele] - -* [setup] added bumpversion tagging again. [mzwiessele] - -* [mrd] init updates and added tests. [mzwiessele] - * Merge pull request #356 from SheffieldML/binomial_laplace. [Max Zwiessele] Binomial laplace #352 @@ -510,6 +492,18 @@ * Merge branch 'deploy' into devel. [Max Zwiessele] +* Merge pull request #368 from SheffieldML/devel. [Max Zwiessele] + + README of pypi now directly in setup + +* Merge pull request #367 from SheffieldML/devel. [Max Zwiessele] + + Update setup to fix problems with slicing + +* Merge pull request #365 from SheffieldML/devel. [Max Zwiessele] + + patch 1.0.5 + * [py3] iterator .next fixes. [Max Zwiessele] * [imports] fix #392. [Max Zwiessele] @@ -578,6 +572,14 @@ * Add eq_ode1 kern and ibp_lfm model. [cdguarnizo] +* Bump version: 1.0.8 → 1.0.9. [mzwiessele] + +* [setxy] was bugged. [mzwiessele] + +* Bump version: 1.0.7 → 1.0.8. [mzwiessele] + +* [examples] dim reduction plotting changes. [Max Zwiessele] + * Integral kernels added, these allow 'histogram' or 'binned' data to be modelled. [Michael T Smith] * Fixed bug, replaced for loops with einsum. [alessandratosi] @@ -596,6 +598,10 @@ * Merge branch devel into gradientsxx. [alessandratosi] +* [fix #380] reloading ep. [mzwiessele] + +* [fix #380] reloading ep. [mzwiessele] + * Fixed bug in kernel_tests for gradients_XX. [alessandratosi] * [slicing] fixed slicing for second order derivatives. [mzwiessele] @@ -604,40 +610,6 @@ * Merge branch 'devel' into gradientsxx. [mzwiessele] -* Bug fix. [alessandratosi] - -* Syntax fix. [alessandratosi] - -* Merge branch 'devel' of https://github.com/SheffieldML/GPy into gradientsxx. [alessandratosi] - -* Modified kernel tests for gradients_XX. [alessandratosi] - -* Added kernel tests for gradients_XX. [alessandratosi] - -* Update function kern.gradients_XX() to compute cross-covariance terms. [alessandratosi] - - -## v1.0.9 (2016-05-11) - -### Other - -* Bump version: 1.0.8 → 1.0.9. [mzwiessele] - -* [setxy] was bugged. [mzwiessele] - - -## v1.0.8 (2016-05-11) - -### Other - -* Bump version: 1.0.7 → 1.0.8. [mzwiessele] - -* [examples] dim reduction plotting changes. [Max Zwiessele] - -* [fix #380] reloading ep. [mzwiessele] - -* [fix #380] reloading ep. [mzwiessele] - * Merge branch 'devel' into kenokabe-devel. [mzwiessele] * [statespace] omg. [mzwiessele] @@ -652,6 +624,12 @@ * [kernel addition] in statespace is bugged for py33 on mac, deactivating it. [mzwiessele] +* Bug fix. [alessandratosi] + +* Syntax fix. [alessandratosi] + +* Merge branch 'devel' of https://github.com/SheffieldML/GPy into gradientsxx. [alessandratosi] + * [statespace] less restrictive test for regular statespace model. [Max Zwiessele] * [travis] condition. [Max Zwiessele] @@ -662,6 +640,12 @@ * Added precomputed kernel class. [vsaase] +* Modified kernel tests for gradients_XX. [alessandratosi] + +* Added kernel tests for gradients_XX. [alessandratosi] + +* Update function kern.gradients_XX() to compute cross-covariance terms. [alessandratosi] + * [readme] added landscape for code cleanines. [Max Zwiessele] * Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] @@ -674,15 +658,6 @@ * [tests] show skipped. [Max Zwiessele] - -## v1.0.7 (2016-04-12) - -### Other - -* Merge pull request #368 from SheffieldML/devel. [Max Zwiessele] - - README of pypi now directly in setup - * [tests] classification tests less strict (sporadic fails) [Max Zwiessele] * [tests] verbose in travis. [Max Zwiessele] @@ -697,31 +672,18 @@ * Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] -* [setup] readme in setup. [Max Zwiessele] - -* []README. [Max Zwiessele] - - -## v1.0.6 (2016-04-11) - -### Other - -* Merge pull request #367 from SheffieldML/devel. [Max Zwiessele] - - Update setup to fix problems with slicing - * Update __version__.py. [Max Zwiessele] * Update setup.cfg. [Max Zwiessele] * Update setup.py. [Max Zwiessele] +* [setup] readme in setup. [Max Zwiessele] + +* []README. [Max Zwiessele] + * [coverage] some more restrictions. [Max Zwiessele] -* Merge pull request #365 from SheffieldML/devel. [Max Zwiessele] - - patch 1.0.5 - * [config] softfail when config cannot be written. [Max Zwiessele] * Bump version: 1.0.4 → 1.0.5. [Max Zwiessele] @@ -2325,43 +2287,38 @@ * Merge branch 'master' into devel. [Max Zwiessele] -* Merge branch 'master' into devel. [Max Zwiessele] - -* Merge branch 'master' into devel. [Max Zwiessele] - -* Merge branch 'master' into devel. [Max Zwiessele] - -* Merge branch 'master' into devel. [Max Zwiessele] - -* Merge branch 'master' into devel. [Max Zwiessele] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] - -* Bug fix for set_XY. [Zhenwen Dai] - -* Add set_XY test. [Zhenwen Dai] - - -## v0.8.8 (2015-09-10) - -### Other - * Bump version: 0.8.7 → 0.8.8. [Max Zwiessele] +* Merge branch 'master' into devel. [Max Zwiessele] + * Bump version: 0.8.6 → 0.8.7. [Max Zwiessele] +* Merge branch 'master' into devel. [Max Zwiessele] + * [py3] print statement. [Max Zwiessele] +* Merge branch 'master' into devel. [Max Zwiessele] + * [pred_var] added predictive variable as property now. [Max Zwiessele] +* Merge branch 'master' into devel. [Max Zwiessele] + * Bump version: 0.8.5 → 0.8.6. [Max Zwiessele] +* Merge branch 'master' into devel. [Max Zwiessele] + * Merge branch 'master' of github.com:SheffieldML/GPy. [Max Zwiessele] * Apply bug fix for set_XY. [Zhenwen Dai] * Converting .md to .rst automatically now. [Max Zwiessele] +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] + +* Bug fix for set_XY. [Zhenwen Dai] + +* Add set_XY test. [Zhenwen Dai] + * Bump version: 0.8.4 → 0.8.5. [Max Zwiessele] * Merge pull request #240 from SheffieldML/devel. [Max Zwiessele] @@ -3744,786 +3701,6 @@ * [huge merge] trying to merge old master and master. [Max Zwiessele] -* Merge pull request #148 from martinsch/normalization_fix. [James Hensman] - -* Normalization: avoid division by zero for constant feature dimensions. [mschiegg] - -* Merge pull request #160 from slinderman/master. [James Hensman] - - Removing set of numpy random seed - - Great spot! We're just about to relase the next iteratino of GPy, we'll make sure it happens there too. - - Thanks. - -* Removing set of numpy random seed. [Scott Linderman] - -* Removed unnecessary spaces from citation. [Nicolo Fusi] - -* Added a way to cite. [Nicolo Fusi] - -* Modified logexp transformation to prevent it returning zero when argument is under -36. [Neil Lawrence] - -* Version change. [Max Zwiessele] - -* Dim reduction examples clearer and init not as much black magic anymore. [Max Zwiessele] - -* Plot_latent now shows selected inputs, even after switching dimensions. [Max Zwiessele] - -* Added hapmap3 as dataset. [Max Zwiessele] - -* Sparse gp stability improved. [Max Zwiessele] - -* HapMap3 dataset added. [Max Zwiessele] - -* Added hapmap download, need to put in data preprocessing for actual usability. [Max Zwiessele] - -* Dim reduction examples. [mzwiessele] - -* Dim reduction examples. [mzwiessele] - -* Bgplvm steepest gradient map update. [mzwiessele] - -* Bgplvm steepest gradient map update. [mzwiessele] - -* Merge branch 'master' of github.com:SheffieldML/GPy. [mzwiessele] - -* Update README.md with funding acknowledgements. [Neil Lawrence] - -* Plotting bug for bgplvm fixed. [mzwiessele] - -* Scg optimizer scale bounds back to 1e-15. [mzwiessele] - -* Parameterized: added warning switch. [mzwiessele] - -* Mrd corrections. [mzwiessele] - -* Version change (early beta, do not change until everythin works. [Max Zwiessele] - -* Fixed the SCG optimizer, thanks to Yarin Gal. [James Hensman] - -* Version now 48. [Max Zwiessele] - -* Version update. [Max Zwiessele] - -* Merge branch 'master' of github.com:SheffieldML/GPy. [Max Zwiessele] - -* Pickling now allways binary as well as protocol -1. [Max Zwiessele] - -* Pickling now allways binary as well as protocol -1. [Max Zwiessele] - -* Using lbfgs algorithm from scipy.minimize, starting to convert all optimizers to minimize format. [Max Zwiessele] - -* Windows -.- [Max Zwiessele] - -* Versions update. [Max Zwiessele] - -* Image is a PIL requirement and should only be imported when actually using it. [Max Zwiessele] - -* Plot handling greatly improved for latent space visualizations. [Max Zwiessele] - -* Version file added. [Max Zwiessele] - -* Fixed Ctrl-C behaviour on Windows. [Nicolo Fusi] - -* Merge branch 'master' of github.com:SheffieldML/GPy. [Nicolo Fusi] - -* Fixed come path issues in sympykern. [James Hensman] - -* Rename and redoc. [Alan Saul] - -* Moving imports, attempting to update RTD. [Alan Saul] - -* Changed initalise_latent to take lower of init=PCA and corrected import. [Alan Saul] - -* Change order of imports for RTD. [Alan Saul] - -* Merge branch 'devel' [Alan Saul] - -* Merge branch 'master' into devel. [Alan Saul] - -* Removed variational.py. [Alan Saul] - -* Fixed plot_latent failure. [Alan Saul] - -* Ignore example tests. [Alan Saul] - -* Removed yes pipe for travis. [Alan Saul] - -* Seems to handle without answering now. [Alan Saul] - -* Reverse travis to see what it asks for. [Alan Saul] - -* Fixed some tests. [Alan Saul] - -* Merged and fixed conflict in ODE_UY.py. [Alan Saul] - -* Merge completed. [Max Zwiessele] - -* Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [mu] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Andreas] - -* Small changes in svigp. [Andreas] - -* ODE UY dkdtheta. [mu] - -* UY dkdtheta. [mu] - -* UY dkdtheta. [mu] - -* BGPLVM with missing data. [Max Zwiessele] - -* Pca adjustements to lvm models. [Max Zwiessele] - -* Ppca added, ppca missing data not working yet. [Max Zwiessele] - -* Diagonal operations. [Max Zwiessele] - -* Subarray indexing. [Max Zwiessele] - -* Documenting. [Max Zwiessele] - -* Bug in ODE_UY fix. [Alan Saul] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Alan Saul] - -* Fixed the numerical quadrature, won't work with large f unless normalized. [Alan Saul] - -* Fixed lots of breaking tests, reduced step size for checkgrad to 1e-4 in tests (perhaps this should be global), added some missing attributes to data_resources.json. [Alan Saul] - -* Set warnings for truncated hessian, it has been noted that that by truncating we can have incorrect posteriors, though at convergence this should not be a problem, could be fixed by not using Cholsky as the decomposition as it cannot handle non-positive definite mats. [Alan Saul] - -* Merge pull request #90 from jamesmcm/master. [Alan Saul] - - Fixing ReadTheDocs reading docstrings, adding data_resources.json - -* Adding data_resources.json to setup data files. [James McMurray] - -* Testing modification for ReadTheDocs to stop docstring errors. [James McMurray] - -* Ensure_defaiult constraints in svigp. [James Hensman] - -* Fixed Ctrl-C behaviour on Windows. [Nicolo Fusi] - -* Removed print statements from config parser, commented out ODE kerns. [Nicolo Fusi] - -* Merge branch 'devel' [Nicolo Fusi] - -* Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [mu] - -* Ode UY. [mu] - -* Dk dparameter. [mu] - -* ODE_UY. [mu] - -* Merge branch 'devel' [Nicolo Fusi] - -* Added some more error checking for downloading datasets. [Neil Lawrence] - -* Fixed some bugs in mocap.py where errors weren't being raised when file format was incorrect and made datasets.py check for 404 errors which previously were occuring silently ... shhhhh. [Neil Lawrence] - -* Fixed bugs in cmu_mocap loader where cmu_url was missing and loading in mocap data twice in same session led to incorrect url through copy error. [Neil Lawrence] - -* Fixed examples tests, started changing datasets code which has a few bugs. [Alan Saul] - -* Minor changes to naming of signitures. [Alan Saul] - -* Changed more examples to accept optimize and plot. [Alan Saul] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Alan Saul] - -* Added comments for models module and adjusted setup. [Max Zwiessele] - -* Merge branch 'naniny' into devel. [Max Zwiessele] - -* Rename _models to models_modules to include in doc. [Max Zwiessele] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] - -* Some tidying in the regression examples. [James Hensman] - -* Added optimize and plot for classification, non_gaussian and stochastic examples. [Alan Saul] - -* Fixed gp_base and svigp for sampling (doesn't use it but needs the arguments) [Alan Saul] - -* Added constant to Z_tilde, now log likelihoods are equal! [Alan Saul] - -* Changed some parameters of the laplace, tidied up examples. [Alan Saul] - -* Dimensionality reduction examples updated with optimize, plot and verbose. [Max Zwiessele] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] - -* Changing the seed seems to fix Alan's bug. [James Hensman] - -* Fixed symmetry in checkgrad issue. [James Hensman] - -* Reverted the brent optimisation in laplace. [James Hensman] - - (For the 1D linesearch using Brent) - -* Improved detectino of sympy. [James Hensman] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] - -* Fixed exception handling bug in GPy/util/linalg.py:29. [Teo de Campos] - -* Argghdfklg. [James Hensman] - -* Better warings for cathcing of blaslib detection. [James Hensman] - -* Changeing models to _models in setup.py. [James Hensman] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Alan Saul] - -* Fixed step size for likelihood tests and allowed randomizing of laplace. [Alan Saul] - -* Fixed student_t approximation demo and changed convergence critera to difference of f. [Alan Saul] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Alan Saul] - -* Added cfg file to manfiest and package_data. [Alan Saul] - -* Dimensionality reduction example (oil) updated. [Max Zwiessele] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] - -* Removed ipdb statement from kern, cleaned up some nasty whitespace. [James Hensman] - -* More readme. [James Hensman] - -* More readme stuff. [James Hensman] - -* More readme edits. [James Hensman] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] - -* Minor edits to the README. [James Hensman] - -* Gradient checker comments and import updates. [Max Zwiessele] - -* Gradientchecker added as a model. [Max Zwiessele] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] - -* Push minor fix to eq_sympy kernel test. [Neil Lawrence] - -* Added some tips to the readme. [James Hensman] - -* Fixed import errors in tests. [Max Zwiessele] - -* Sympykern kern_tests now passing, code is inefficient but should be numerically stable. [Neil Lawrence] - -* Modified to improve part of stability, gradient checks still passing. [Neil Lawrence] - -* Added gradient of sympy kernel, seems to pass tests, but know it's not numerically stable. Checking in before making numerically stable. [Neil Lawrence] - -* Fixed test in kern.py to request correct output dim for multioutput covariances. [Neil Lawrence] - -* ODE_UY. [mu] - -* Fixing up the blas detectino in linalg. [James Hensman] - -* Removing ipdb statements. [James Hensman] - -* Lots of medding with the likelihoods to get the tests working. the tests still don;t work. [James Hensman] - -* Formatting docstring. [James Hensman] - -* Changed nasty whitespace. [James Hensman] - -* Eq_ode1 working but test failing? [Neil Lawrence] - -* Added eq_ode1 to constructors.py. [Neil Lawrence] - -* Fixed bug in sympy kernel and added sympolic.py back into utils __init__.py. [Neil Lawrence] - -* Merge with James's changes. [Neil Lawrence] - -* Merge. [James Hensman] - -* Removed some sympy stuff. [James Hensman] - -* Merge with James's changes. [Neil Lawrence] - -* Committing change for master check out. [Neil Lawrence] - -* Skipping crossterm tests instead of expected failure. [Max Zwiessele] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] - -* Added a path for the data resources. not all users will be working in the GPy directory. [James Hensman] - -* Moved data resource information to a json file. [Neil Lawrence] - -* Bug fix for single output sympy kernel. [Neil Lawrence] - -* Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [Neil Lawrence] - -* Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [mu] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Nicolo Fusi] - -* Fixed problem in warping. [Nicolo Fusi] - -* Constructor and init for ODE_UY. [mu] - -* Working eq_ode1 in sympy now. [Neil Lawrence] - -* Dim reduction imports. [Max Zwiessele] - -* Testing imports update and expected failure for crossterms. [Max Zwiessele] - -* Rename models to _models and import models in models.py. [Max Zwiessele] - -* Psistattests update. [Max Zwiessele] - -* Updated crossterms, rbf x any not working yet (derivatives) [Max Zwiessele] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] - -* Weird merge. [James Hensman] - -* Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [Neil Lawrence] - -* Part implementation of ode_eq functionality. Not yet numerically stable or efficient (some horrible use of cut and paste to get things working ...) [Neil Lawrence] - -* A trial namespace renaming. [James Hensman] - -* Better handling of missing config files. [James Hensman] - -* Debugging the config paths. [James Hensman] - -* Allowing the passing of 1D X to a GP. with warning of course. [James Hensman] - -* More fiddling with the windows path for config. [James Hensman] - - Where is the windows guru? out playing beach volley? - -* Changed how we search for config files on windows. [James Hensman] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] - -* Fixed up symmetric kern. [James Hensman] - -* Half way through crossterm objective. [Max Zwiessele] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] - -* Added block matrix utility. [James Hensman] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] - -* Added **likelihood_params to predictive_values. [Ricardo] - -* Changes in plot function: sampling vs numerical approximation. [Ricardo] - -* Adding docstring for symmetric kern. [James Hensman] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] - -* Bug fixed in numerical approx. to the predictive variance. [Ricardo] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Ricardo] - -* Numerical predictions fixed, sampling predictions are not working. [Ricardo] - -* Predictive_mean and predictive_variance now use gp_var as a parameter, rather than gp_std. [Ricardo] - -* Fixed product kern get and set state. [James Hensman] - -* Added getstate/setstate for product kernel. [James Hensman] - -* In the middle of crossterms. [Max Zwiessele] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] - -* 2D plots fixed. [Ricardo] - -* Passing **noise_args into predictive_values. [Ricardo] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Ricardo] - -* Missing term in the likelihood. [Ricardo] - -* Reverted broken kern. [Max Zwiessele] - -* Added variational distribution for latent space. [Max Zwiessele] - -* BGPLVM test for crossterms. [Max Zwiessele] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] - -* Even more data plotting. [James Hensman] - -* Fixed plotting isue with plot_f. [James Hensman] - -* Fixed the dpotrs use.. [Alan Saul] - -* Added dpotrs instead of cho_solve. [Alan Saul] - -* Removed ipython dependency from kern. [Alan Saul] - -* Sped up sampling a lot for student t, bernoulli and poisson, added sampling for gaussian and exponential (untested) [Alan Saul] - -* Merge remote-tracking branch 'gpy_real/devel' [Alan Saul] - -* Ignoring examples tests again. [Alan Saul] - -* Added sampling to student_t noise distribution, very slow and is possible to speed up. predictive mean analytical and variance need checking. [Alan Saul] - -* Tidying up and fixed objective being vector. [Alan Saul] - -* Added sampling for predictive quantiles and also mean and variance where necessary. [Alan Saul] - -* Rederived gamma distribution. [Alan Saul] - -* Added derivatives for poisson and a couple of examples, need to fix for EP. [Alan Saul] - -* Merged with devel. [Alan Saul] - -* Reimplemented gradients for exponential, seems to work for laplace now, needs a visual test though. [Alan Saul] - -* Was a bug in the examples_tests.py, fixed and added brendan faces to ignore list. [Alan Saul] - -* Minor clean up. [Alan Saul] - -* Removed unnecessary laplace examples. [Alan Saul] - -* Updated laplace example to use predictive density aswell as RMSE. [Alan Saul] - -* Added log predictive density, ln p(y*|D) [Alan Saul] - -* Merge remote-tracking branch 'gpy_real/devel' [Alan Saul] - -* Updated boston tests (more folds, allow a bias as the datasets are not normalized once split) and more folds. Tweaked some laplace line search parameters, added basis tests for ep. [Alan Saul] - -* Fixed bug in gradient checker where it worked differently given a integer parameter to a float. [Alan Saul] - -* Removed derivatives of variance wrt gp and derivatives of means with respect to gp from noise models. [Alan Saul] - -* Tore out code no longer used from noise_distributions due to rewriting using quadrature. [Alan Saul] - -* Added numerical mean and variance with quadrature, about to clean up. [Alan Saul] - -* Merge branch 'master' into merge_branch. [Alan Saul] - -* Changed the gradients (perhaps for the worse) [Alan Saul] - -* A few typos. [Alan Saul] - -* Gaussian likelihood errors, still not working. [Alan Saul] - -* Added gaussian checker and gaussian likelihood, not checkgrading yet. [Alan Saul] - -* Started adding gaussian sanity checker. [Alan Saul] - -* Got rid of some overdoing the approximation. [Alan Saul] - -* Started adding gaussian likelihood, changed round preloading old_a. [Alan Saul] - -* Trying to debug kernel parameters learning (fails even when noise fixed) may be some instablility, seems like it can get it if it starts close. [Alan Saul] - -* Fixed 2*variance plotting instead of 2*std plotting, tidied up. [Alan Saul] - -* Changed incorrect naming. [Alan Saul] - -* Reparameratised in terms of sigma2. [Alan Saul] - -* Playing trying to find what makes it want to go so low. [Alan Saul] - -* Fixed bug where B wasn't refering to current f location. [Alan Saul] - -* Everything seems to be gradchecking again. [Alan Saul] - -* Added minimizer for finding f, doesn't help. [Alan Saul] - -* Now checkgrads a lot more of the time, but still fails in optimisation, seems also odd that when parameter is fixed kernel parameters go to infinity. [Alan Saul] - -* Added another optimisation which doesn't use gradients. Seems like F is almost always found, but Y can be off, suggesting that Wi__Ki_W is wrong, maybe W? [Alan Saul] - -* Trying to fix optimisation problem, fixed a few bugs but still fails at very low noise. [Alan Saul] - -* Starting to fiddle with mode finding code. [Alan Saul] - -* Fixed a sign wrong, now gradchecks weirdly only above certain points. [Alan Saul] - -* Now gradchecks everytime but student_t fit is bad, noise is underestimated by a long way. [Alan Saul] - -* Checkgrads with explicit and implicit components half the time. [Alan Saul] - -* About to input new derivations for Z's... [Alan Saul] - -* Took out all the asserts and using pure broadcasting method of diagonal now. [Alan Saul] - -* Made it use the fact that W is diagonal and put assertions in to ensure that the results are the same. [Alan Saul] - -* Broken it by getting rid of squeeze, but now working on making it faster using proper vector multiplciation for diagonals. [Alan Saul] - -* Made more numerically stable in a hope that it will work and I will find a bug... [Alan Saul] - -* Lots of name changing and went through all likelihood gradients again. [Alan Saul] - -* Ripped out all things Laplace parameter estimation, starting again with new tactic. [Alan Saul] - -* About to rip out old chain rule method of learning gradients. [Alan Saul] - -* Trying to fix dL_dytil gradient. [Alan Saul] - -* Changed name. [Alan Saul] - -* FIXED DYTIL_DFHAT. [Alan Saul] - -* Workong on doing explicit gradients. [Alan Saul] - -* Gradients almost there for dytil_dfhat, diagonal terms are right. [Alan Saul] - -* Merged with devel. [Alan Saul] - -* Still getting closer to grads for likelihood. [Alan Saul] - -* Almost have likelihood gradients working but kernels still way off. [Alan Saul] - -* Working on putting callback to update laplace in callback. [Alan Saul] - -* Seem to have gradients much closer now. [Alan Saul] - -* Scale and switch KW+I. [Alan Saul] - -* Merged with upstream. [Alan Saul] - -* Added a debug examples. [Alan Saul] - -* Merging. [Alan Saul] - -* Merged with master. [Alan Saul] - -* Plotting problematic kernel. [Alan Saul] - -* Adding gradients, shapes starting to make sense. [Alan Saul] - -* Attempted to introduce gradient methods, won't work yet I doubt. [Alan Saul] - -* Merge remote-tracking branch 'upstream/devel' [Alan Saul] - -* Merged likelihood functions. [Alan Saul] - -* Should be working now, needed to change relative path names. [Alan Saul] - -* Merge branch 'merge_trial' [Alan Saul] - -* Merge branch 'old_repo' into merge_trial. [Alan Saul] - -* Make directory structure match that of GPy. [Alan Saul] - -* Tidy up comments. [Alan Saul] - -* Merged in the GPy upstream. [Alan Saul] - -* Merged in branch which had old_repo merged in. [Alan Saul] - -* Adding weibull likelihood, requires 'extra_data' to be passed to likelihood, i.e. the censoring information. [Alan Saul] - -* Fixed the z scalings. [Alan Saul] - -* Fixed laplace approximation and made more numerically stable with cholesky decompositions, and commented. [Alan Saul] - -* Stabalised most of the algorithm (apart from the end inversion which is impossible) [Alan Saul] - -* Added timing and realised mdot can be faster as its almost always a diagonal matrix its multiplying with. [Alan Saul] - -* Got the mode finding without computing Ki. [Alan Saul] - -* Fixed broadcasting bug, rasm now appears to work. [Alan Saul] - -* Still working on rasmussen, link function needs vectorizing I think. [Alan Saul] - -* Tidying up. [Alan Saul] - -* Added predicted values for student t, works well. [Alan Saul] - -* Working laplace, just needs predictive values. [Alan Saul] - -* Seemed to be working, now its not. [Alan Saul] - -* Changing definitions again... [Alan Saul] - -* Worked out in terms of W, needs gradients implementing. [Alan Saul] - -* Just breaking some things... [Alan Saul] - -* Following naming convention better, lots of inverses which should be able to get rid of one or two, unsure if it works. [Alan Saul] - -* Got an idea of how to implement! written in docs. [Alan Saul] - -* Trying to 'debug' [Alan Saul] - -* Got most of laplace approximation working. [Alan Saul] - -* Added some comments. [Alan Saul] - -* Initial commit, setting up the laplace approximation for a student t. [Alan Saul] - -* Initial commit. [James Hensman] - -* Use bfgs for laplace instead. [Alan Saul] - -* Moved transf_data to make data -1 or 1 from 0 or 1 for bernoulli with probit into the analytical moment match (but it 10% slower), needs removing from epmixednoise. [Alan Saul] - -* Changed naming from old derivatives of likelihoods to new ones in noise distributions. [Alan Saul] - -* Fixed breakage of dvar, tidied up to make more efficient. [Alan Saul] - -* Doc stringing. [Alan Saul] - -* Added quadrature numerical moment matching (but not predictive yet) [Alan Saul] - -* Fixed a few laplace bits. [Alan Saul] - -* Refactored gradients wrt parameters slightly, need to future proof against _get_param_names() disappearing. [Alan Saul] - -* Added more options to generic tests (constraining link function values as bernoulli requies R^{0,1}) and implemented new gradients for bernoulli. [Alan Saul] - -* Rename Binomial to Bernoulli (maybe generalise it with the constant later, but tilted distribution may change) [Alan Saul] - -* Added pdf_link's for gaussian and student t, added third derivatives for transformations and tests for them. [Alan Saul] - -* Renamed laplace_tests to likelihoods_tests. [Alan Saul] - -* Tidying up laplace_tests.py. [Alan Saul] - -* Fixed some bugs, added third derivative for log transformation, and did some doccing. [Alan Saul] - -* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] - -* Still tidying up, laplace now working again, gaussian and student_t likelihoods now done. [Alan Saul] - -* Started on chaining, must remember to chain _laplace_gradients aswell! [Alan Saul] - -* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] - -* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] - -* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] - -* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] - -* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] - -* Finished tearing gaussian noise down, time for student t. [Alan Saul] - -* Beginning to merge lik_functions and derivatives with richardos. [Alan Saul] - -* Docs. [Alan Saul] - -* Removed fit as it is unused. [Alan Saul] - -* More doc strings. [Alan Saul] - -* Doccing and testing for D dimensional input (not multiple dimensional Y yet) [Alan Saul] - -* Tidying up a lot, works for 1D, need to check for more dimensions. [Alan Saul] - -* Tidied up laplace. [Alan Saul] - -* Integrated Laplace and merged Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] - - Conflicts: - GPy/core/gp.py - GPy/likelihoods/__init__.py - GPy/likelihoods/likelihood_functions.py - GPy/likelihoods/link_functions.py - -* Fixed white variance. [Alan Saul] - -* Boston housing works (apart from variance of student t is not valid below 2) [Alan Saul] - -* Tests setup but not fitting properly yet. [Alan Saul] - -* Changed the examples (started boston data) and increased tolerance of finding fhat. [Alan Saul] - -* Added some stability and tidied up. [Alan Saul] - -* Tidying up. [Alan Saul] - -* Student t likelihood function checkgrads (summed gradients wrt to sigma2), maybe some numerical instability in laplace. [Alan Saul] - -* Now checkgrads for gaussian, and ALMOST for student t. [Alan Saul] - -* All gradients now gradcheck. [Alan Saul] - -* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] - -* Refactored tests. [Alan Saul] - -* Tidied up grad checking. [Alan Saul] - -* Added tests and fixed some naming. [Alan Saul] - -* Modified gradient_checker to allow for variable 'f' [Alan Saul] - -* Renamed some things, made some small (incorrect) gradient changes, generalised the gp regression for any likelihood, and added a place holder link function waiting for Richardos changes. [Alan Saul] - -* Removed unneeded dependency. [Alan Saul] - -* Merged GP models. [Alan Saul] - -* Dragged likelihood_function changes in. [Alan Saul] - -* Checked out relavent files. [Alan Saul] - -* Merged in real gpy. [Alan Saul] - -* Empty branch. [Alan Saul] - -* SPELLAFSDIUN. [Max Zwiessele] - -* Fixed up plotting in sparse_gp also. [James Hensman] - -* Fixed up the plotting. [James Hensman] - -* Fixed up plot in GP_base. [James Hensman] - -* Started changing the plotting in examples to remove plot_single_output. [James Hensman] - -* General tidying in models. [James Hensman] - -* Improved docstrings in svigp. [James Hensman] - -* Some tidying in gp.py. [James Hensman] - -* Docstrings and removal of duplicated plotting code in gp_base. [James Hensman] - -* Turned omp off by default as discussed. [James Hensman] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] - -* Added configuration file. [Nicolò Fusi] - - this was done to solve the OpenMP problem on Windows/mac, but I think it - is useful in general. All unit tests pass except the sympy kern ones. - -* Added xw_pen data. [Neil Lawrence] - -* Added xw_pen data. [Neil Lawrence] - -* Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [Neil Lawrence] - -* Basic sim code functional. [Neil Lawrence] - -* Minor change in tutorial. [mu] - -* Added link to user mailing list. [James Hensman] - - -## v0.6.0 (2014-11-21) - -### Other - * Small random perturbations in kernel tests helps with the symmetry gradcheck bug. [James Hensman] * More cooooooopyrighting. [James Hensman] @@ -7398,8 +6575,841 @@ * Merged dimen reduction. [Max Zwiessele] +* Domain and trtansformations namespace prettyfying. [Max Zwiessele] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into params. [Max Zwiessele] + +* Tests updated. [Max Zwiessele] + +* Dim reduction examples Q= > input_dim= [Max Zwiessele] + +* Numpy non hashable AHHHHHH. [Max Zwiessele] + +* Dont print brackets in transformations. [Max Zwiessele] + +* Parameterized first beta test. [Max Zwiessele] + +* Merge branch 'devel' into params. [Max Zwiessele] + +* Cross term testing switched on. [Max Zwiessele] + +* Comment linear x linear for testing purposes. [Max Zwiessele] + +* Parameter_testing. [Max Zwiessele] + +* Index operations now work on flattened indices. [Max Zwiessele] + +* Unoptimized parameter, still slower than current implementation. [Max Zwiessele] + +* Ndarray subclass cleaned up. [Max Zwiessele] + +* Subclassing ndarray almost functional. [Max Zwiessele] + +* Parameters ndarray, stuck at using strides for transformations. [Max Zwiessele] + +* Transformations are singletons now, weak refs for memory managment. [Max Zwiessele] + +* Almost there with array inheriting. [Max Zwiessele] + +* With subclassing ndarray, current_slice problems not solved... [Max Zwiessele] + +* Without inheriting from numpy.ndarray. ndarray functionality missing. [Max Zwiessele] + +* NegativeLogexp Pep8ted. [Max Zwiessele] + +* Index operations finalized. [Max Zwiessele] + +* Added index_operations and deleted them from paramter. [Max Zwießele] + +* Added parameter files - Alan. [Max Zwiessele] + +* Transformations singleton. [Max Zwiessele] + +* Merge branch 'devel' into params. [Max Zwiessele] + + Conflicts: + GPy/core/transformations.py + GPy/kern/parts/kernpart.py + +* Parameterization changes take a while. [Max Zwiessele] + +* Correcterd minor errors (imports etc) [Max Zwiessele] + +* Pep8'ed transformations module. [Max Zwiessele] + +* Merge pull request #148 from martinsch/normalization_fix. [James Hensman] + +* Normalization: avoid division by zero for constant feature dimensions. [mschiegg] + +* Merge pull request #160 from slinderman/master. [James Hensman] + + Removing set of numpy random seed + + Great spot! We're just about to relase the next iteratino of GPy, we'll make sure it happens there too. + + Thanks. + +* Removing set of numpy random seed. [Scott Linderman] + +* Removed unnecessary spaces from citation. [Nicolo Fusi] + +* Added a way to cite. [Nicolo Fusi] + +* Modified logexp transformation to prevent it returning zero when argument is under -36. [Neil Lawrence] + +* Version change. [Max Zwiessele] + +* Dim reduction examples clearer and init not as much black magic anymore. [Max Zwiessele] + +* Plot_latent now shows selected inputs, even after switching dimensions. [Max Zwiessele] + +* Added hapmap3 as dataset. [Max Zwiessele] + +* Sparse gp stability improved. [Max Zwiessele] + +* HapMap3 dataset added. [Max Zwiessele] + +* Added hapmap download, need to put in data preprocessing for actual usability. [Max Zwiessele] + +* Dim reduction examples. [mzwiessele] + +* Dim reduction examples. [mzwiessele] + +* Bgplvm steepest gradient map update. [mzwiessele] + +* Bgplvm steepest gradient map update. [mzwiessele] + +* Merge branch 'master' of github.com:SheffieldML/GPy. [mzwiessele] + +* Update README.md with funding acknowledgements. [Neil Lawrence] + +* Plotting bug for bgplvm fixed. [mzwiessele] + +* Scg optimizer scale bounds back to 1e-15. [mzwiessele] + +* Parameterized: added warning switch. [mzwiessele] + +* Mrd corrections. [mzwiessele] + +* Version change (early beta, do not change until everythin works. [Max Zwiessele] + +* Fixed the SCG optimizer, thanks to Yarin Gal. [James Hensman] + +* Version now 48. [Max Zwiessele] + +* Version update. [Max Zwiessele] + +* Merge branch 'master' of github.com:SheffieldML/GPy. [Max Zwiessele] + +* Pickling now allways binary as well as protocol -1. [Max Zwiessele] + +* Pickling now allways binary as well as protocol -1. [Max Zwiessele] + +* Using lbfgs algorithm from scipy.minimize, starting to convert all optimizers to minimize format. [Max Zwiessele] + +* Windows -.- [Max Zwiessele] + +* Versions update. [Max Zwiessele] + +* Image is a PIL requirement and should only be imported when actually using it. [Max Zwiessele] + +* Plot handling greatly improved for latent space visualizations. [Max Zwiessele] + +* Version file added. [Max Zwiessele] + +* Fixed Ctrl-C behaviour on Windows. [Nicolo Fusi] + +* Merge branch 'master' of github.com:SheffieldML/GPy. [Nicolo Fusi] + +* Fixed come path issues in sympykern. [James Hensman] + +* Rename and redoc. [Alan Saul] + +* Moving imports, attempting to update RTD. [Alan Saul] + +* Changed initalise_latent to take lower of init=PCA and corrected import. [Alan Saul] + +* Change order of imports for RTD. [Alan Saul] + +* Merge branch 'devel' [Alan Saul] + +* Merge branch 'master' into devel. [Alan Saul] + +* Removed variational.py. [Alan Saul] + +* Fixed plot_latent failure. [Alan Saul] + +* Ignore example tests. [Alan Saul] + +* Removed yes pipe for travis. [Alan Saul] + +* Seems to handle without answering now. [Alan Saul] + +* Reverse travis to see what it asks for. [Alan Saul] + +* Fixed some tests. [Alan Saul] + +* Merged and fixed conflict in ODE_UY.py. [Alan Saul] + +* Merge completed. [Max Zwiessele] + +* Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [mu] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Andreas] + +* Small changes in svigp. [Andreas] + +* ODE UY dkdtheta. [mu] + +* UY dkdtheta. [mu] + +* UY dkdtheta. [mu] + +* BGPLVM with missing data. [Max Zwiessele] + +* Pca adjustements to lvm models. [Max Zwiessele] + +* Ppca added, ppca missing data not working yet. [Max Zwiessele] + +* Diagonal operations. [Max Zwiessele] + +* Subarray indexing. [Max Zwiessele] + +* Documenting. [Max Zwiessele] + +* Bug in ODE_UY fix. [Alan Saul] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Alan Saul] + +* Fixed the numerical quadrature, won't work with large f unless normalized. [Alan Saul] + +* Fixed lots of breaking tests, reduced step size for checkgrad to 1e-4 in tests (perhaps this should be global), added some missing attributes to data_resources.json. [Alan Saul] + +* Set warnings for truncated hessian, it has been noted that that by truncating we can have incorrect posteriors, though at convergence this should not be a problem, could be fixed by not using Cholsky as the decomposition as it cannot handle non-positive definite mats. [Alan Saul] + +* Merge pull request #90 from jamesmcm/master. [Alan Saul] + + Fixing ReadTheDocs reading docstrings, adding data_resources.json + +* Adding data_resources.json to setup data files. [James McMurray] + +* Testing modification for ReadTheDocs to stop docstring errors. [James McMurray] + +* Ensure_defaiult constraints in svigp. [James Hensman] + +* Fixed Ctrl-C behaviour on Windows. [Nicolo Fusi] + +* Removed print statements from config parser, commented out ODE kerns. [Nicolo Fusi] + +* Merge branch 'devel' [Nicolo Fusi] + +* Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [mu] + +* Ode UY. [mu] + +* Dk dparameter. [mu] + +* ODE_UY. [mu] + +* Merge branch 'devel' [Nicolo Fusi] + +* Added some more error checking for downloading datasets. [Neil Lawrence] + +* Fixed some bugs in mocap.py where errors weren't being raised when file format was incorrect and made datasets.py check for 404 errors which previously were occuring silently ... shhhhh. [Neil Lawrence] + +* Fixed bugs in cmu_mocap loader where cmu_url was missing and loading in mocap data twice in same session led to incorrect url through copy error. [Neil Lawrence] + +* Fixed examples tests, started changing datasets code which has a few bugs. [Alan Saul] + +* Minor changes to naming of signitures. [Alan Saul] + +* Changed more examples to accept optimize and plot. [Alan Saul] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Alan Saul] + +* Added comments for models module and adjusted setup. [Max Zwiessele] + +* Merge branch 'naniny' into devel. [Max Zwiessele] + +* Rename _models to models_modules to include in doc. [Max Zwiessele] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] + +* Some tidying in the regression examples. [James Hensman] + +* Added optimize and plot for classification, non_gaussian and stochastic examples. [Alan Saul] + +* Fixed gp_base and svigp for sampling (doesn't use it but needs the arguments) [Alan Saul] + +* Added constant to Z_tilde, now log likelihoods are equal! [Alan Saul] + +* Changed some parameters of the laplace, tidied up examples. [Alan Saul] + +* Dimensionality reduction examples updated with optimize, plot and verbose. [Max Zwiessele] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] + +* Changing the seed seems to fix Alan's bug. [James Hensman] + +* Fixed symmetry in checkgrad issue. [James Hensman] + +* Reverted the brent optimisation in laplace. [James Hensman] + + (For the 1D linesearch using Brent) + +* Improved detectino of sympy. [James Hensman] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] + +* Fixed exception handling bug in GPy/util/linalg.py:29. [Teo de Campos] + +* Argghdfklg. [James Hensman] + +* Better warings for cathcing of blaslib detection. [James Hensman] + +* Changeing models to _models in setup.py. [James Hensman] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Alan Saul] + +* Fixed step size for likelihood tests and allowed randomizing of laplace. [Alan Saul] + +* Fixed student_t approximation demo and changed convergence critera to difference of f. [Alan Saul] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Alan Saul] + +* Added cfg file to manfiest and package_data. [Alan Saul] + +* Dimensionality reduction example (oil) updated. [Max Zwiessele] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] + +* Removed ipdb statement from kern, cleaned up some nasty whitespace. [James Hensman] + +* More readme. [James Hensman] + +* More readme stuff. [James Hensman] + +* More readme edits. [James Hensman] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] + +* Minor edits to the README. [James Hensman] + +* Gradient checker comments and import updates. [Max Zwiessele] + +* Gradientchecker added as a model. [Max Zwiessele] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] + +* Push minor fix to eq_sympy kernel test. [Neil Lawrence] + +* Added some tips to the readme. [James Hensman] + +* Fixed import errors in tests. [Max Zwiessele] + +* Sympykern kern_tests now passing, code is inefficient but should be numerically stable. [Neil Lawrence] + +* Modified to improve part of stability, gradient checks still passing. [Neil Lawrence] + +* Added gradient of sympy kernel, seems to pass tests, but know it's not numerically stable. Checking in before making numerically stable. [Neil Lawrence] + +* Fixed test in kern.py to request correct output dim for multioutput covariances. [Neil Lawrence] + +* ODE_UY. [mu] + +* Fixing up the blas detectino in linalg. [James Hensman] + +* Removing ipdb statements. [James Hensman] + +* Lots of medding with the likelihoods to get the tests working. the tests still don;t work. [James Hensman] + +* Formatting docstring. [James Hensman] + +* Changed nasty whitespace. [James Hensman] + +* Eq_ode1 working but test failing? [Neil Lawrence] + +* Added eq_ode1 to constructors.py. [Neil Lawrence] + +* Fixed bug in sympy kernel and added sympolic.py back into utils __init__.py. [Neil Lawrence] + +* Merge with James's changes. [Neil Lawrence] + +* Merge. [James Hensman] + +* Removed some sympy stuff. [James Hensman] + +* Merge with James's changes. [Neil Lawrence] + +* Committing change for master check out. [Neil Lawrence] + +* Skipping crossterm tests instead of expected failure. [Max Zwiessele] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] + +* Added a path for the data resources. not all users will be working in the GPy directory. [James Hensman] + +* Moved data resource information to a json file. [Neil Lawrence] + +* Bug fix for single output sympy kernel. [Neil Lawrence] + +* Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [Neil Lawrence] + +* Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [mu] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Nicolo Fusi] + +* Fixed problem in warping. [Nicolo Fusi] + +* Constructor and init for ODE_UY. [mu] + +* Working eq_ode1 in sympy now. [Neil Lawrence] + +* Dim reduction imports. [Max Zwiessele] + +* Testing imports update and expected failure for crossterms. [Max Zwiessele] + +* Rename models to _models and import models in models.py. [Max Zwiessele] + +* Psistattests update. [Max Zwiessele] + +* Updated crossterms, rbf x any not working yet (derivatives) [Max Zwiessele] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] + +* Weird merge. [James Hensman] + +* Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [Neil Lawrence] + +* Part implementation of ode_eq functionality. Not yet numerically stable or efficient (some horrible use of cut and paste to get things working ...) [Neil Lawrence] + +* A trial namespace renaming. [James Hensman] + +* Better handling of missing config files. [James Hensman] + +* Debugging the config paths. [James Hensman] + +* Allowing the passing of 1D X to a GP. with warning of course. [James Hensman] + +* More fiddling with the windows path for config. [James Hensman] + + Where is the windows guru? out playing beach volley? + +* Changed how we search for config files on windows. [James Hensman] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] + +* Fixed up symmetric kern. [James Hensman] + +* Half way through crossterm objective. [Max Zwiessele] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] + +* Added block matrix utility. [James Hensman] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] + +* Added **likelihood_params to predictive_values. [Ricardo] + +* Changes in plot function: sampling vs numerical approximation. [Ricardo] + +* Adding docstring for symmetric kern. [James Hensman] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] + +* Bug fixed in numerical approx. to the predictive variance. [Ricardo] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Ricardo] + +* Numerical predictions fixed, sampling predictions are not working. [Ricardo] + +* Predictive_mean and predictive_variance now use gp_var as a parameter, rather than gp_std. [Ricardo] + +* Fixed product kern get and set state. [James Hensman] + +* Added getstate/setstate for product kernel. [James Hensman] + +* In the middle of crossterms. [Max Zwiessele] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] + +* 2D plots fixed. [Ricardo] + +* Passing **noise_args into predictive_values. [Ricardo] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Ricardo] + +* Missing term in the likelihood. [Ricardo] + +* Reverted broken kern. [Max Zwiessele] + +* Added variational distribution for latent space. [Max Zwiessele] + +* BGPLVM test for crossterms. [Max Zwiessele] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] + +* Even more data plotting. [James Hensman] + +* Fixed plotting isue with plot_f. [James Hensman] + +* Fixed the dpotrs use.. [Alan Saul] + +* Added dpotrs instead of cho_solve. [Alan Saul] + +* Removed ipython dependency from kern. [Alan Saul] + +* Sped up sampling a lot for student t, bernoulli and poisson, added sampling for gaussian and exponential (untested) [Alan Saul] + +* Merge remote-tracking branch 'gpy_real/devel' [Alan Saul] + +* Ignoring examples tests again. [Alan Saul] + +* Added sampling to student_t noise distribution, very slow and is possible to speed up. predictive mean analytical and variance need checking. [Alan Saul] + +* Tidying up and fixed objective being vector. [Alan Saul] + +* Added sampling for predictive quantiles and also mean and variance where necessary. [Alan Saul] + +* Rederived gamma distribution. [Alan Saul] + +* Added derivatives for poisson and a couple of examples, need to fix for EP. [Alan Saul] + +* Merged with devel. [Alan Saul] + +* Reimplemented gradients for exponential, seems to work for laplace now, needs a visual test though. [Alan Saul] + +* Was a bug in the examples_tests.py, fixed and added brendan faces to ignore list. [Alan Saul] + +* Minor clean up. [Alan Saul] + +* Removed unnecessary laplace examples. [Alan Saul] + +* Updated laplace example to use predictive density aswell as RMSE. [Alan Saul] + +* Added log predictive density, ln p(y*|D) [Alan Saul] + +* Merge remote-tracking branch 'gpy_real/devel' [Alan Saul] + +* Updated boston tests (more folds, allow a bias as the datasets are not normalized once split) and more folds. Tweaked some laplace line search parameters, added basis tests for ep. [Alan Saul] + +* Fixed bug in gradient checker where it worked differently given a integer parameter to a float. [Alan Saul] + +* Removed derivatives of variance wrt gp and derivatives of means with respect to gp from noise models. [Alan Saul] + +* Tore out code no longer used from noise_distributions due to rewriting using quadrature. [Alan Saul] + +* Added numerical mean and variance with quadrature, about to clean up. [Alan Saul] + +* Merge branch 'master' into merge_branch. [Alan Saul] + +* Changed the gradients (perhaps for the worse) [Alan Saul] + +* A few typos. [Alan Saul] + +* Gaussian likelihood errors, still not working. [Alan Saul] + +* Added gaussian checker and gaussian likelihood, not checkgrading yet. [Alan Saul] + +* Started adding gaussian sanity checker. [Alan Saul] + +* Got rid of some overdoing the approximation. [Alan Saul] + +* Started adding gaussian likelihood, changed round preloading old_a. [Alan Saul] + +* Trying to debug kernel parameters learning (fails even when noise fixed) may be some instablility, seems like it can get it if it starts close. [Alan Saul] + +* Fixed 2*variance plotting instead of 2*std plotting, tidied up. [Alan Saul] + +* Changed incorrect naming. [Alan Saul] + +* Reparameratised in terms of sigma2. [Alan Saul] + +* Playing trying to find what makes it want to go so low. [Alan Saul] + +* Fixed bug where B wasn't refering to current f location. [Alan Saul] + +* Everything seems to be gradchecking again. [Alan Saul] + +* Added minimizer for finding f, doesn't help. [Alan Saul] + +* Now checkgrads a lot more of the time, but still fails in optimisation, seems also odd that when parameter is fixed kernel parameters go to infinity. [Alan Saul] + +* Added another optimisation which doesn't use gradients. Seems like F is almost always found, but Y can be off, suggesting that Wi__Ki_W is wrong, maybe W? [Alan Saul] + +* Trying to fix optimisation problem, fixed a few bugs but still fails at very low noise. [Alan Saul] + +* Starting to fiddle with mode finding code. [Alan Saul] + +* Fixed a sign wrong, now gradchecks weirdly only above certain points. [Alan Saul] + +* Now gradchecks everytime but student_t fit is bad, noise is underestimated by a long way. [Alan Saul] + +* Checkgrads with explicit and implicit components half the time. [Alan Saul] + +* About to input new derivations for Z's... [Alan Saul] + +* Took out all the asserts and using pure broadcasting method of diagonal now. [Alan Saul] + +* Made it use the fact that W is diagonal and put assertions in to ensure that the results are the same. [Alan Saul] + +* Broken it by getting rid of squeeze, but now working on making it faster using proper vector multiplciation for diagonals. [Alan Saul] + +* Made more numerically stable in a hope that it will work and I will find a bug... [Alan Saul] + +* Lots of name changing and went through all likelihood gradients again. [Alan Saul] + +* Ripped out all things Laplace parameter estimation, starting again with new tactic. [Alan Saul] + +* About to rip out old chain rule method of learning gradients. [Alan Saul] + +* Trying to fix dL_dytil gradient. [Alan Saul] + +* Changed name. [Alan Saul] + +* FIXED DYTIL_DFHAT. [Alan Saul] + +* Workong on doing explicit gradients. [Alan Saul] + +* Gradients almost there for dytil_dfhat, diagonal terms are right. [Alan Saul] + +* Merged with devel. [Alan Saul] + +* Still getting closer to grads for likelihood. [Alan Saul] + +* Almost have likelihood gradients working but kernels still way off. [Alan Saul] + +* Working on putting callback to update laplace in callback. [Alan Saul] + +* Seem to have gradients much closer now. [Alan Saul] + +* Scale and switch KW+I. [Alan Saul] + +* Merged with upstream. [Alan Saul] + +* Added a debug examples. [Alan Saul] + +* Merging. [Alan Saul] + +* Merged with master. [Alan Saul] + +* Plotting problematic kernel. [Alan Saul] + +* Adding gradients, shapes starting to make sense. [Alan Saul] + +* Attempted to introduce gradient methods, won't work yet I doubt. [Alan Saul] + +* Merge remote-tracking branch 'upstream/devel' [Alan Saul] + +* Merged likelihood functions. [Alan Saul] + +* Should be working now, needed to change relative path names. [Alan Saul] + +* Merge branch 'merge_trial' [Alan Saul] + +* Merge branch 'old_repo' into merge_trial. [Alan Saul] + +* Make directory structure match that of GPy. [Alan Saul] + +* Tidy up comments. [Alan Saul] + +* Merged in the GPy upstream. [Alan Saul] + +* Merged in branch which had old_repo merged in. [Alan Saul] + +* Adding weibull likelihood, requires 'extra_data' to be passed to likelihood, i.e. the censoring information. [Alan Saul] + +* Fixed the z scalings. [Alan Saul] + +* Fixed laplace approximation and made more numerically stable with cholesky decompositions, and commented. [Alan Saul] + +* Stabalised most of the algorithm (apart from the end inversion which is impossible) [Alan Saul] + +* Added timing and realised mdot can be faster as its almost always a diagonal matrix its multiplying with. [Alan Saul] + +* Got the mode finding without computing Ki. [Alan Saul] + +* Fixed broadcasting bug, rasm now appears to work. [Alan Saul] + +* Still working on rasmussen, link function needs vectorizing I think. [Alan Saul] + +* Tidying up. [Alan Saul] + +* Added predicted values for student t, works well. [Alan Saul] + +* Working laplace, just needs predictive values. [Alan Saul] + +* Seemed to be working, now its not. [Alan Saul] + +* Changing definitions again... [Alan Saul] + +* Worked out in terms of W, needs gradients implementing. [Alan Saul] + +* Just breaking some things... [Alan Saul] + +* Following naming convention better, lots of inverses which should be able to get rid of one or two, unsure if it works. [Alan Saul] + +* Got an idea of how to implement! written in docs. [Alan Saul] + +* Trying to 'debug' [Alan Saul] + +* Got most of laplace approximation working. [Alan Saul] + +* Added some comments. [Alan Saul] + +* Initial commit, setting up the laplace approximation for a student t. [Alan Saul] + +* Initial commit. [James Hensman] + +* Use bfgs for laplace instead. [Alan Saul] + +* Moved transf_data to make data -1 or 1 from 0 or 1 for bernoulli with probit into the analytical moment match (but it 10% slower), needs removing from epmixednoise. [Alan Saul] + +* Changed naming from old derivatives of likelihoods to new ones in noise distributions. [Alan Saul] + +* Fixed breakage of dvar, tidied up to make more efficient. [Alan Saul] + +* Doc stringing. [Alan Saul] + +* Added quadrature numerical moment matching (but not predictive yet) [Alan Saul] + +* Fixed a few laplace bits. [Alan Saul] + +* Refactored gradients wrt parameters slightly, need to future proof against _get_param_names() disappearing. [Alan Saul] + +* Added more options to generic tests (constraining link function values as bernoulli requies R^{0,1}) and implemented new gradients for bernoulli. [Alan Saul] + +* Rename Binomial to Bernoulli (maybe generalise it with the constant later, but tilted distribution may change) [Alan Saul] + +* Added pdf_link's for gaussian and student t, added third derivatives for transformations and tests for them. [Alan Saul] + +* Renamed laplace_tests to likelihoods_tests. [Alan Saul] + +* Tidying up laplace_tests.py. [Alan Saul] + +* Fixed some bugs, added third derivative for log transformation, and did some doccing. [Alan Saul] + +* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] + +* Still tidying up, laplace now working again, gaussian and student_t likelihoods now done. [Alan Saul] + +* Started on chaining, must remember to chain _laplace_gradients aswell! [Alan Saul] + +* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] + +* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] + +* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] + +* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] + +* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] + +* Finished tearing gaussian noise down, time for student t. [Alan Saul] + +* Beginning to merge lik_functions and derivatives with richardos. [Alan Saul] + +* Docs. [Alan Saul] + +* Removed fit as it is unused. [Alan Saul] + +* More doc strings. [Alan Saul] + +* Doccing and testing for D dimensional input (not multiple dimensional Y yet) [Alan Saul] + +* Tidying up a lot, works for 1D, need to check for more dimensions. [Alan Saul] + +* Tidied up laplace. [Alan Saul] + +* Integrated Laplace and merged Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] + + Conflicts: + GPy/core/gp.py + GPy/likelihoods/__init__.py + GPy/likelihoods/likelihood_functions.py + GPy/likelihoods/link_functions.py + +* Fixed white variance. [Alan Saul] + +* Boston housing works (apart from variance of student t is not valid below 2) [Alan Saul] + +* Tests setup but not fitting properly yet. [Alan Saul] + +* Changed the examples (started boston data) and increased tolerance of finding fhat. [Alan Saul] + +* Added some stability and tidied up. [Alan Saul] + +* Tidying up. [Alan Saul] + +* Student t likelihood function checkgrads (summed gradients wrt to sigma2), maybe some numerical instability in laplace. [Alan Saul] + +* Now checkgrads for gaussian, and ALMOST for student t. [Alan Saul] + +* All gradients now gradcheck. [Alan Saul] + +* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] + +* Refactored tests. [Alan Saul] + +* Tidied up grad checking. [Alan Saul] + +* Added tests and fixed some naming. [Alan Saul] + +* Modified gradient_checker to allow for variable 'f' [Alan Saul] + +* Renamed some things, made some small (incorrect) gradient changes, generalised the gp regression for any likelihood, and added a place holder link function waiting for Richardos changes. [Alan Saul] + +* Removed unneeded dependency. [Alan Saul] + +* Merged GP models. [Alan Saul] + +* Dragged likelihood_function changes in. [Alan Saul] + +* Checked out relavent files. [Alan Saul] + +* Merged in real gpy. [Alan Saul] + +* Empty branch. [Alan Saul] + +* SPELLAFSDIUN. [Max Zwiessele] + +* Fixed up plotting in sparse_gp also. [James Hensman] + +* Fixed up the plotting. [James Hensman] + +* Fixed up plot in GP_base. [James Hensman] + +* Started changing the plotting in examples to remove plot_single_output. [James Hensman] + +* General tidying in models. [James Hensman] + +* Improved docstrings in svigp. [James Hensman] + +* Some tidying in gp.py. [James Hensman] + +* Docstrings and removal of duplicated plotting code in gp_base. [James Hensman] + +* Turned omp off by default as discussed. [James Hensman] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] + +* Added configuration file. [Nicolò Fusi] + + this was done to solve the OpenMP problem on Windows/mac, but I think it + is useful in general. All unit tests pass except the sympy kern ones. + +* Added xw_pen data. [Neil Lawrence] + +* Added xw_pen data. [Neil Lawrence] + +* Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [Neil Lawrence] + * Docstrinfs in kern.py. [James Hensman] +* Basic sim code functional. [Neil Lawrence] + * Updated sympy code, multioutput grad checks pass apart from wrt X. Similar problems with prediction as to sinc covariance, needs investigation. [Neil Lawrence] * Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [Neil Lawrence] @@ -7410,10 +7420,6 @@ * Added olivetti faces data set. It required adding netpbmfile.py a bsd licensed pgm file reader from Christoph Gohlke, which doesn't seem to have a spearate installer. Also modified image_show to assume by default that array ordering is python instead of fortran. Modified brendan_faces demo to explicilty force fortran ordering. Notified Teo of change. [Neil Lawrence] -* Domain and trtansformations namespace prettyfying. [Max Zwiessele] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into params. [Max Zwiessele] - * Some fixes and changes to the sympykern. [Neil Lawrence] * Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [Neil Lawrence] @@ -7442,18 +7448,6 @@ * Added first draft of functionality for multiple output sympy kernels. [Neil Lawrence] -* Tests updated. [Max Zwiessele] - -* Dim reduction examples Q= > input_dim= [Max Zwiessele] - -* Numpy non hashable AHHHHHH. [Max Zwiessele] - -* Dont print brackets in transformations. [Max Zwiessele] - -* Parameterized first beta test. [Max Zwiessele] - -* Merge branch 'devel' into params. [Max Zwiessele] - * Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] * Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [Neil Lawrence] @@ -7550,64 +7544,12 @@ * Added hetero back to the init. [James Hensman] -* Epsilon and power_ep now are parameters of update_likelihood. [Ricardo] - -* Crescent data example is better organized. [Ricardo] - -* Cross term testing switched on. [Max Zwiessele] - -* Comment linear x linear for testing purposes. [Max Zwiessele] - -* Parameter_testing. [Max Zwiessele] - -* Index operations now work on flattened indices. [Max Zwiessele] - -* Unoptimized parameter, still slower than current implementation. [Max Zwiessele] - -* Ndarray subclass cleaned up. [Max Zwiessele] - -* Subclassing ndarray almost functional. [Max Zwiessele] - -* Parameters ndarray, stuck at using strides for transformations. [Max Zwiessele] - -* Transformations are singletons now, weak refs for memory managment. [Max Zwiessele] - -* Almost there with array inheriting. [Max Zwiessele] - -* With subclassing ndarray, current_slice problems not solved... [Max Zwiessele] - -* Without inheriting from numpy.ndarray. ndarray functionality missing. [Max Zwiessele] - -* NegativeLogexp Pep8ted. [Max Zwiessele] - -* Index operations finalized. [Max Zwiessele] - -* Added index_operations and deleted them from paramter. [Max Zwießele] - -* Added parameter files - Alan. [Max Zwiessele] - -* Transformations singleton. [Max Zwiessele] - -* Merge branch 'devel' into params. [Max Zwiessele] - - Conflicts: - GPy/core/transformations.py - GPy/kern/parts/kernpart.py - * Merge pull request #77 from jamesmcm/devel. [James McMurray] Devel * Merge branch 'devel' of git://github.com/SheffieldML/GPy into devel. [James McMurray] -* Bug in prod-coreg kernels fixed, not in the most elegant way though. [Ricardo] - -* Disp=False in scipy optimizations. [Ricardo] - -* Function grep_model added, works like print model, but accepts regexp. [Ricardo] - -* Reverting error. [Ricardo] - * Rebuilt documentation. [James McMurray] * Changes to fix autodoc - need to test with ReadTheDocs still. [James McMurray] @@ -7616,6 +7558,18 @@ * Updated readme with instructions for compiling documentation, running unit tests. [James McMurray] +* Epsilon and power_ep now are parameters of update_likelihood. [Ricardo] + +* Crescent data example is better organized. [Ricardo] + +* Bug in prod-coreg kernels fixed, not in the most elegant way though. [Ricardo] + +* Disp=False in scipy optimizations. [Ricardo] + +* Function grep_model added, works like print model, but accepts regexp. [Ricardo] + +* Reverting error. [Ricardo] + * Normalize_Y is passed to core function. [Ricardo] * Redundant code commented. [Ricardo] @@ -7777,18 +7731,6 @@ * Fixed Alans checkgrad bug. [James Hensman] -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Ricardo] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Ricardo] - -* Working on the Poisson likelihood. [Ricardo] - -* Parameterization changes take a while. [Max Zwiessele] - -* Correcterd minor errors (imports etc) [Max Zwiessele] - -* Pep8'ed transformations module. [Max Zwiessele] - * Mrd_sim stable and deprecated. [Max Zwiessele] * Expectation test slicing. [Max Zwiessele] @@ -8113,6 +8055,10 @@ * Merge branch 'devel' into pickle. [Max Zwiessele] +* Pcikling? [Max Zwiessele] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Ricardo] + * Fixed an import. [Teo de Campos] * Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Nicolo Fusi] @@ -8121,31 +8067,53 @@ * Removed sympy import. [Nicolo Fusi] +* Added an include. [Teo de Campos] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Ricardo] + * Added missing import in util.linalg. [Nicolo Fusi] * Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Nicolo Fusi] -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Ricardo] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Ricardo] - -* Merge remote-tracking branch 'origin/devel' into devel. [Ricardo] - -* Bug fix in the confusion matrix. [Ricardo] - * Removed sympy helpers from init. [Nicolo Fusi] +* Working on the Poisson likelihood. [Ricardo] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Ricardo] + * Added init. [Nicolo Fusi] * Removed unnecessary gitignore line. [Nicolo Fusi] +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Ricardo] + * Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Nicolo Fusi] +* Merging by hand... [James Hensman] + * Kernels are now consistent with pep8 and common reason. [Nicolo Fusi] -* Added an include. [Teo de Campos] +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Nicolo Fusi] -* Merging by hand... [James Hensman] +* Fixed fixed kernel (aha!) [Nicolo Fusi] + +* Merge remote-tracking branch 'origin/devel' into devel. [Ricardo] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] + +* Corrected minor bug in Brownian kernel. [Nicolas] + +* Comment import visual in visualize due to issues in Windows/OSX. [Neil Lawrence] + +* Mods to visualize and dimensionality to make stick demos work for summer school. [Neil Lawrence] + +* Changes to the efficiency of the sparse GP when there are many outputs. [James Hensman] + +* Bug fix in the confusion matrix. [Ricardo] + +* Minor change in tutorial. [mu] + +* Added link to user mailing list. [James Hensman] * Merge branch 'master' of github.com:SheffieldML/GPy. [James Hensman] @@ -8165,22 +8133,6 @@ * Changed manifest from docs to doc. [James Hensman] -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Nicolo Fusi] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] - -* Corrected minor bug in Brownian kernel. [Nicolas] - -* Comment import visual in visualize due to issues in Windows/OSX. [Neil Lawrence] - -* Mods to visualize and dimensionality to make stick demos work for summer school. [Neil Lawrence] - -* Changes to the efficiency of the sparse GP when there are many outputs. [James Hensman] - -* Fixed fixed kernel (aha!) [Nicolo Fusi] - -* Pcikling? [Max Zwiessele] - * Pypi release update. [Max Zwiessele] * Merge branch 'master' into devel. [Max Zwiessele] @@ -9499,6 +9451,11 @@ * Added fixed effect kernel. [Nicolo Fusi] + +## 0.2 (2013-03-13) + +### Other + * Merge branch 'master' of github.com:SheffieldML/GPy. [Nicolo Fusi] * Merge branch 'debug' [James Hensman] From 53fa0b51f2b0ac3cf9c6c2686cd08d9238444250 Mon Sep 17 00:00:00 2001 From: Max Zwiessele Date: Mon, 3 Oct 2016 13:21:51 +0100 Subject: [PATCH 05/11] chg: pkg: added gitchangelog config file to repo --- .gitchangelog.rc | 192 +++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 192 insertions(+) create mode 100644 .gitchangelog.rc diff --git a/.gitchangelog.rc b/.gitchangelog.rc new file mode 100644 index 00000000..58d5e788 --- /dev/null +++ b/.gitchangelog.rc @@ -0,0 +1,192 @@ +## +## Format +## +## ACTION: [AUDIENCE:] COMMIT_MSG [!TAG ...] +## +## Description +## +## ACTION is one of 'chg', 'fix', 'new' +## +## Is WHAT the change is about. +## +## 'chg' is for refactor, small improvement, cosmetic changes... +## 'fix' is for bug fixes +## 'new' is for new features, big improvement +## +## AUDIENCE is optional and one of 'dev', 'usr', 'pkg', 'test', 'doc' +## +## Is WHO is concerned by the change. +## +## 'dev' is for developpers (API changes, refactors...) +## 'usr' is for final users (UI changes) +## 'pkg' is for packagers (packaging changes) +## 'test' is for testers (test only related changes) +## 'doc' is for doc guys (doc only changes) +## +## COMMIT_MSG is ... well ... the commit message itself. +## +## TAGs are additionnal adjective as 'refactor' 'minor' 'cosmetic' +## +## They are preceded with a '!' or a '@' (prefer the former, as the +## latter is wrongly interpreted in github.) Commonly used tags are: +## +## 'refactor' is obviously for refactoring code only +## 'minor' is for a very meaningless change (a typo, adding a comment) +## 'cosmetic' is for cosmetic driven change (re-indentation, 80-col...) +## 'wip' is for partial functionality but complete subfunctionality. +## +## Example: +## +## new: usr: support of bazaar implemented +## chg: re-indentend some lines !cosmetic +## new: dev: updated code to be compatible with last version of killer lib. +## fix: pkg: updated year of licence coverage. +## new: test: added a bunch of test around user usability of feature X. +## fix: typo in spelling my name in comment. !minor +## +## Please note that multi-line commit message are supported, and only the +## first line will be considered as the "summary" of the commit message. So +## tags, and other rules only applies to the summary. The body of the commit +## message will be displayed in the changelog without reformatting. + + +## +## ``ignore_regexps`` is a line of regexps +## +## Any commit having its full commit message matching any regexp listed here +## will be ignored and won't be reported in the changelog. +## +ignore_regexps = [ + r'@minor', r'!minor', + r'@cosmetic', r'!cosmetic', + r'@refactor', r'!refactor', + r'@wip', r'!wip', + r'^([cC]hg|[fF]ix|[nN]ew)\s*:\s*[p|P]kg:', + r'^([cC]hg|[fF]ix|[nN]ew)\s*:\s*[d|D]ev:', + r'^(.{3,3}\s*:)?\s*[fF]irst commit.?\s*$', + ] + + +## ``section_regexps`` is a list of 2-tuples associating a string label and a +## list of regexp +## +## Commit messages will be classified in sections thanks to this. Section +## titles are the label, and a commit is classified under this section if any +## of the regexps associated is matching. +## +section_regexps = [ + ('New', [ + r'^[nN]ew\s*:\s*((dev|use?r|pkg|test|doc)\s*:\s*)?([^\n]*)$', + ]), + ('Changes', [ + r'^[cC]hg\s*:\s*((dev|use?r|pkg|test|doc)\s*:\s*)?([^\n]*)$', + ]), + ('Fix', [ + r'^[fF]ix\s*:\s*((dev|use?r|pkg|test|doc)\s*:\s*)?([^\n]*)$', + ]), + + ('Other', None ## Match all lines + ), + +] + + +## ``body_process`` is a callable +## +## This callable will be given the original body and result will +## be used in the changelog. +## +## Available constructs are: +## +## - any python callable that take one txt argument and return txt argument. +## +## - ReSub(pattern, replacement): will apply regexp substitution. +## +## - Indent(chars=" "): will indent the text with the prefix +## Please remember that template engines gets also to modify the text and +## will usually indent themselves the text if needed. +## +## - Wrap(regexp=r"\n\n"): re-wrap text in separate paragraph to fill 80-Columns +## +## - noop: do nothing +## +## - ucfirst: ensure the first letter is uppercase. +## (usually used in the ``subject_process`` pipeline) +## +## - final_dot: ensure text finishes with a dot +## (usually used in the ``subject_process`` pipeline) +## +## - strip: remove any spaces before or after the content of the string +## +## Additionally, you can `pipe` the provided filters, for instance: +#body_process = Wrap(regexp=r'\n(?=\w+\s*:)') | Indent(chars=" ") +#body_process = Wrap(regexp=r'\n(?=\w+\s*:)') +#body_process = noop +body_process = ReSub(r'((^|\n)[A-Z]\w+(-\w+)*: .*(\n\s+.*)*)+$', r'') | strip + + +## ``subject_process`` is a callable +## +## This callable will be given the original subject and result will +## be used in the changelog. +## +## Available constructs are those listed in ``body_process`` doc. +subject_process = (strip | + ReSub(r'^([cC]hg|[fF]ix|[nN]ew)\s*:\s*((dev|use?r|pkg|test|doc)\s*:\s*)?([^\n@]*)(@[a-z]+\s+)*$', r'\4') | + ucfirst | final_dot) + + +## ``tag_filter_regexp`` is a regexp +## +## Tags that will be used for the changelog must match this regexp. +## +tag_filter_regexp = r'^v[0-9]+\.[0-9]+(\.[0-9]+)?$' + + +## ``unreleased_version_label`` is a string +## +## This label will be used as the changelog Title of the last set of changes +## between last valid tag and HEAD if any. +unreleased_version_label = "%%__version__%% (unreleased)" + + +## ``output_engine`` is a callable +## +## This will change the output format of the generated changelog file +## +## Available choices are: +## +## - rest_py +## +## Legacy pure python engine, outputs ReSTructured text. +## This is the default. +## +## - mustache() +## +## Template name could be any of the available templates in +## ``templates/mustache/*.tpl``. +## Requires python package ``pystache``. +## Examples: +## - mustache("markdown") +## - mustache("restructuredtext") +## +## - makotemplate() +## +## Template name could be any of the available templates in +## ``templates/mako/*.tpl``. +## Requires python package ``mako``. +## Examples: +## - makotemplate("restructuredtext") +## + +#output_engine = rest_py +#output_engine = mustache("restructuredtext") +output_engine = mustache("markdown") +#output_engine = makotemplate("restructuredtext") + + +## ``include_merge`` is a boolean +## +## This option tells git-log whether to include merge commits in the log. +## The default is to include them. +include_merge = True \ No newline at end of file From 7f8b287f63de21b8b1c891f5535c370e4c7d84bf Mon Sep 17 00:00:00 2001 From: Max Zwiessele Date: Mon, 3 Oct 2016 13:22:44 +0100 Subject: [PATCH 06/11] fix: pkg: README update --- README.md | 34 +++++++++++++++++----------------- 1 file changed, 17 insertions(+), 17 deletions(-) diff --git a/README.md b/README.md index 838c97e7..5a771e1b 100644 --- a/README.md +++ b/README.md @@ -13,9 +13,9 @@ The Gaussian processes framework in Python. ## What's new: -From now on we keep track of changes in the CHANGELOG.md. +From now on we keep track of changes in the CHANGELOG.md. If you want your changes to show up there follow the [guidelines](#gl). -In particular tag your commits by the [gitchangelog](https://github.com/vaab/gitchangelog) commit message format. +In particular tag your commits by the [gitchangelog](https://github.com/vaab/gitchangelog) commit message format. ## Contributing to GPy @@ -33,7 +33,7 @@ For an in depth description of pull requests, please visit https://help.github.c 6. The tests will be running on your pull request. In the comments section we will be able to discuss the changes and help you with any problems. Let us know if there are any in the comments, so we can help. 7. The pull request gets accepted and your awsome new feature will be in the next GPy release :) -For any further questions/suggestions head over to the issues section in GPy. +For any further questions/suggestions head over to the issues section in GPy. ### Pull Request Guidelines @@ -126,14 +126,14 @@ If you're having trouble installing GPy via `pip install GPy` here is a probable # Saving models in a consistent way across versions: -As pickle is inconsistent across python versions and heavily dependent on class structure, it behaves inconsistent across versions. +As pickle is inconsistent across python versions and heavily dependent on class structure, it behaves inconsistent across versions. Pickling as meant to serialize models within the same environment, and not to store models on disk to be used later on. To save a model it is best to save the m.param_array of it to disk (using numpy’s np.save). -Additionally, you save the script, which creates the model. -In this script you can create the model using initialize=False as a keyword argument and with the data loaded as normal. -You then set the model parameters by setting m.param_array[:] = loaded_params as the previously saved parameters. -Then you initialize the model by m.initialize_parameter(), which will make the model usable. +Additionally, you save the script, which creates the model. +In this script you can create the model using initialize=False as a keyword argument and with the data loaded as normal. +You then set the model parameters by setting m.param_array[:] = loaded_params as the previously saved parameters. +Then you initialize the model by m.initialize_parameter(), which will make the model usable. Be aware that up to this point the model is in an inconsistent state and cannot be used to produce any results. ```python @@ -161,15 +161,15 @@ New way of running tests is using coverage: Ensure nose and coverage is installed: pip install nose coverage - + Run nosetests from root directory of repository: coverage run travis_tests.py - + Create coverage report in htmlcov/ coverage html - + The coverage report is located in htmlcov/index.html ##### Legacy: using nosetests @@ -223,7 +223,7 @@ If you want to merge a branch into devel make sure the following steps are met: - Check that tests are there and are checking code where applicable. - [optional] Make changes if necessary and commit and push to run tests. - [optional] Repeat the above until tests pass. - - [optional] bump up the version of GPy using bumpversion. The configuration is done, so all you need is bumpversion [major|minor|patch]. + - [optional] bump up the version of GPy using bumpversion. The configuration is done, so all you need is bumpversion [major|minor|patch]. - Update the changelog using gitchangelog: `gitchangelog > CHANGELOG.md` - Commit the changes of the changelog as silent update: `git commit -m "chg: pkg: CHANGELOG update" CHANGELOG.md - Push the changes into devel. @@ -246,16 +246,16 @@ Then we are ready to make the last changes for the changelog and versioning: Now we can merge the pull request into devel: $ git checkout devel - $ git merge --no-ff kurtCutajar-devel + $ git merge --no-ff -devel $ git push origin devel - + This will update the devel branch of GPy. ### Deploying GPy -We have set up all deployment automatic. -Thus, all you need to do is create a pull request from devel to deploy. -Wait for the tests to finish (successfully!) and merge the pull request. +We have set up all deployment automatic. +Thus, all you need to do is create a pull request from devel to deploy. +Wait for the tests to finish (successfully!) and merge the pull request. This will update the package on pypi for all platforms fully automatically. ## Funding Acknowledgements From c4276dba90f2fc5274fae81ca179ef13484b9669 Mon Sep 17 00:00:00 2001 From: Max Zwiessele Date: Mon, 3 Oct 2016 13:23:34 +0100 Subject: [PATCH 07/11] chg: pkg: CHANGELOG update --- CHANGELOG.md | 1935 +++++++++++++++++++++++++------------------------- 1 file changed, 985 insertions(+), 950 deletions(-) diff --git a/CHANGELOG.md b/CHANGELOG.md index 4125feb9..c6822937 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -1,6 +1,6 @@ # Changelog -## %%__version__%% (unreleased) +## v1.5.4 (2016-10-03) ### New @@ -52,12 +52,6 @@ * [kern] fix #440. [mzwiessele] -* Bump version: 1.4.0 → 1.4.1. [mzwiessele] - -* [setup] added bumpversion tagging again. [mzwiessele] - -* [mrd] init updates and added tests. [mzwiessele] - * [doc] cleanup. [mzwiessele] * [merge] into new devel. [Max Zwiessele] @@ -90,6 +84,17 @@ * Added core code for GpSSM and GpGrid. [kcutajar] + +## v1.4.1 (2016-09-06) + +### Other + +* Bump version: 1.4.0 → 1.4.1. [mzwiessele] + +* [setup] added bumpversion tagging again. [mzwiessele] + +* [mrd] init updates and added tests. [mzwiessele] + * Merge pull request #356 from SheffieldML/binomial_laplace. [Max Zwiessele] Binomial laplace #352 @@ -492,18 +497,6 @@ * Merge branch 'deploy' into devel. [Max Zwiessele] -* Merge pull request #368 from SheffieldML/devel. [Max Zwiessele] - - README of pypi now directly in setup - -* Merge pull request #367 from SheffieldML/devel. [Max Zwiessele] - - Update setup to fix problems with slicing - -* Merge pull request #365 from SheffieldML/devel. [Max Zwiessele] - - patch 1.0.5 - * [py3] iterator .next fixes. [Max Zwiessele] * [imports] fix #392. [Max Zwiessele] @@ -572,14 +565,6 @@ * Add eq_ode1 kern and ibp_lfm model. [cdguarnizo] -* Bump version: 1.0.8 → 1.0.9. [mzwiessele] - -* [setxy] was bugged. [mzwiessele] - -* Bump version: 1.0.7 → 1.0.8. [mzwiessele] - -* [examples] dim reduction plotting changes. [Max Zwiessele] - * Integral kernels added, these allow 'histogram' or 'binned' data to be modelled. [Michael T Smith] * Fixed bug, replaced for loops with einsum. [alessandratosi] @@ -598,10 +583,6 @@ * Merge branch devel into gradientsxx. [alessandratosi] -* [fix #380] reloading ep. [mzwiessele] - -* [fix #380] reloading ep. [mzwiessele] - * Fixed bug in kernel_tests for gradients_XX. [alessandratosi] * [slicing] fixed slicing for second order derivatives. [mzwiessele] @@ -610,6 +591,40 @@ * Merge branch 'devel' into gradientsxx. [mzwiessele] +* Bug fix. [alessandratosi] + +* Syntax fix. [alessandratosi] + +* Merge branch 'devel' of https://github.com/SheffieldML/GPy into gradientsxx. [alessandratosi] + +* Modified kernel tests for gradients_XX. [alessandratosi] + +* Added kernel tests for gradients_XX. [alessandratosi] + +* Update function kern.gradients_XX() to compute cross-covariance terms. [alessandratosi] + + +## v1.0.9 (2016-05-11) + +### Other + +* Bump version: 1.0.8 → 1.0.9. [mzwiessele] + +* [setxy] was bugged. [mzwiessele] + + +## v1.0.8 (2016-05-11) + +### Other + +* Bump version: 1.0.7 → 1.0.8. [mzwiessele] + +* [examples] dim reduction plotting changes. [Max Zwiessele] + +* [fix #380] reloading ep. [mzwiessele] + +* [fix #380] reloading ep. [mzwiessele] + * Merge branch 'devel' into kenokabe-devel. [mzwiessele] * [statespace] omg. [mzwiessele] @@ -624,12 +639,6 @@ * [kernel addition] in statespace is bugged for py33 on mac, deactivating it. [mzwiessele] -* Bug fix. [alessandratosi] - -* Syntax fix. [alessandratosi] - -* Merge branch 'devel' of https://github.com/SheffieldML/GPy into gradientsxx. [alessandratosi] - * [statespace] less restrictive test for regular statespace model. [Max Zwiessele] * [travis] condition. [Max Zwiessele] @@ -640,12 +649,6 @@ * Added precomputed kernel class. [vsaase] -* Modified kernel tests for gradients_XX. [alessandratosi] - -* Added kernel tests for gradients_XX. [alessandratosi] - -* Update function kern.gradients_XX() to compute cross-covariance terms. [alessandratosi] - * [readme] added landscape for code cleanines. [Max Zwiessele] * Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] @@ -658,6 +661,15 @@ * [tests] show skipped. [Max Zwiessele] + +## v1.0.7 (2016-04-12) + +### Other + +* Merge pull request #368 from SheffieldML/devel. [Max Zwiessele] + + README of pypi now directly in setup + * [tests] classification tests less strict (sporadic fails) [Max Zwiessele] * [tests] verbose in travis. [Max Zwiessele] @@ -672,18 +684,36 @@ * Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] +* [setup] readme in setup. [Max Zwiessele] + +* []README. [Max Zwiessele] + + +## v1.0.6 (2016-04-11) + +### Other + +* Merge pull request #367 from SheffieldML/devel. [Max Zwiessele] + + Update setup to fix problems with slicing + * Update __version__.py. [Max Zwiessele] * Update setup.cfg. [Max Zwiessele] * Update setup.py. [Max Zwiessele] -* [setup] readme in setup. [Max Zwiessele] - -* []README. [Max Zwiessele] - * [coverage] some more restrictions. [Max Zwiessele] + +## v1.0.5 (2016-04-08) + +### Other + +* Merge pull request #365 from SheffieldML/devel. [Max Zwiessele] + + patch 1.0.5 + * [config] softfail when config cannot be written. [Max Zwiessele] * Bump version: 1.0.4 → 1.0.5. [Max Zwiessele] @@ -2287,38 +2317,43 @@ * Merge branch 'master' into devel. [Max Zwiessele] -* Bump version: 0.8.7 → 0.8.8. [Max Zwiessele] +* Merge branch 'master' into devel. [Max Zwiessele] * Merge branch 'master' into devel. [Max Zwiessele] -* Bump version: 0.8.6 → 0.8.7. [Max Zwiessele] +* Merge branch 'master' into devel. [Max Zwiessele] * Merge branch 'master' into devel. [Max Zwiessele] -* [py3] print statement. [Max Zwiessele] - * Merge branch 'master' into devel. [Max Zwiessele] -* [pred_var] added predictive variable as property now. [Max Zwiessele] - -* Merge branch 'master' into devel. [Max Zwiessele] - -* Bump version: 0.8.5 → 0.8.6. [Max Zwiessele] - -* Merge branch 'master' into devel. [Max Zwiessele] - -* Merge branch 'master' of github.com:SheffieldML/GPy. [Max Zwiessele] - -* Apply bug fix for set_XY. [Zhenwen Dai] - -* Converting .md to .rst automatically now. [Max Zwiessele] - * Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] * Bug fix for set_XY. [Zhenwen Dai] * Add set_XY test. [Zhenwen Dai] + +## v0.8.8 (2015-09-10) + +### Other + +* Bump version: 0.8.7 → 0.8.8. [Max Zwiessele] + +* Bump version: 0.8.6 → 0.8.7. [Max Zwiessele] + +* [py3] print statement. [Max Zwiessele] + +* [pred_var] added predictive variable as property now. [Max Zwiessele] + +* Bump version: 0.8.5 → 0.8.6. [Max Zwiessele] + +* Merge branch 'master' of github.com:SheffieldML/GPy. [Max Zwiessele] + +* Apply bug fix for set_XY. [Zhenwen Dai] + +* Converting .md to .rst automatically now. [Max Zwiessele] + * Bump version: 0.8.4 → 0.8.5. [Max Zwiessele] * Merge pull request #240 from SheffieldML/devel. [Max Zwiessele] @@ -3701,6 +3736,786 @@ * [huge merge] trying to merge old master and master. [Max Zwiessele] +* Merge pull request #148 from martinsch/normalization_fix. [James Hensman] + +* Normalization: avoid division by zero for constant feature dimensions. [mschiegg] + +* Merge pull request #160 from slinderman/master. [James Hensman] + + Removing set of numpy random seed + + Great spot! We're just about to relase the next iteratino of GPy, we'll make sure it happens there too. + + Thanks. + +* Removing set of numpy random seed. [Scott Linderman] + +* Removed unnecessary spaces from citation. [Nicolo Fusi] + +* Added a way to cite. [Nicolo Fusi] + +* Modified logexp transformation to prevent it returning zero when argument is under -36. [Neil Lawrence] + +* Version change. [Max Zwiessele] + +* Dim reduction examples clearer and init not as much black magic anymore. [Max Zwiessele] + +* Plot_latent now shows selected inputs, even after switching dimensions. [Max Zwiessele] + +* Added hapmap3 as dataset. [Max Zwiessele] + +* Sparse gp stability improved. [Max Zwiessele] + +* HapMap3 dataset added. [Max Zwiessele] + +* Added hapmap download, need to put in data preprocessing for actual usability. [Max Zwiessele] + +* Dim reduction examples. [mzwiessele] + +* Dim reduction examples. [mzwiessele] + +* Bgplvm steepest gradient map update. [mzwiessele] + +* Bgplvm steepest gradient map update. [mzwiessele] + +* Merge branch 'master' of github.com:SheffieldML/GPy. [mzwiessele] + +* Update README.md with funding acknowledgements. [Neil Lawrence] + +* Plotting bug for bgplvm fixed. [mzwiessele] + +* Scg optimizer scale bounds back to 1e-15. [mzwiessele] + +* Parameterized: added warning switch. [mzwiessele] + +* Mrd corrections. [mzwiessele] + +* Version change (early beta, do not change until everythin works. [Max Zwiessele] + +* Fixed the SCG optimizer, thanks to Yarin Gal. [James Hensman] + +* Version now 48. [Max Zwiessele] + +* Version update. [Max Zwiessele] + +* Merge branch 'master' of github.com:SheffieldML/GPy. [Max Zwiessele] + +* Pickling now allways binary as well as protocol -1. [Max Zwiessele] + +* Pickling now allways binary as well as protocol -1. [Max Zwiessele] + +* Using lbfgs algorithm from scipy.minimize, starting to convert all optimizers to minimize format. [Max Zwiessele] + +* Windows -.- [Max Zwiessele] + +* Versions update. [Max Zwiessele] + +* Image is a PIL requirement and should only be imported when actually using it. [Max Zwiessele] + +* Plot handling greatly improved for latent space visualizations. [Max Zwiessele] + +* Version file added. [Max Zwiessele] + +* Fixed Ctrl-C behaviour on Windows. [Nicolo Fusi] + +* Merge branch 'master' of github.com:SheffieldML/GPy. [Nicolo Fusi] + +* Fixed come path issues in sympykern. [James Hensman] + +* Rename and redoc. [Alan Saul] + +* Moving imports, attempting to update RTD. [Alan Saul] + +* Changed initalise_latent to take lower of init=PCA and corrected import. [Alan Saul] + +* Change order of imports for RTD. [Alan Saul] + +* Merge branch 'devel' [Alan Saul] + +* Merge branch 'master' into devel. [Alan Saul] + +* Removed variational.py. [Alan Saul] + +* Fixed plot_latent failure. [Alan Saul] + +* Ignore example tests. [Alan Saul] + +* Removed yes pipe for travis. [Alan Saul] + +* Seems to handle without answering now. [Alan Saul] + +* Reverse travis to see what it asks for. [Alan Saul] + +* Fixed some tests. [Alan Saul] + +* Merged and fixed conflict in ODE_UY.py. [Alan Saul] + +* Merge completed. [Max Zwiessele] + +* Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [mu] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Andreas] + +* Small changes in svigp. [Andreas] + +* ODE UY dkdtheta. [mu] + +* UY dkdtheta. [mu] + +* UY dkdtheta. [mu] + +* BGPLVM with missing data. [Max Zwiessele] + +* Pca adjustements to lvm models. [Max Zwiessele] + +* Ppca added, ppca missing data not working yet. [Max Zwiessele] + +* Diagonal operations. [Max Zwiessele] + +* Subarray indexing. [Max Zwiessele] + +* Documenting. [Max Zwiessele] + +* Bug in ODE_UY fix. [Alan Saul] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Alan Saul] + +* Fixed the numerical quadrature, won't work with large f unless normalized. [Alan Saul] + +* Fixed lots of breaking tests, reduced step size for checkgrad to 1e-4 in tests (perhaps this should be global), added some missing attributes to data_resources.json. [Alan Saul] + +* Set warnings for truncated hessian, it has been noted that that by truncating we can have incorrect posteriors, though at convergence this should not be a problem, could be fixed by not using Cholsky as the decomposition as it cannot handle non-positive definite mats. [Alan Saul] + +* Merge pull request #90 from jamesmcm/master. [Alan Saul] + + Fixing ReadTheDocs reading docstrings, adding data_resources.json + +* Adding data_resources.json to setup data files. [James McMurray] + +* Testing modification for ReadTheDocs to stop docstring errors. [James McMurray] + +* Ensure_defaiult constraints in svigp. [James Hensman] + +* Fixed Ctrl-C behaviour on Windows. [Nicolo Fusi] + +* Removed print statements from config parser, commented out ODE kerns. [Nicolo Fusi] + +* Merge branch 'devel' [Nicolo Fusi] + +* Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [mu] + +* Ode UY. [mu] + +* Dk dparameter. [mu] + +* ODE_UY. [mu] + +* Merge branch 'devel' [Nicolo Fusi] + +* Added some more error checking for downloading datasets. [Neil Lawrence] + +* Fixed some bugs in mocap.py where errors weren't being raised when file format was incorrect and made datasets.py check for 404 errors which previously were occuring silently ... shhhhh. [Neil Lawrence] + +* Fixed bugs in cmu_mocap loader where cmu_url was missing and loading in mocap data twice in same session led to incorrect url through copy error. [Neil Lawrence] + +* Fixed examples tests, started changing datasets code which has a few bugs. [Alan Saul] + +* Minor changes to naming of signitures. [Alan Saul] + +* Changed more examples to accept optimize and plot. [Alan Saul] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Alan Saul] + +* Added comments for models module and adjusted setup. [Max Zwiessele] + +* Merge branch 'naniny' into devel. [Max Zwiessele] + +* Rename _models to models_modules to include in doc. [Max Zwiessele] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] + +* Some tidying in the regression examples. [James Hensman] + +* Added optimize and plot for classification, non_gaussian and stochastic examples. [Alan Saul] + +* Fixed gp_base and svigp for sampling (doesn't use it but needs the arguments) [Alan Saul] + +* Added constant to Z_tilde, now log likelihoods are equal! [Alan Saul] + +* Changed some parameters of the laplace, tidied up examples. [Alan Saul] + +* Dimensionality reduction examples updated with optimize, plot and verbose. [Max Zwiessele] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] + +* Changing the seed seems to fix Alan's bug. [James Hensman] + +* Fixed symmetry in checkgrad issue. [James Hensman] + +* Reverted the brent optimisation in laplace. [James Hensman] + + (For the 1D linesearch using Brent) + +* Improved detectino of sympy. [James Hensman] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] + +* Fixed exception handling bug in GPy/util/linalg.py:29. [Teo de Campos] + +* Argghdfklg. [James Hensman] + +* Better warings for cathcing of blaslib detection. [James Hensman] + +* Changeing models to _models in setup.py. [James Hensman] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Alan Saul] + +* Fixed step size for likelihood tests and allowed randomizing of laplace. [Alan Saul] + +* Fixed student_t approximation demo and changed convergence critera to difference of f. [Alan Saul] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Alan Saul] + +* Added cfg file to manfiest and package_data. [Alan Saul] + +* Dimensionality reduction example (oil) updated. [Max Zwiessele] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] + +* Removed ipdb statement from kern, cleaned up some nasty whitespace. [James Hensman] + +* More readme. [James Hensman] + +* More readme stuff. [James Hensman] + +* More readme edits. [James Hensman] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] + +* Minor edits to the README. [James Hensman] + +* Gradient checker comments and import updates. [Max Zwiessele] + +* Gradientchecker added as a model. [Max Zwiessele] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] + +* Push minor fix to eq_sympy kernel test. [Neil Lawrence] + +* Added some tips to the readme. [James Hensman] + +* Fixed import errors in tests. [Max Zwiessele] + +* Sympykern kern_tests now passing, code is inefficient but should be numerically stable. [Neil Lawrence] + +* Modified to improve part of stability, gradient checks still passing. [Neil Lawrence] + +* Added gradient of sympy kernel, seems to pass tests, but know it's not numerically stable. Checking in before making numerically stable. [Neil Lawrence] + +* Fixed test in kern.py to request correct output dim for multioutput covariances. [Neil Lawrence] + +* ODE_UY. [mu] + +* Fixing up the blas detectino in linalg. [James Hensman] + +* Removing ipdb statements. [James Hensman] + +* Lots of medding with the likelihoods to get the tests working. the tests still don;t work. [James Hensman] + +* Formatting docstring. [James Hensman] + +* Changed nasty whitespace. [James Hensman] + +* Eq_ode1 working but test failing? [Neil Lawrence] + +* Added eq_ode1 to constructors.py. [Neil Lawrence] + +* Fixed bug in sympy kernel and added sympolic.py back into utils __init__.py. [Neil Lawrence] + +* Merge with James's changes. [Neil Lawrence] + +* Merge. [James Hensman] + +* Removed some sympy stuff. [James Hensman] + +* Merge with James's changes. [Neil Lawrence] + +* Committing change for master check out. [Neil Lawrence] + +* Skipping crossterm tests instead of expected failure. [Max Zwiessele] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] + +* Added a path for the data resources. not all users will be working in the GPy directory. [James Hensman] + +* Moved data resource information to a json file. [Neil Lawrence] + +* Bug fix for single output sympy kernel. [Neil Lawrence] + +* Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [Neil Lawrence] + +* Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [mu] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Nicolo Fusi] + +* Fixed problem in warping. [Nicolo Fusi] + +* Constructor and init for ODE_UY. [mu] + +* Working eq_ode1 in sympy now. [Neil Lawrence] + +* Dim reduction imports. [Max Zwiessele] + +* Testing imports update and expected failure for crossterms. [Max Zwiessele] + +* Rename models to _models and import models in models.py. [Max Zwiessele] + +* Psistattests update. [Max Zwiessele] + +* Updated crossterms, rbf x any not working yet (derivatives) [Max Zwiessele] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] + +* Weird merge. [James Hensman] + +* Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [Neil Lawrence] + +* Part implementation of ode_eq functionality. Not yet numerically stable or efficient (some horrible use of cut and paste to get things working ...) [Neil Lawrence] + +* A trial namespace renaming. [James Hensman] + +* Better handling of missing config files. [James Hensman] + +* Debugging the config paths. [James Hensman] + +* Allowing the passing of 1D X to a GP. with warning of course. [James Hensman] + +* More fiddling with the windows path for config. [James Hensman] + + Where is the windows guru? out playing beach volley? + +* Changed how we search for config files on windows. [James Hensman] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] + +* Fixed up symmetric kern. [James Hensman] + +* Half way through crossterm objective. [Max Zwiessele] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] + +* Added block matrix utility. [James Hensman] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] + +* Added **likelihood_params to predictive_values. [Ricardo] + +* Changes in plot function: sampling vs numerical approximation. [Ricardo] + +* Adding docstring for symmetric kern. [James Hensman] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] + +* Bug fixed in numerical approx. to the predictive variance. [Ricardo] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Ricardo] + +* Numerical predictions fixed, sampling predictions are not working. [Ricardo] + +* Predictive_mean and predictive_variance now use gp_var as a parameter, rather than gp_std. [Ricardo] + +* Fixed product kern get and set state. [James Hensman] + +* Added getstate/setstate for product kernel. [James Hensman] + +* In the middle of crossterms. [Max Zwiessele] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] + +* 2D plots fixed. [Ricardo] + +* Passing **noise_args into predictive_values. [Ricardo] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Ricardo] + +* Missing term in the likelihood. [Ricardo] + +* Reverted broken kern. [Max Zwiessele] + +* Added variational distribution for latent space. [Max Zwiessele] + +* BGPLVM test for crossterms. [Max Zwiessele] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] + +* Even more data plotting. [James Hensman] + +* Fixed plotting isue with plot_f. [James Hensman] + +* Fixed the dpotrs use.. [Alan Saul] + +* Added dpotrs instead of cho_solve. [Alan Saul] + +* Removed ipython dependency from kern. [Alan Saul] + +* Sped up sampling a lot for student t, bernoulli and poisson, added sampling for gaussian and exponential (untested) [Alan Saul] + +* Merge remote-tracking branch 'gpy_real/devel' [Alan Saul] + +* Ignoring examples tests again. [Alan Saul] + +* Added sampling to student_t noise distribution, very slow and is possible to speed up. predictive mean analytical and variance need checking. [Alan Saul] + +* Tidying up and fixed objective being vector. [Alan Saul] + +* Added sampling for predictive quantiles and also mean and variance where necessary. [Alan Saul] + +* Rederived gamma distribution. [Alan Saul] + +* Added derivatives for poisson and a couple of examples, need to fix for EP. [Alan Saul] + +* Merged with devel. [Alan Saul] + +* Reimplemented gradients for exponential, seems to work for laplace now, needs a visual test though. [Alan Saul] + +* Was a bug in the examples_tests.py, fixed and added brendan faces to ignore list. [Alan Saul] + +* Minor clean up. [Alan Saul] + +* Removed unnecessary laplace examples. [Alan Saul] + +* Updated laplace example to use predictive density aswell as RMSE. [Alan Saul] + +* Added log predictive density, ln p(y*|D) [Alan Saul] + +* Merge remote-tracking branch 'gpy_real/devel' [Alan Saul] + +* Updated boston tests (more folds, allow a bias as the datasets are not normalized once split) and more folds. Tweaked some laplace line search parameters, added basis tests for ep. [Alan Saul] + +* Fixed bug in gradient checker where it worked differently given a integer parameter to a float. [Alan Saul] + +* Removed derivatives of variance wrt gp and derivatives of means with respect to gp from noise models. [Alan Saul] + +* Tore out code no longer used from noise_distributions due to rewriting using quadrature. [Alan Saul] + +* Added numerical mean and variance with quadrature, about to clean up. [Alan Saul] + +* Merge branch 'master' into merge_branch. [Alan Saul] + +* Changed the gradients (perhaps for the worse) [Alan Saul] + +* A few typos. [Alan Saul] + +* Gaussian likelihood errors, still not working. [Alan Saul] + +* Added gaussian checker and gaussian likelihood, not checkgrading yet. [Alan Saul] + +* Started adding gaussian sanity checker. [Alan Saul] + +* Got rid of some overdoing the approximation. [Alan Saul] + +* Started adding gaussian likelihood, changed round preloading old_a. [Alan Saul] + +* Trying to debug kernel parameters learning (fails even when noise fixed) may be some instablility, seems like it can get it if it starts close. [Alan Saul] + +* Fixed 2*variance plotting instead of 2*std plotting, tidied up. [Alan Saul] + +* Changed incorrect naming. [Alan Saul] + +* Reparameratised in terms of sigma2. [Alan Saul] + +* Playing trying to find what makes it want to go so low. [Alan Saul] + +* Fixed bug where B wasn't refering to current f location. [Alan Saul] + +* Everything seems to be gradchecking again. [Alan Saul] + +* Added minimizer for finding f, doesn't help. [Alan Saul] + +* Now checkgrads a lot more of the time, but still fails in optimisation, seems also odd that when parameter is fixed kernel parameters go to infinity. [Alan Saul] + +* Added another optimisation which doesn't use gradients. Seems like F is almost always found, but Y can be off, suggesting that Wi__Ki_W is wrong, maybe W? [Alan Saul] + +* Trying to fix optimisation problem, fixed a few bugs but still fails at very low noise. [Alan Saul] + +* Starting to fiddle with mode finding code. [Alan Saul] + +* Fixed a sign wrong, now gradchecks weirdly only above certain points. [Alan Saul] + +* Now gradchecks everytime but student_t fit is bad, noise is underestimated by a long way. [Alan Saul] + +* Checkgrads with explicit and implicit components half the time. [Alan Saul] + +* About to input new derivations for Z's... [Alan Saul] + +* Took out all the asserts and using pure broadcasting method of diagonal now. [Alan Saul] + +* Made it use the fact that W is diagonal and put assertions in to ensure that the results are the same. [Alan Saul] + +* Broken it by getting rid of squeeze, but now working on making it faster using proper vector multiplciation for diagonals. [Alan Saul] + +* Made more numerically stable in a hope that it will work and I will find a bug... [Alan Saul] + +* Lots of name changing and went through all likelihood gradients again. [Alan Saul] + +* Ripped out all things Laplace parameter estimation, starting again with new tactic. [Alan Saul] + +* About to rip out old chain rule method of learning gradients. [Alan Saul] + +* Trying to fix dL_dytil gradient. [Alan Saul] + +* Changed name. [Alan Saul] + +* FIXED DYTIL_DFHAT. [Alan Saul] + +* Workong on doing explicit gradients. [Alan Saul] + +* Gradients almost there for dytil_dfhat, diagonal terms are right. [Alan Saul] + +* Merged with devel. [Alan Saul] + +* Still getting closer to grads for likelihood. [Alan Saul] + +* Almost have likelihood gradients working but kernels still way off. [Alan Saul] + +* Working on putting callback to update laplace in callback. [Alan Saul] + +* Seem to have gradients much closer now. [Alan Saul] + +* Scale and switch KW+I. [Alan Saul] + +* Merged with upstream. [Alan Saul] + +* Added a debug examples. [Alan Saul] + +* Merging. [Alan Saul] + +* Merged with master. [Alan Saul] + +* Plotting problematic kernel. [Alan Saul] + +* Adding gradients, shapes starting to make sense. [Alan Saul] + +* Attempted to introduce gradient methods, won't work yet I doubt. [Alan Saul] + +* Merge remote-tracking branch 'upstream/devel' [Alan Saul] + +* Merged likelihood functions. [Alan Saul] + +* Should be working now, needed to change relative path names. [Alan Saul] + +* Merge branch 'merge_trial' [Alan Saul] + +* Merge branch 'old_repo' into merge_trial. [Alan Saul] + +* Make directory structure match that of GPy. [Alan Saul] + +* Tidy up comments. [Alan Saul] + +* Merged in the GPy upstream. [Alan Saul] + +* Merged in branch which had old_repo merged in. [Alan Saul] + +* Adding weibull likelihood, requires 'extra_data' to be passed to likelihood, i.e. the censoring information. [Alan Saul] + +* Fixed the z scalings. [Alan Saul] + +* Fixed laplace approximation and made more numerically stable with cholesky decompositions, and commented. [Alan Saul] + +* Stabalised most of the algorithm (apart from the end inversion which is impossible) [Alan Saul] + +* Added timing and realised mdot can be faster as its almost always a diagonal matrix its multiplying with. [Alan Saul] + +* Got the mode finding without computing Ki. [Alan Saul] + +* Fixed broadcasting bug, rasm now appears to work. [Alan Saul] + +* Still working on rasmussen, link function needs vectorizing I think. [Alan Saul] + +* Tidying up. [Alan Saul] + +* Added predicted values for student t, works well. [Alan Saul] + +* Working laplace, just needs predictive values. [Alan Saul] + +* Seemed to be working, now its not. [Alan Saul] + +* Changing definitions again... [Alan Saul] + +* Worked out in terms of W, needs gradients implementing. [Alan Saul] + +* Just breaking some things... [Alan Saul] + +* Following naming convention better, lots of inverses which should be able to get rid of one or two, unsure if it works. [Alan Saul] + +* Got an idea of how to implement! written in docs. [Alan Saul] + +* Trying to 'debug' [Alan Saul] + +* Got most of laplace approximation working. [Alan Saul] + +* Added some comments. [Alan Saul] + +* Initial commit, setting up the laplace approximation for a student t. [Alan Saul] + +* Initial commit. [James Hensman] + +* Use bfgs for laplace instead. [Alan Saul] + +* Moved transf_data to make data -1 or 1 from 0 or 1 for bernoulli with probit into the analytical moment match (but it 10% slower), needs removing from epmixednoise. [Alan Saul] + +* Changed naming from old derivatives of likelihoods to new ones in noise distributions. [Alan Saul] + +* Fixed breakage of dvar, tidied up to make more efficient. [Alan Saul] + +* Doc stringing. [Alan Saul] + +* Added quadrature numerical moment matching (but not predictive yet) [Alan Saul] + +* Fixed a few laplace bits. [Alan Saul] + +* Refactored gradients wrt parameters slightly, need to future proof against _get_param_names() disappearing. [Alan Saul] + +* Added more options to generic tests (constraining link function values as bernoulli requies R^{0,1}) and implemented new gradients for bernoulli. [Alan Saul] + +* Rename Binomial to Bernoulli (maybe generalise it with the constant later, but tilted distribution may change) [Alan Saul] + +* Added pdf_link's for gaussian and student t, added third derivatives for transformations and tests for them. [Alan Saul] + +* Renamed laplace_tests to likelihoods_tests. [Alan Saul] + +* Tidying up laplace_tests.py. [Alan Saul] + +* Fixed some bugs, added third derivative for log transformation, and did some doccing. [Alan Saul] + +* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] + +* Still tidying up, laplace now working again, gaussian and student_t likelihoods now done. [Alan Saul] + +* Started on chaining, must remember to chain _laplace_gradients aswell! [Alan Saul] + +* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] + +* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] + +* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] + +* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] + +* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] + +* Finished tearing gaussian noise down, time for student t. [Alan Saul] + +* Beginning to merge lik_functions and derivatives with richardos. [Alan Saul] + +* Docs. [Alan Saul] + +* Removed fit as it is unused. [Alan Saul] + +* More doc strings. [Alan Saul] + +* Doccing and testing for D dimensional input (not multiple dimensional Y yet) [Alan Saul] + +* Tidying up a lot, works for 1D, need to check for more dimensions. [Alan Saul] + +* Tidied up laplace. [Alan Saul] + +* Integrated Laplace and merged Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] + + Conflicts: + GPy/core/gp.py + GPy/likelihoods/__init__.py + GPy/likelihoods/likelihood_functions.py + GPy/likelihoods/link_functions.py + +* Fixed white variance. [Alan Saul] + +* Boston housing works (apart from variance of student t is not valid below 2) [Alan Saul] + +* Tests setup but not fitting properly yet. [Alan Saul] + +* Changed the examples (started boston data) and increased tolerance of finding fhat. [Alan Saul] + +* Added some stability and tidied up. [Alan Saul] + +* Tidying up. [Alan Saul] + +* Student t likelihood function checkgrads (summed gradients wrt to sigma2), maybe some numerical instability in laplace. [Alan Saul] + +* Now checkgrads for gaussian, and ALMOST for student t. [Alan Saul] + +* All gradients now gradcheck. [Alan Saul] + +* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] + +* Refactored tests. [Alan Saul] + +* Tidied up grad checking. [Alan Saul] + +* Added tests and fixed some naming. [Alan Saul] + +* Modified gradient_checker to allow for variable 'f' [Alan Saul] + +* Renamed some things, made some small (incorrect) gradient changes, generalised the gp regression for any likelihood, and added a place holder link function waiting for Richardos changes. [Alan Saul] + +* Removed unneeded dependency. [Alan Saul] + +* Merged GP models. [Alan Saul] + +* Dragged likelihood_function changes in. [Alan Saul] + +* Checked out relavent files. [Alan Saul] + +* Merged in real gpy. [Alan Saul] + +* Empty branch. [Alan Saul] + +* SPELLAFSDIUN. [Max Zwiessele] + +* Fixed up plotting in sparse_gp also. [James Hensman] + +* Fixed up the plotting. [James Hensman] + +* Fixed up plot in GP_base. [James Hensman] + +* Started changing the plotting in examples to remove plot_single_output. [James Hensman] + +* General tidying in models. [James Hensman] + +* Improved docstrings in svigp. [James Hensman] + +* Some tidying in gp.py. [James Hensman] + +* Docstrings and removal of duplicated plotting code in gp_base. [James Hensman] + +* Turned omp off by default as discussed. [James Hensman] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] + +* Added configuration file. [Nicolò Fusi] + + this was done to solve the OpenMP problem on Windows/mac, but I think it + is useful in general. All unit tests pass except the sympy kern ones. + +* Added xw_pen data. [Neil Lawrence] + +* Added xw_pen data. [Neil Lawrence] + +* Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [Neil Lawrence] + +* Basic sim code functional. [Neil Lawrence] + +* Minor change in tutorial. [mu] + +* Added link to user mailing list. [James Hensman] + + +## v0.6.0 (2014-11-21) + +### Other + * Small random perturbations in kernel tests helps with the symmetry gradcheck bug. [James Hensman] * More cooooooopyrighting. [James Hensman] @@ -6575,841 +7390,8 @@ * Merged dimen reduction. [Max Zwiessele] -* Domain and trtansformations namespace prettyfying. [Max Zwiessele] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into params. [Max Zwiessele] - -* Tests updated. [Max Zwiessele] - -* Dim reduction examples Q= > input_dim= [Max Zwiessele] - -* Numpy non hashable AHHHHHH. [Max Zwiessele] - -* Dont print brackets in transformations. [Max Zwiessele] - -* Parameterized first beta test. [Max Zwiessele] - -* Merge branch 'devel' into params. [Max Zwiessele] - -* Cross term testing switched on. [Max Zwiessele] - -* Comment linear x linear for testing purposes. [Max Zwiessele] - -* Parameter_testing. [Max Zwiessele] - -* Index operations now work on flattened indices. [Max Zwiessele] - -* Unoptimized parameter, still slower than current implementation. [Max Zwiessele] - -* Ndarray subclass cleaned up. [Max Zwiessele] - -* Subclassing ndarray almost functional. [Max Zwiessele] - -* Parameters ndarray, stuck at using strides for transformations. [Max Zwiessele] - -* Transformations are singletons now, weak refs for memory managment. [Max Zwiessele] - -* Almost there with array inheriting. [Max Zwiessele] - -* With subclassing ndarray, current_slice problems not solved... [Max Zwiessele] - -* Without inheriting from numpy.ndarray. ndarray functionality missing. [Max Zwiessele] - -* NegativeLogexp Pep8ted. [Max Zwiessele] - -* Index operations finalized. [Max Zwiessele] - -* Added index_operations and deleted them from paramter. [Max Zwießele] - -* Added parameter files - Alan. [Max Zwiessele] - -* Transformations singleton. [Max Zwiessele] - -* Merge branch 'devel' into params. [Max Zwiessele] - - Conflicts: - GPy/core/transformations.py - GPy/kern/parts/kernpart.py - -* Parameterization changes take a while. [Max Zwiessele] - -* Correcterd minor errors (imports etc) [Max Zwiessele] - -* Pep8'ed transformations module. [Max Zwiessele] - -* Merge pull request #148 from martinsch/normalization_fix. [James Hensman] - -* Normalization: avoid division by zero for constant feature dimensions. [mschiegg] - -* Merge pull request #160 from slinderman/master. [James Hensman] - - Removing set of numpy random seed - - Great spot! We're just about to relase the next iteratino of GPy, we'll make sure it happens there too. - - Thanks. - -* Removing set of numpy random seed. [Scott Linderman] - -* Removed unnecessary spaces from citation. [Nicolo Fusi] - -* Added a way to cite. [Nicolo Fusi] - -* Modified logexp transformation to prevent it returning zero when argument is under -36. [Neil Lawrence] - -* Version change. [Max Zwiessele] - -* Dim reduction examples clearer and init not as much black magic anymore. [Max Zwiessele] - -* Plot_latent now shows selected inputs, even after switching dimensions. [Max Zwiessele] - -* Added hapmap3 as dataset. [Max Zwiessele] - -* Sparse gp stability improved. [Max Zwiessele] - -* HapMap3 dataset added. [Max Zwiessele] - -* Added hapmap download, need to put in data preprocessing for actual usability. [Max Zwiessele] - -* Dim reduction examples. [mzwiessele] - -* Dim reduction examples. [mzwiessele] - -* Bgplvm steepest gradient map update. [mzwiessele] - -* Bgplvm steepest gradient map update. [mzwiessele] - -* Merge branch 'master' of github.com:SheffieldML/GPy. [mzwiessele] - -* Update README.md with funding acknowledgements. [Neil Lawrence] - -* Plotting bug for bgplvm fixed. [mzwiessele] - -* Scg optimizer scale bounds back to 1e-15. [mzwiessele] - -* Parameterized: added warning switch. [mzwiessele] - -* Mrd corrections. [mzwiessele] - -* Version change (early beta, do not change until everythin works. [Max Zwiessele] - -* Fixed the SCG optimizer, thanks to Yarin Gal. [James Hensman] - -* Version now 48. [Max Zwiessele] - -* Version update. [Max Zwiessele] - -* Merge branch 'master' of github.com:SheffieldML/GPy. [Max Zwiessele] - -* Pickling now allways binary as well as protocol -1. [Max Zwiessele] - -* Pickling now allways binary as well as protocol -1. [Max Zwiessele] - -* Using lbfgs algorithm from scipy.minimize, starting to convert all optimizers to minimize format. [Max Zwiessele] - -* Windows -.- [Max Zwiessele] - -* Versions update. [Max Zwiessele] - -* Image is a PIL requirement and should only be imported when actually using it. [Max Zwiessele] - -* Plot handling greatly improved for latent space visualizations. [Max Zwiessele] - -* Version file added. [Max Zwiessele] - -* Fixed Ctrl-C behaviour on Windows. [Nicolo Fusi] - -* Merge branch 'master' of github.com:SheffieldML/GPy. [Nicolo Fusi] - -* Fixed come path issues in sympykern. [James Hensman] - -* Rename and redoc. [Alan Saul] - -* Moving imports, attempting to update RTD. [Alan Saul] - -* Changed initalise_latent to take lower of init=PCA and corrected import. [Alan Saul] - -* Change order of imports for RTD. [Alan Saul] - -* Merge branch 'devel' [Alan Saul] - -* Merge branch 'master' into devel. [Alan Saul] - -* Removed variational.py. [Alan Saul] - -* Fixed plot_latent failure. [Alan Saul] - -* Ignore example tests. [Alan Saul] - -* Removed yes pipe for travis. [Alan Saul] - -* Seems to handle without answering now. [Alan Saul] - -* Reverse travis to see what it asks for. [Alan Saul] - -* Fixed some tests. [Alan Saul] - -* Merged and fixed conflict in ODE_UY.py. [Alan Saul] - -* Merge completed. [Max Zwiessele] - -* Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [mu] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Andreas] - -* Small changes in svigp. [Andreas] - -* ODE UY dkdtheta. [mu] - -* UY dkdtheta. [mu] - -* UY dkdtheta. [mu] - -* BGPLVM with missing data. [Max Zwiessele] - -* Pca adjustements to lvm models. [Max Zwiessele] - -* Ppca added, ppca missing data not working yet. [Max Zwiessele] - -* Diagonal operations. [Max Zwiessele] - -* Subarray indexing. [Max Zwiessele] - -* Documenting. [Max Zwiessele] - -* Bug in ODE_UY fix. [Alan Saul] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Alan Saul] - -* Fixed the numerical quadrature, won't work with large f unless normalized. [Alan Saul] - -* Fixed lots of breaking tests, reduced step size for checkgrad to 1e-4 in tests (perhaps this should be global), added some missing attributes to data_resources.json. [Alan Saul] - -* Set warnings for truncated hessian, it has been noted that that by truncating we can have incorrect posteriors, though at convergence this should not be a problem, could be fixed by not using Cholsky as the decomposition as it cannot handle non-positive definite mats. [Alan Saul] - -* Merge pull request #90 from jamesmcm/master. [Alan Saul] - - Fixing ReadTheDocs reading docstrings, adding data_resources.json - -* Adding data_resources.json to setup data files. [James McMurray] - -* Testing modification for ReadTheDocs to stop docstring errors. [James McMurray] - -* Ensure_defaiult constraints in svigp. [James Hensman] - -* Fixed Ctrl-C behaviour on Windows. [Nicolo Fusi] - -* Removed print statements from config parser, commented out ODE kerns. [Nicolo Fusi] - -* Merge branch 'devel' [Nicolo Fusi] - -* Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [mu] - -* Ode UY. [mu] - -* Dk dparameter. [mu] - -* ODE_UY. [mu] - -* Merge branch 'devel' [Nicolo Fusi] - -* Added some more error checking for downloading datasets. [Neil Lawrence] - -* Fixed some bugs in mocap.py where errors weren't being raised when file format was incorrect and made datasets.py check for 404 errors which previously were occuring silently ... shhhhh. [Neil Lawrence] - -* Fixed bugs in cmu_mocap loader where cmu_url was missing and loading in mocap data twice in same session led to incorrect url through copy error. [Neil Lawrence] - -* Fixed examples tests, started changing datasets code which has a few bugs. [Alan Saul] - -* Minor changes to naming of signitures. [Alan Saul] - -* Changed more examples to accept optimize and plot. [Alan Saul] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Alan Saul] - -* Added comments for models module and adjusted setup. [Max Zwiessele] - -* Merge branch 'naniny' into devel. [Max Zwiessele] - -* Rename _models to models_modules to include in doc. [Max Zwiessele] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] - -* Some tidying in the regression examples. [James Hensman] - -* Added optimize and plot for classification, non_gaussian and stochastic examples. [Alan Saul] - -* Fixed gp_base and svigp for sampling (doesn't use it but needs the arguments) [Alan Saul] - -* Added constant to Z_tilde, now log likelihoods are equal! [Alan Saul] - -* Changed some parameters of the laplace, tidied up examples. [Alan Saul] - -* Dimensionality reduction examples updated with optimize, plot and verbose. [Max Zwiessele] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] - -* Changing the seed seems to fix Alan's bug. [James Hensman] - -* Fixed symmetry in checkgrad issue. [James Hensman] - -* Reverted the brent optimisation in laplace. [James Hensman] - - (For the 1D linesearch using Brent) - -* Improved detectino of sympy. [James Hensman] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] - -* Fixed exception handling bug in GPy/util/linalg.py:29. [Teo de Campos] - -* Argghdfklg. [James Hensman] - -* Better warings for cathcing of blaslib detection. [James Hensman] - -* Changeing models to _models in setup.py. [James Hensman] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Alan Saul] - -* Fixed step size for likelihood tests and allowed randomizing of laplace. [Alan Saul] - -* Fixed student_t approximation demo and changed convergence critera to difference of f. [Alan Saul] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Alan Saul] - -* Added cfg file to manfiest and package_data. [Alan Saul] - -* Dimensionality reduction example (oil) updated. [Max Zwiessele] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] - -* Removed ipdb statement from kern, cleaned up some nasty whitespace. [James Hensman] - -* More readme. [James Hensman] - -* More readme stuff. [James Hensman] - -* More readme edits. [James Hensman] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] - -* Minor edits to the README. [James Hensman] - -* Gradient checker comments and import updates. [Max Zwiessele] - -* Gradientchecker added as a model. [Max Zwiessele] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] - -* Push minor fix to eq_sympy kernel test. [Neil Lawrence] - -* Added some tips to the readme. [James Hensman] - -* Fixed import errors in tests. [Max Zwiessele] - -* Sympykern kern_tests now passing, code is inefficient but should be numerically stable. [Neil Lawrence] - -* Modified to improve part of stability, gradient checks still passing. [Neil Lawrence] - -* Added gradient of sympy kernel, seems to pass tests, but know it's not numerically stable. Checking in before making numerically stable. [Neil Lawrence] - -* Fixed test in kern.py to request correct output dim for multioutput covariances. [Neil Lawrence] - -* ODE_UY. [mu] - -* Fixing up the blas detectino in linalg. [James Hensman] - -* Removing ipdb statements. [James Hensman] - -* Lots of medding with the likelihoods to get the tests working. the tests still don;t work. [James Hensman] - -* Formatting docstring. [James Hensman] - -* Changed nasty whitespace. [James Hensman] - -* Eq_ode1 working but test failing? [Neil Lawrence] - -* Added eq_ode1 to constructors.py. [Neil Lawrence] - -* Fixed bug in sympy kernel and added sympolic.py back into utils __init__.py. [Neil Lawrence] - -* Merge with James's changes. [Neil Lawrence] - -* Merge. [James Hensman] - -* Removed some sympy stuff. [James Hensman] - -* Merge with James's changes. [Neil Lawrence] - -* Committing change for master check out. [Neil Lawrence] - -* Skipping crossterm tests instead of expected failure. [Max Zwiessele] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] - -* Added a path for the data resources. not all users will be working in the GPy directory. [James Hensman] - -* Moved data resource information to a json file. [Neil Lawrence] - -* Bug fix for single output sympy kernel. [Neil Lawrence] - -* Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [Neil Lawrence] - -* Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [mu] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Nicolo Fusi] - -* Fixed problem in warping. [Nicolo Fusi] - -* Constructor and init for ODE_UY. [mu] - -* Working eq_ode1 in sympy now. [Neil Lawrence] - -* Dim reduction imports. [Max Zwiessele] - -* Testing imports update and expected failure for crossterms. [Max Zwiessele] - -* Rename models to _models and import models in models.py. [Max Zwiessele] - -* Psistattests update. [Max Zwiessele] - -* Updated crossterms, rbf x any not working yet (derivatives) [Max Zwiessele] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] - -* Weird merge. [James Hensman] - -* Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [Neil Lawrence] - -* Part implementation of ode_eq functionality. Not yet numerically stable or efficient (some horrible use of cut and paste to get things working ...) [Neil Lawrence] - -* A trial namespace renaming. [James Hensman] - -* Better handling of missing config files. [James Hensman] - -* Debugging the config paths. [James Hensman] - -* Allowing the passing of 1D X to a GP. with warning of course. [James Hensman] - -* More fiddling with the windows path for config. [James Hensman] - - Where is the windows guru? out playing beach volley? - -* Changed how we search for config files on windows. [James Hensman] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] - -* Fixed up symmetric kern. [James Hensman] - -* Half way through crossterm objective. [Max Zwiessele] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] - -* Added block matrix utility. [James Hensman] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] - -* Added **likelihood_params to predictive_values. [Ricardo] - -* Changes in plot function: sampling vs numerical approximation. [Ricardo] - -* Adding docstring for symmetric kern. [James Hensman] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] - -* Bug fixed in numerical approx. to the predictive variance. [Ricardo] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Ricardo] - -* Numerical predictions fixed, sampling predictions are not working. [Ricardo] - -* Predictive_mean and predictive_variance now use gp_var as a parameter, rather than gp_std. [Ricardo] - -* Fixed product kern get and set state. [James Hensman] - -* Added getstate/setstate for product kernel. [James Hensman] - -* In the middle of crossterms. [Max Zwiessele] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] - -* 2D plots fixed. [Ricardo] - -* Passing **noise_args into predictive_values. [Ricardo] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Ricardo] - -* Missing term in the likelihood. [Ricardo] - -* Reverted broken kern. [Max Zwiessele] - -* Added variational distribution for latent space. [Max Zwiessele] - -* BGPLVM test for crossterms. [Max Zwiessele] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] - -* Even more data plotting. [James Hensman] - -* Fixed plotting isue with plot_f. [James Hensman] - -* Fixed the dpotrs use.. [Alan Saul] - -* Added dpotrs instead of cho_solve. [Alan Saul] - -* Removed ipython dependency from kern. [Alan Saul] - -* Sped up sampling a lot for student t, bernoulli and poisson, added sampling for gaussian and exponential (untested) [Alan Saul] - -* Merge remote-tracking branch 'gpy_real/devel' [Alan Saul] - -* Ignoring examples tests again. [Alan Saul] - -* Added sampling to student_t noise distribution, very slow and is possible to speed up. predictive mean analytical and variance need checking. [Alan Saul] - -* Tidying up and fixed objective being vector. [Alan Saul] - -* Added sampling for predictive quantiles and also mean and variance where necessary. [Alan Saul] - -* Rederived gamma distribution. [Alan Saul] - -* Added derivatives for poisson and a couple of examples, need to fix for EP. [Alan Saul] - -* Merged with devel. [Alan Saul] - -* Reimplemented gradients for exponential, seems to work for laplace now, needs a visual test though. [Alan Saul] - -* Was a bug in the examples_tests.py, fixed and added brendan faces to ignore list. [Alan Saul] - -* Minor clean up. [Alan Saul] - -* Removed unnecessary laplace examples. [Alan Saul] - -* Updated laplace example to use predictive density aswell as RMSE. [Alan Saul] - -* Added log predictive density, ln p(y*|D) [Alan Saul] - -* Merge remote-tracking branch 'gpy_real/devel' [Alan Saul] - -* Updated boston tests (more folds, allow a bias as the datasets are not normalized once split) and more folds. Tweaked some laplace line search parameters, added basis tests for ep. [Alan Saul] - -* Fixed bug in gradient checker where it worked differently given a integer parameter to a float. [Alan Saul] - -* Removed derivatives of variance wrt gp and derivatives of means with respect to gp from noise models. [Alan Saul] - -* Tore out code no longer used from noise_distributions due to rewriting using quadrature. [Alan Saul] - -* Added numerical mean and variance with quadrature, about to clean up. [Alan Saul] - -* Merge branch 'master' into merge_branch. [Alan Saul] - -* Changed the gradients (perhaps for the worse) [Alan Saul] - -* A few typos. [Alan Saul] - -* Gaussian likelihood errors, still not working. [Alan Saul] - -* Added gaussian checker and gaussian likelihood, not checkgrading yet. [Alan Saul] - -* Started adding gaussian sanity checker. [Alan Saul] - -* Got rid of some overdoing the approximation. [Alan Saul] - -* Started adding gaussian likelihood, changed round preloading old_a. [Alan Saul] - -* Trying to debug kernel parameters learning (fails even when noise fixed) may be some instablility, seems like it can get it if it starts close. [Alan Saul] - -* Fixed 2*variance plotting instead of 2*std plotting, tidied up. [Alan Saul] - -* Changed incorrect naming. [Alan Saul] - -* Reparameratised in terms of sigma2. [Alan Saul] - -* Playing trying to find what makes it want to go so low. [Alan Saul] - -* Fixed bug where B wasn't refering to current f location. [Alan Saul] - -* Everything seems to be gradchecking again. [Alan Saul] - -* Added minimizer for finding f, doesn't help. [Alan Saul] - -* Now checkgrads a lot more of the time, but still fails in optimisation, seems also odd that when parameter is fixed kernel parameters go to infinity. [Alan Saul] - -* Added another optimisation which doesn't use gradients. Seems like F is almost always found, but Y can be off, suggesting that Wi__Ki_W is wrong, maybe W? [Alan Saul] - -* Trying to fix optimisation problem, fixed a few bugs but still fails at very low noise. [Alan Saul] - -* Starting to fiddle with mode finding code. [Alan Saul] - -* Fixed a sign wrong, now gradchecks weirdly only above certain points. [Alan Saul] - -* Now gradchecks everytime but student_t fit is bad, noise is underestimated by a long way. [Alan Saul] - -* Checkgrads with explicit and implicit components half the time. [Alan Saul] - -* About to input new derivations for Z's... [Alan Saul] - -* Took out all the asserts and using pure broadcasting method of diagonal now. [Alan Saul] - -* Made it use the fact that W is diagonal and put assertions in to ensure that the results are the same. [Alan Saul] - -* Broken it by getting rid of squeeze, but now working on making it faster using proper vector multiplciation for diagonals. [Alan Saul] - -* Made more numerically stable in a hope that it will work and I will find a bug... [Alan Saul] - -* Lots of name changing and went through all likelihood gradients again. [Alan Saul] - -* Ripped out all things Laplace parameter estimation, starting again with new tactic. [Alan Saul] - -* About to rip out old chain rule method of learning gradients. [Alan Saul] - -* Trying to fix dL_dytil gradient. [Alan Saul] - -* Changed name. [Alan Saul] - -* FIXED DYTIL_DFHAT. [Alan Saul] - -* Workong on doing explicit gradients. [Alan Saul] - -* Gradients almost there for dytil_dfhat, diagonal terms are right. [Alan Saul] - -* Merged with devel. [Alan Saul] - -* Still getting closer to grads for likelihood. [Alan Saul] - -* Almost have likelihood gradients working but kernels still way off. [Alan Saul] - -* Working on putting callback to update laplace in callback. [Alan Saul] - -* Seem to have gradients much closer now. [Alan Saul] - -* Scale and switch KW+I. [Alan Saul] - -* Merged with upstream. [Alan Saul] - -* Added a debug examples. [Alan Saul] - -* Merging. [Alan Saul] - -* Merged with master. [Alan Saul] - -* Plotting problematic kernel. [Alan Saul] - -* Adding gradients, shapes starting to make sense. [Alan Saul] - -* Attempted to introduce gradient methods, won't work yet I doubt. [Alan Saul] - -* Merge remote-tracking branch 'upstream/devel' [Alan Saul] - -* Merged likelihood functions. [Alan Saul] - -* Should be working now, needed to change relative path names. [Alan Saul] - -* Merge branch 'merge_trial' [Alan Saul] - -* Merge branch 'old_repo' into merge_trial. [Alan Saul] - -* Make directory structure match that of GPy. [Alan Saul] - -* Tidy up comments. [Alan Saul] - -* Merged in the GPy upstream. [Alan Saul] - -* Merged in branch which had old_repo merged in. [Alan Saul] - -* Adding weibull likelihood, requires 'extra_data' to be passed to likelihood, i.e. the censoring information. [Alan Saul] - -* Fixed the z scalings. [Alan Saul] - -* Fixed laplace approximation and made more numerically stable with cholesky decompositions, and commented. [Alan Saul] - -* Stabalised most of the algorithm (apart from the end inversion which is impossible) [Alan Saul] - -* Added timing and realised mdot can be faster as its almost always a diagonal matrix its multiplying with. [Alan Saul] - -* Got the mode finding without computing Ki. [Alan Saul] - -* Fixed broadcasting bug, rasm now appears to work. [Alan Saul] - -* Still working on rasmussen, link function needs vectorizing I think. [Alan Saul] - -* Tidying up. [Alan Saul] - -* Added predicted values for student t, works well. [Alan Saul] - -* Working laplace, just needs predictive values. [Alan Saul] - -* Seemed to be working, now its not. [Alan Saul] - -* Changing definitions again... [Alan Saul] - -* Worked out in terms of W, needs gradients implementing. [Alan Saul] - -* Just breaking some things... [Alan Saul] - -* Following naming convention better, lots of inverses which should be able to get rid of one or two, unsure if it works. [Alan Saul] - -* Got an idea of how to implement! written in docs. [Alan Saul] - -* Trying to 'debug' [Alan Saul] - -* Got most of laplace approximation working. [Alan Saul] - -* Added some comments. [Alan Saul] - -* Initial commit, setting up the laplace approximation for a student t. [Alan Saul] - -* Initial commit. [James Hensman] - -* Use bfgs for laplace instead. [Alan Saul] - -* Moved transf_data to make data -1 or 1 from 0 or 1 for bernoulli with probit into the analytical moment match (but it 10% slower), needs removing from epmixednoise. [Alan Saul] - -* Changed naming from old derivatives of likelihoods to new ones in noise distributions. [Alan Saul] - -* Fixed breakage of dvar, tidied up to make more efficient. [Alan Saul] - -* Doc stringing. [Alan Saul] - -* Added quadrature numerical moment matching (but not predictive yet) [Alan Saul] - -* Fixed a few laplace bits. [Alan Saul] - -* Refactored gradients wrt parameters slightly, need to future proof against _get_param_names() disappearing. [Alan Saul] - -* Added more options to generic tests (constraining link function values as bernoulli requies R^{0,1}) and implemented new gradients for bernoulli. [Alan Saul] - -* Rename Binomial to Bernoulli (maybe generalise it with the constant later, but tilted distribution may change) [Alan Saul] - -* Added pdf_link's for gaussian and student t, added third derivatives for transformations and tests for them. [Alan Saul] - -* Renamed laplace_tests to likelihoods_tests. [Alan Saul] - -* Tidying up laplace_tests.py. [Alan Saul] - -* Fixed some bugs, added third derivative for log transformation, and did some doccing. [Alan Saul] - -* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] - -* Still tidying up, laplace now working again, gaussian and student_t likelihoods now done. [Alan Saul] - -* Started on chaining, must remember to chain _laplace_gradients aswell! [Alan Saul] - -* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] - -* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] - -* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] - -* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] - -* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] - -* Finished tearing gaussian noise down, time for student t. [Alan Saul] - -* Beginning to merge lik_functions and derivatives with richardos. [Alan Saul] - -* Docs. [Alan Saul] - -* Removed fit as it is unused. [Alan Saul] - -* More doc strings. [Alan Saul] - -* Doccing and testing for D dimensional input (not multiple dimensional Y yet) [Alan Saul] - -* Tidying up a lot, works for 1D, need to check for more dimensions. [Alan Saul] - -* Tidied up laplace. [Alan Saul] - -* Integrated Laplace and merged Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] - - Conflicts: - GPy/core/gp.py - GPy/likelihoods/__init__.py - GPy/likelihoods/likelihood_functions.py - GPy/likelihoods/link_functions.py - -* Fixed white variance. [Alan Saul] - -* Boston housing works (apart from variance of student t is not valid below 2) [Alan Saul] - -* Tests setup but not fitting properly yet. [Alan Saul] - -* Changed the examples (started boston data) and increased tolerance of finding fhat. [Alan Saul] - -* Added some stability and tidied up. [Alan Saul] - -* Tidying up. [Alan Saul] - -* Student t likelihood function checkgrads (summed gradients wrt to sigma2), maybe some numerical instability in laplace. [Alan Saul] - -* Now checkgrads for gaussian, and ALMOST for student t. [Alan Saul] - -* All gradients now gradcheck. [Alan Saul] - -* Merge remote-tracking branch 'gpy_real/devel' into merge_branch. [Alan Saul] - -* Refactored tests. [Alan Saul] - -* Tidied up grad checking. [Alan Saul] - -* Added tests and fixed some naming. [Alan Saul] - -* Modified gradient_checker to allow for variable 'f' [Alan Saul] - -* Renamed some things, made some small (incorrect) gradient changes, generalised the gp regression for any likelihood, and added a place holder link function waiting for Richardos changes. [Alan Saul] - -* Removed unneeded dependency. [Alan Saul] - -* Merged GP models. [Alan Saul] - -* Dragged likelihood_function changes in. [Alan Saul] - -* Checked out relavent files. [Alan Saul] - -* Merged in real gpy. [Alan Saul] - -* Empty branch. [Alan Saul] - -* SPELLAFSDIUN. [Max Zwiessele] - -* Fixed up plotting in sparse_gp also. [James Hensman] - -* Fixed up the plotting. [James Hensman] - -* Fixed up plot in GP_base. [James Hensman] - -* Started changing the plotting in examples to remove plot_single_output. [James Hensman] - -* General tidying in models. [James Hensman] - -* Improved docstrings in svigp. [James Hensman] - -* Some tidying in gp.py. [James Hensman] - -* Docstrings and removal of duplicated plotting code in gp_base. [James Hensman] - -* Turned omp off by default as discussed. [James Hensman] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] - -* Added configuration file. [Nicolò Fusi] - - this was done to solve the OpenMP problem on Windows/mac, but I think it - is useful in general. All unit tests pass except the sympy kern ones. - -* Added xw_pen data. [Neil Lawrence] - -* Added xw_pen data. [Neil Lawrence] - -* Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [Neil Lawrence] - * Docstrinfs in kern.py. [James Hensman] -* Basic sim code functional. [Neil Lawrence] - * Updated sympy code, multioutput grad checks pass apart from wrt X. Similar problems with prediction as to sinc covariance, needs investigation. [Neil Lawrence] * Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [Neil Lawrence] @@ -7420,6 +7402,10 @@ * Added olivetti faces data set. It required adding netpbmfile.py a bsd licensed pgm file reader from Christoph Gohlke, which doesn't seem to have a spearate installer. Also modified image_show to assume by default that array ordering is python instead of fortran. Modified brendan_faces demo to explicilty force fortran ordering. Notified Teo of change. [Neil Lawrence] +* Domain and trtansformations namespace prettyfying. [Max Zwiessele] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into params. [Max Zwiessele] + * Some fixes and changes to the sympykern. [Neil Lawrence] * Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [Neil Lawrence] @@ -7448,6 +7434,18 @@ * Added first draft of functionality for multiple output sympy kernels. [Neil Lawrence] +* Tests updated. [Max Zwiessele] + +* Dim reduction examples Q= > input_dim= [Max Zwiessele] + +* Numpy non hashable AHHHHHH. [Max Zwiessele] + +* Dont print brackets in transformations. [Max Zwiessele] + +* Parameterized first beta test. [Max Zwiessele] + +* Merge branch 'devel' into params. [Max Zwiessele] + * Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Max Zwiessele] * Merge branch 'devel' of https://github.com/SheffieldML/GPy into devel. [Neil Lawrence] @@ -7544,24 +7542,56 @@ * Added hetero back to the init. [James Hensman] +* Epsilon and power_ep now are parameters of update_likelihood. [Ricardo] + +* Crescent data example is better organized. [Ricardo] + +* Cross term testing switched on. [Max Zwiessele] + +* Comment linear x linear for testing purposes. [Max Zwiessele] + +* Parameter_testing. [Max Zwiessele] + +* Index operations now work on flattened indices. [Max Zwiessele] + +* Unoptimized parameter, still slower than current implementation. [Max Zwiessele] + +* Ndarray subclass cleaned up. [Max Zwiessele] + +* Subclassing ndarray almost functional. [Max Zwiessele] + +* Parameters ndarray, stuck at using strides for transformations. [Max Zwiessele] + +* Transformations are singletons now, weak refs for memory managment. [Max Zwiessele] + +* Almost there with array inheriting. [Max Zwiessele] + +* With subclassing ndarray, current_slice problems not solved... [Max Zwiessele] + +* Without inheriting from numpy.ndarray. ndarray functionality missing. [Max Zwiessele] + +* NegativeLogexp Pep8ted. [Max Zwiessele] + +* Index operations finalized. [Max Zwiessele] + +* Added index_operations and deleted them from paramter. [Max Zwießele] + +* Added parameter files - Alan. [Max Zwiessele] + +* Transformations singleton. [Max Zwiessele] + +* Merge branch 'devel' into params. [Max Zwiessele] + + Conflicts: + GPy/core/transformations.py + GPy/kern/parts/kernpart.py + * Merge pull request #77 from jamesmcm/devel. [James McMurray] Devel * Merge branch 'devel' of git://github.com/SheffieldML/GPy into devel. [James McMurray] -* Rebuilt documentation. [James McMurray] - -* Changes to fix autodoc - need to test with ReadTheDocs still. [James McMurray] - -* Fixed readme. [James McMurray] - -* Updated readme with instructions for compiling documentation, running unit tests. [James McMurray] - -* Epsilon and power_ep now are parameters of update_likelihood. [Ricardo] - -* Crescent data example is better organized. [Ricardo] - * Bug in prod-coreg kernels fixed, not in the most elegant way though. [Ricardo] * Disp=False in scipy optimizations. [Ricardo] @@ -7570,6 +7600,14 @@ * Reverting error. [Ricardo] +* Rebuilt documentation. [James McMurray] + +* Changes to fix autodoc - need to test with ReadTheDocs still. [James McMurray] + +* Fixed readme. [James McMurray] + +* Updated readme with instructions for compiling documentation, running unit tests. [James McMurray] + * Normalize_Y is passed to core function. [Ricardo] * Redundant code commented. [Ricardo] @@ -7731,6 +7769,18 @@ * Fixed Alans checkgrad bug. [James Hensman] +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Ricardo] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Ricardo] + +* Working on the Poisson likelihood. [Ricardo] + +* Parameterization changes take a while. [Max Zwiessele] + +* Correcterd minor errors (imports etc) [Max Zwiessele] + +* Pep8'ed transformations module. [Max Zwiessele] + * Mrd_sim stable and deprecated. [Max Zwiessele] * Expectation test slicing. [Max Zwiessele] @@ -8055,10 +8105,6 @@ * Merge branch 'devel' into pickle. [Max Zwiessele] -* Pcikling? [Max Zwiessele] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Ricardo] - * Fixed an import. [Teo de Campos] * Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Nicolo Fusi] @@ -8067,53 +8113,31 @@ * Removed sympy import. [Nicolo Fusi] -* Added an include. [Teo de Campos] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Ricardo] - * Added missing import in util.linalg. [Nicolo Fusi] * Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Nicolo Fusi] -* Removed sympy helpers from init. [Nicolo Fusi] - -* Working on the Poisson likelihood. [Ricardo] +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Ricardo] * Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Ricardo] +* Merge remote-tracking branch 'origin/devel' into devel. [Ricardo] + +* Bug fix in the confusion matrix. [Ricardo] + +* Removed sympy helpers from init. [Nicolo Fusi] + * Added init. [Nicolo Fusi] * Removed unnecessary gitignore line. [Nicolo Fusi] -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Ricardo] - * Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Nicolo Fusi] -* Merging by hand... [James Hensman] - * Kernels are now consistent with pep8 and common reason. [Nicolo Fusi] -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Nicolo Fusi] +* Added an include. [Teo de Campos] -* Fixed fixed kernel (aha!) [Nicolo Fusi] - -* Merge remote-tracking branch 'origin/devel' into devel. [Ricardo] - -* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] - -* Corrected minor bug in Brownian kernel. [Nicolas] - -* Comment import visual in visualize due to issues in Windows/OSX. [Neil Lawrence] - -* Mods to visualize and dimensionality to make stick demos work for summer school. [Neil Lawrence] - -* Changes to the efficiency of the sparse GP when there are many outputs. [James Hensman] - -* Bug fix in the confusion matrix. [Ricardo] - -* Minor change in tutorial. [mu] - -* Added link to user mailing list. [James Hensman] +* Merging by hand... [James Hensman] * Merge branch 'master' of github.com:SheffieldML/GPy. [James Hensman] @@ -8133,6 +8157,22 @@ * Changed manifest from docs to doc. [James Hensman] +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [Nicolo Fusi] + +* Merge branch 'devel' of github.com:SheffieldML/GPy into devel. [James Hensman] + +* Corrected minor bug in Brownian kernel. [Nicolas] + +* Comment import visual in visualize due to issues in Windows/OSX. [Neil Lawrence] + +* Mods to visualize and dimensionality to make stick demos work for summer school. [Neil Lawrence] + +* Changes to the efficiency of the sparse GP when there are many outputs. [James Hensman] + +* Fixed fixed kernel (aha!) [Nicolo Fusi] + +* Pcikling? [Max Zwiessele] + * Pypi release update. [Max Zwiessele] * Merge branch 'master' into devel. [Max Zwiessele] @@ -9451,11 +9491,6 @@ * Added fixed effect kernel. [Nicolo Fusi] - -## 0.2 (2013-03-13) - -### Other - * Merge branch 'master' of github.com:SheffieldML/GPy. [Nicolo Fusi] * Merge branch 'debug' [James Hensman] From 491a81015723e9221bf543914522aebb4ade9e0f Mon Sep 17 00:00:00 2001 From: Max Zwiessele Date: Mon, 3 Oct 2016 13:23:53 +0100 Subject: [PATCH 08/11] =?UTF-8?q?Bump=20version:=201.5.4=20=E2=86=92=201.5?= =?UTF-8?q?.5?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- GPy/__version__.py | 2 +- appveyor.yml | 2 +- setup.cfg | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) diff --git a/GPy/__version__.py b/GPy/__version__.py index 3a384fb2..8c118e61 100644 --- a/GPy/__version__.py +++ b/GPy/__version__.py @@ -1 +1 @@ -__version__ = "1.5.4" +__version__ = "1.5.5" diff --git a/appveyor.yml b/appveyor.yml index 36092f90..9e7351c1 100644 --- a/appveyor.yml +++ b/appveyor.yml @@ -3,7 +3,7 @@ environment: secure: 8/ZjXFwtd1S7ixd7PJOpptupKKEDhm2da/q3unabJ00= COVERALLS_REPO_TOKEN: secure: d3Luic/ESkGaWnZrvWZTKrzO+xaVwJWaRCEP0F+K/9DQGPSRZsJ/Du5g3s4XF+tS - gpy_version: 1.5.4 + gpy_version: 1.5.5 matrix: - PYTHON_VERSION: 2.7 MINICONDA: C:\Miniconda-x64 diff --git a/setup.cfg b/setup.cfg index 2ca0f9e6..11dee055 100644 --- a/setup.cfg +++ b/setup.cfg @@ -1,5 +1,5 @@ [bumpversion] -current_version = 1.5.4 +current_version = 1.5.5 tag = True commit = True From a140ba9607475beb16caf7fd094368fd91047489 Mon Sep 17 00:00:00 2001 From: Max Zwiessele Date: Mon, 3 Oct 2016 13:24:05 +0100 Subject: [PATCH 09/11] chg: pkg: CHANGELOG update --- CHANGELOG.md | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/CHANGELOG.md b/CHANGELOG.md index c6822937..e553ffb6 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -1,5 +1,12 @@ # Changelog +## v1.5.5 (2016-10-03) + +### Other + +* Bump version: 1.5.4 → 1.5.5. [Max Zwiessele] + + ## v1.5.4 (2016-10-03) ### New From f3a57b457e4613e4741697b66b6661e961c135a2 Mon Sep 17 00:00:00 2001 From: mzwiessele Date: Thu, 13 Oct 2016 12:21:47 +0100 Subject: [PATCH 10/11] new: gitchangelogrc --- .gitchangelog.rc | 191 +++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 191 insertions(+) create mode 100644 .gitchangelog.rc diff --git a/.gitchangelog.rc b/.gitchangelog.rc new file mode 100644 index 00000000..1e8ac088 --- /dev/null +++ b/.gitchangelog.rc @@ -0,0 +1,191 @@ +## +## Format +## +## ACTION: [AUDIENCE:] COMMIT_MSG [!TAG ...] +## +## Description +## +## ACTION is one of 'chg', 'fix', 'new' +## +## Is WHAT the change is about. +## +## 'chg' is for refactor, small improvement, cosmetic changes... +## 'fix' is for bug fixes +## 'new' is for new features, big improvement +## +## AUDIENCE is optional and one of 'dev', 'usr', 'pkg', 'test', 'doc' +## +## Is WHO is concerned by the change. +## +## 'dev' is for developpers (API changes, refactors...) +## 'usr' is for final users (UI changes) +## 'pkg' is for packagers (packaging changes) +## 'test' is for testers (test only related changes) +## 'doc' is for doc guys (doc only changes) +## +## COMMIT_MSG is ... well ... the commit message itself. +## +## TAGs are additionnal adjective as 'refactor' 'minor' 'cosmetic' +## +## They are preceded with a '!' or a '@' (prefer the former, as the +## latter is wrongly interpreted in github.) Commonly used tags are: +## +## 'refactor' is obviously for refactoring code only +## 'minor' is for a very meaningless change (a typo, adding a comment) +## 'cosmetic' is for cosmetic driven change (re-indentation, 80-col...) +## 'wip' is for partial functionality but complete subfunctionality. +## +## Example: +## +## new: usr: support of bazaar implemented +## chg: re-indentend some lines !cosmetic +## new: dev: updated code to be compatible with last version of killer lib. +## fix: pkg: updated year of licence coverage. +## new: test: added a bunch of test around user usability of feature X. +## fix: typo in spelling my name in comment. !minor +## +## Please note that multi-line commit message are supported, and only the +## first line will be considered as the "summary" of the commit message. So +## tags, and other rules only applies to the summary. The body of the commit +## message will be displayed in the changelog without reformatting. + + +## +## ``ignore_regexps`` is a line of regexps +## +## Any commit having its full commit message matching any regexp listed here +## will be ignored and won't be reported in the changelog. +## +ignore_regexps = [ + r'@minor', r'!minor', + r'@cosmetic', r'!cosmetic', + r'@refactor', r'!refactor', + r'@wip', r'!wip', + r'^([cC]hg|[fF]ix|[nN]ew)\s*:\s*[p|P]kg:', + r'^([cC]hg|[fF]ix|[nN]ew)\s*:\s*[d|D]ev:', + r'^(.{3,3}\s*:)?\s*[fF]irst commit.?\s*$', + ] + + +## ``section_regexps`` is a list of 2-tuples associating a string label and a +## list of regexp +## +## Commit messages will be classified in sections thanks to this. Section +## titles are the label, and a commit is classified under this section if any +## of the regexps associated is matching. +## +section_regexps = [ + ('New', [ + r'^[nN]ew\s*:\s*((dev|use?r|pkg|test|doc)\s*:\s*)?([^\n]*)$', + ]), + ('Changes', [ + r'^[cC]hg\s*:\s*((dev|use?r|pkg|test|doc)\s*:\s*)?([^\n]*)$', + ]), + ('Fix', [ + r'^[fF]ix\s*:\s*((dev|use?r|pkg|test|doc)\s*:\s*)?([^\n]*)$', + ]), + + ('Other', None ## Match all lines + ), + +] + + +## ``body_process`` is a callable +## +## This callable will be given the original body and result will +## be used in the changelog. +## +## Available constructs are: +## +## - any python callable that take one txt argument and return txt argument. +## +## - ReSub(pattern, replacement): will apply regexp substitution. +## +## - Indent(chars=" "): will indent the text with the prefix +## Please remember that template engines gets also to modify the text and +## will usually indent themselves the text if needed. +## +## - Wrap(regexp=r"\n\n"): re-wrap text in separate paragraph to fill 80-Columns +## +## - noop: do nothing +## +## - ucfirst: ensure the first letter is uppercase. +## (usually used in the ``subject_process`` pipeline) +## +## - final_dot: ensure text finishes with a dot +## (usually used in the ``subject_process`` pipeline) +## +## - strip: remove any spaces before or after the content of the string +## +## Additionally, you can `pipe` the provided filters, for instance: +#body_process = Wrap(regexp=r'\n(?=\w+\s*:)') | Indent(chars=" ") +#body_process = Wrap(regexp=r'\n(?=\w+\s*:)') +#body_process = noop +body_process = ReSub(r'((^|\n)[A-Z]\w+(-\w+)*: .*(\n\s+.*)*)+$', r'') | strip + + +## ``subject_process`` is a callable +## +## This callable will be given the original subject and result will +## be used in the changelog. +## +## Available constructs are those listed in ``body_process`` doc. +subject_process = (strip | + ReSub(r'^([cC]hg|[fF]ix|[nN]ew)\s*:\s*((dev|use?r|pkg|test|doc)\s*:\s*)?([^\n@]*)(@[a-z]+\s+)*$', r'\4') | + ucfirst | final_dot) + + +## ``tag_filter_regexp`` is a regexp +## +## Tags that will be used for the changelog must match this regexp. +## +tag_filter_regexp = r'^v[0-9]+\.[0-9]+(\.[0-9]+)?$' + + +## ``unreleased_version_label`` is a string +## +## This label will be used as the changelog Title of the last set of changes +## between last valid tag and HEAD if any. +unreleased_version_label = "Unreleased" + + +## ``output_engine`` is a callable +## +## This will change the output format of the generated changelog file +## +## Available choices are: +## +## - rest_py +## +## Legacy pure python engine, outputs ReSTructured text. +## This is the default. +## +## - mustache() +## +## Template name could be any of the available templates in +## ``templates/mustache/*.tpl``. +## Requires python package ``pystache``. +## Examples: +## - mustache("markdown") +## - mustache("restructuredtext") +## +## - makotemplate() +## +## Template name could be any of the available templates in +## ``templates/mako/*.tpl``. +## Requires python package ``mako``. +## Examples: +## - makotemplate("restructuredtext") +## +#output_engine = rest_py +#output_engine = mustache("restructuredtext") +output_engine = mustache("markdown") +#output_engine = makotemplate("restructuredtext") + + +## ``include_merge`` is a boolean +## +## This option tells git-log whether to include merge commits in the log. +## The default is to include them. +include_merge = True From 0007eccd79fafdcde177e9ac30866b41d25811ba Mon Sep 17 00:00:00 2001 From: mzwiessele Date: Thu, 13 Oct 2016 12:23:34 +0100 Subject: [PATCH 11/11] fix: pkg: minor fixes (imports and empty spaces) --- GPy/kern/src/kern.py | 1 - GPy/kern/src/rbf.py | 1 + 2 files changed, 1 insertion(+), 1 deletion(-) diff --git a/GPy/kern/src/kern.py b/GPy/kern/src/kern.py index 537d0f7c..931591fd 100644 --- a/GPy/kern/src/kern.py +++ b/GPy/kern/src/kern.py @@ -3,7 +3,6 @@ import sys import numpy as np from ...core.parameterization.parameterized import Parameterized -from paramz.core.observable_array import ObsAr from paramz.caching import Cache_this from .kernel_slice_operations import KernCallsViaSlicerMeta from functools import reduce diff --git a/GPy/kern/src/rbf.py b/GPy/kern/src/rbf.py index 5f3fc97a..9059ccc0 100644 --- a/GPy/kern/src/rbf.py +++ b/GPy/kern/src/rbf.py @@ -42,6 +42,7 @@ class RBF(Stationary): def dK2_drdr_diag(self): return -self.variance # as the diagonal of r is always filled with zeros + def __getstate__(self): dc = super(RBF, self).__getstate__() if self.useGPU: