From 688d6ac7a51547e2906f5956912c20a1a1bca3db Mon Sep 17 00:00:00 2001 From: James Hensman Date: Wed, 16 Jan 2013 11:21:06 +0000 Subject: [PATCH 01/36] fixed up dK_dX in the exponential and Matern kerns --- GPy/kern/Matern32.py | 6 +++--- GPy/kern/Matern52.py | 11 +++++------ GPy/kern/exponential.py | 2 +- 3 files changed, 9 insertions(+), 10 deletions(-) diff --git a/GPy/kern/Matern32.py b/GPy/kern/Matern32.py index 8223b37a..377b687f 100644 --- a/GPy/kern/Matern32.py +++ b/GPy/kern/Matern32.py @@ -78,14 +78,14 @@ class Matern32(kernpart): """derivative of the diagonal of the covariance matrix with respect to the parameters.""" target[0] += np.sum(partial) - def dK_dX(self,X,X2,target): + def dK_dX(self,partial,X,X2,target): """derivative of the covariance matrix with respect to X.""" if X2 is None: X2 = X dist = np.sqrt(np.sum(np.square((X[:,None,:]-X2[None,:,:])/self.lengthscales),-1))[:,:,None] ddist_dX = (X[:,None,:]-X2[None,:,:])/self.lengthscales**2/np.where(dist!=0.,dist,np.inf) - dK_dX += - np.transpose(3*self.variance*dist*np.exp(-np.sqrt(3)*dist)*ddist_dX,(1,0,2)) + dK_dX = - np.transpose(3*self.variance*dist*np.exp(-np.sqrt(3)*dist)*ddist_dX,(1,0,2)) target += np.sum(dK_dX*partial.T[:,:,None],0) - + def dKdiag_dX(self,X,target): pass diff --git a/GPy/kern/Matern52.py b/GPy/kern/Matern52.py index 65059a5b..172c0117 100644 --- a/GPy/kern/Matern52.py +++ b/GPy/kern/Matern52.py @@ -33,7 +33,6 @@ class Matern52(kernpart): self.Nparam = self.D + 1 self.name = 'Mat52' self.set_param(np.hstack((variance,lengthscales))) - def get_param(self): """return the value of the parameters.""" @@ -77,12 +76,12 @@ class Matern52(kernpart): """derivative of the diagonal of the covariance matrix with respect to the parameters.""" target[0] += np.sum(partial) - def dK_dX(self,X,X2,target): + def dK_dX(self,partial,X,X2,target): """derivative of the covariance matrix with respect to X.""" if X2 is None: X2 = X dist = np.sqrt(np.sum(np.square((X[:,None,:]-X2[None,:,:])/self.lengthscales),-1))[:,:,None] ddist_dX = (X[:,None,:]-X2[None,:,:])/self.lengthscales**2/np.where(dist!=0.,dist,np.inf) - dK_dX += - np.transpose(self.variance*5./3*dist*(1+np.sqrt(5)*dist)*np.exp(-np.sqrt(5)*dist)*ddist_dX,(1,0,2)) + dK_dX = - np.transpose(self.variance*5./3*dist*(1+np.sqrt(5)*dist)*np.exp(-np.sqrt(5)*dist)*ddist_dX,(1,0,2)) target += np.sum(dK_dX*partial.T[:,:,None],0) def dKdiag_dX(self,X,target): @@ -97,11 +96,11 @@ class Matern52(kernpart): :param F1: vector of derivatives of F :type F1: np.array :param F2: vector of second derivatives of F - :type F2: np.array + :type F2: np.array :param F3: vector of third derivatives of F - :type F3: np.array + :type F3: np.array :param lower,upper: boundaries of the input domain - :type lower,upper: floats + :type lower,upper: floats """ assert self.D == 1 def L(x,i): diff --git a/GPy/kern/exponential.py b/GPy/kern/exponential.py index ba97881e..8ad5b813 100644 --- a/GPy/kern/exponential.py +++ b/GPy/kern/exponential.py @@ -77,7 +77,7 @@ class exponential(kernpart): #NB: derivative of diagonal elements wrt lengthscale is 0 target[0] += np.sum(partial) - def dK_dX(self,X,X2,target): + def dK_dX(self,partial,X,X2,target): """derivative of the covariance matrix with respect to X.""" if X2 is None: X2 = X dist = np.sqrt(np.sum(np.square((X[:,None,:]-X2[None,:,:])/self.lengthscales),-1))[:,:,None] From 8da741f7257ca1ad8103935289cd8988dc4d7389 Mon Sep 17 00:00:00 2001 From: Alan Saul Date: Fri, 18 Jan 2013 11:36:58 +0000 Subject: [PATCH 02/36] Adding travis.yml file for Travis continuous intregration service, may not work yet --- .travis.yml | 12 ++++++++++++ 1 file changed, 12 insertions(+) create mode 100644 .travis.yml diff --git a/.travis.yml b/.travis.yml new file mode 100644 index 00000000..bc8eb3ab --- /dev/null +++ b/.travis.yml @@ -0,0 +1,12 @@ +language: python +python: + - "2.7" +# command to install dependencies, e.g. pip install -r requirements.txt --use-mirrors +install: + - "pip install scipy>=0.10.1" + - "pip install numpy>=1.6.2" + - "pip install sphinx" + - "pip install ./" +# command to run tests, e.g. python setup.py test +script: + - "nosetests --with-xcoverage --with-xunit --cover-package=GPy --cover-erase GPy/testing" From 52a97763300b7926c5ceb2a97ffe328af0759705 Mon Sep 17 00:00:00 2001 From: Alan Saul Date: Fri, 18 Jan 2013 11:42:46 +0000 Subject: [PATCH 03/36] Changing travis installation --- .travis.yml | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/.travis.yml b/.travis.yml index bc8eb3ab..a4faeaee 100644 --- a/.travis.yml +++ b/.travis.yml @@ -3,10 +3,10 @@ python: - "2.7" # command to install dependencies, e.g. pip install -r requirements.txt --use-mirrors install: - - "pip install scipy>=0.10.1" - - "pip install numpy>=1.6.2" - - "pip install sphinx" - - "pip install ./" + - pip install scipy>=0.10.1 + - pip install numpy>=1.6.2 + - pip install sphinx + - pip install . --use-mirrors # command to run tests, e.g. python setup.py test script: - - "nosetests --with-xcoverage --with-xunit --cover-package=GPy --cover-erase GPy/testing" + - nosetests --with-xcoverage --with-xunit --cover-package=GPy --cover-erase GPy/testing From 1aca88d27fb59a5e10c9a61dc9555c8cd9417507 Mon Sep 17 00:00:00 2001 From: Alan Saul Date: Fri, 18 Jan 2013 12:14:31 +0000 Subject: [PATCH 04/36] Fixed version setting for numpy scipy installatioN --- .travis.yml | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/.travis.yml b/.travis.yml index a4faeaee..921cf083 100644 --- a/.travis.yml +++ b/.travis.yml @@ -3,8 +3,8 @@ python: - "2.7" # command to install dependencies, e.g. pip install -r requirements.txt --use-mirrors install: - - pip install scipy>=0.10.1 - - pip install numpy>=1.6.2 + - pip install "scipy>=0.10.1" + - pip install "numpy>=1.6.2" - pip install sphinx - pip install . --use-mirrors # command to run tests, e.g. python setup.py test From 3c700dfa503f9e6fbd38b8da7cced139f2c4619f Mon Sep 17 00:00:00 2001 From: Neil Lawrence Date: Fri, 18 Jan 2013 12:31:37 +0000 Subject: [PATCH 05/36] Changed get_param and set_param to _get_params and _set_params --- GPy/core/model.py | 36 ++++----- GPy/core/parameterised.py | 34 ++++----- GPy/examples/GPLVM_demo.py | 28 ------- GPy/examples/GP_regression_demo.py | 51 ------------- GPy/examples/GP_regression_kern_demo.py | 33 --------- GPy/examples/classification.py | 6 +- GPy/examples/oil_flow_demo.py | 53 -------------- GPy/examples/regression.py | 97 ++++++++++++++++++++++++- GPy/examples/warped_GP_demo.py | 45 ------------ GPy/inference/samplers.py | 6 +- GPy/kern/Brownian.py | 8 +- GPy/kern/Matern32.py | 8 +- GPy/kern/Matern52.py | 8 +- GPy/kern/bias.py | 8 +- GPy/kern/exponential.py | 8 +- GPy/kern/finite_dimensional.py | 8 +- GPy/kern/kern.py | 12 +-- GPy/kern/kernpart.py | 6 +- GPy/kern/linear.py | 8 +- GPy/kern/linear_ARD.py | 8 +- GPy/kern/rbf.py | 8 +- GPy/kern/rbf_ARD.py | 18 ++--- GPy/kern/spline.py | 8 +- GPy/kern/sympykern.py | 8 +- GPy/kern/white.py | 8 +- GPy/models/GPLVM.py | 8 +- GPy/models/GP_EP.py | 10 +-- GPy/models/GP_regression.py | 6 +- GPy/models/generalized_FITC.py | 10 +-- GPy/models/sparse_GPLVM.py | 12 +-- GPy/models/sparse_GP_regression.py | 8 +- GPy/models/warped_GP.py | 18 ++--- GPy/util/warping_functions.py | 4 +- 33 files changed, 239 insertions(+), 358 deletions(-) delete mode 100644 GPy/examples/GPLVM_demo.py delete mode 100644 GPy/examples/GP_regression_demo.py delete mode 100644 GPy/examples/GP_regression_kern_demo.py delete mode 100644 GPy/examples/oil_flow_demo.py delete mode 100644 GPy/examples/warped_GP_demo.py diff --git a/GPy/core/model.py b/GPy/core/model.py index 46cf6ac9..d3c6e582 100644 --- a/GPy/core/model.py +++ b/GPy/core/model.py @@ -14,14 +14,14 @@ from ..inference import optimization class model(parameterised): def __init__(self): parameterised.__init__(self) - self.priors = [None for i in range(self.get_param().size)] + self.priors = [None for i in range(self._get_params().size)] self.optimization_runs = [] self.sampling_runs = [] - self.set_param(self.get_param()) + self._set_params(self._get_params()) self.preferred_optimizer = 'tnc' - def get_param(self): + def _get_params(self): raise NotImplementedError, "this needs to be implemented to utilise the model class" - def set_param(self,x): + def _set_params(self,x): raise NotImplementedError, "this needs to be implemented to utilise the model class" def log_likelihood(self): raise NotImplementedError, "this needs to be implemented to utilise the model class" @@ -67,7 +67,7 @@ class model(parameterised): unconst = np.setdiff1d(which, self.constrained_positive_indices) if len(unconst): print "Warning: constraining parameters to be positive:" - print '\n'.join([n for i,n in enumerate(self.get_param_names()) if i in unconst]) + print '\n'.join([n for i,n in enumerate(self._get_param_names()) if i in unconst]) print '\n' self.constrain_positive(unconst) elif isinstance(what,priors.Gaussian): @@ -86,7 +86,7 @@ class model(parameterised): """ matches = self.grep_param_names(name) if len(matches): - return self.get_param()[matches] + return self._get_params()[matches] else: raise AttributeError, "no parameter matches %s"%name @@ -96,9 +96,9 @@ class model(parameterised): """ matches = self.grep_param_names(name) if len(matches): - x = self.get_param() + x = self._get_params() x[matches] = val - self.set_param(x) + self._set_params(x) else: raise AttributeError, "no parameter matches %s"%name @@ -106,11 +106,11 @@ class model(parameterised): def log_prior(self): """evaluate the prior""" - return np.sum([p.lnpdf(x) for p, x in zip(self.priors,self.get_param()) if p is not None]) + return np.sum([p.lnpdf(x) for p, x in zip(self.priors,self._get_params()) if p is not None]) def log_prior_gradients(self): """evaluate the gradients of the priors""" - x = self.get_param() + x = self._get_params() ret = np.zeros(x.size) [np.put(ret,i,p.lnpdf_grad(xx)) for i,(p,xx) in enumerate(zip(self.priors,x)) if not p is None] return ret @@ -121,7 +121,7 @@ class model(parameterised): Adjust the gradient for constraints and ties, return. """ g = self.log_likelihood_gradients() + self.log_prior_gradients() - x = self.get_param() + x = self._get_params() g[self.constrained_positive_indices] = g[self.constrained_positive_indices]*x[self.constrained_positive_indices] g[self.constrained_negative_indices] = g[self.constrained_negative_indices]*x[self.constrained_negative_indices] [np.put(g,i,g[i]*(x[i]-l)*(h-x[i])/(h-l)) for i,l,h in zip(self.constrained_bounded_indices, self.constrained_bounded_lowers, self.constrained_bounded_uppers)] @@ -142,9 +142,9 @@ class model(parameterised): x = np.random.randn(x.size) self.expand_param(x) #now draw from prior where possible - x = self.get_param() + x = self._get_params() [np.put(x,i,p.rvs(1)) for i,p in enumerate(self.priors) if not p is None] - self.set_param(x) + self._set_params(x) self.expand_param(self.extract_param())#makes sure all of the tied parameters get the same init (since there's only one prior object...) @@ -194,7 +194,7 @@ class model(parameterised): for s in positive_strings: for i in self.grep_param_names(s): if not (i in self.all_constrained_indices()): - name = self.get_param_names()[i] + name = self._get_param_names()[i] self.constrain_positive(name) if warn: print "Warning! constraining %s postive"%name @@ -248,13 +248,13 @@ class model(parameterised): else: print "numerically calculating hessian. please be patient!" - x = self.get_param() + x = self._get_params() def f(x): - self.set_param(x) + self._set_params(x) return self.log_likelihood() h = ndt.Hessian(f) A = -h(x) - self.set_param(x) + self._set_params(x) # check for almost zero components on the diagonal which screw up the cholesky aa = np.nonzero((np.diag(A)<1e-6) & (np.diag(A)>0.))[0] A[aa,aa] = 0. @@ -268,7 +268,7 @@ class model(parameterised): hld = np.sum(np.log(np.diag(jitchol(A)[0]))) except: return np.nan - return 0.5*self.get_param().size*np.log(2*np.pi) + self.log_likelihood() - hld + return 0.5*self._get_params().size*np.log(2*np.pi) + self.log_likelihood() - hld def __str__(self): s = parameterised.__str__(self).split('\n') diff --git a/GPy/core/parameterised.py b/GPy/core/parameterised.py index da0b6056..738bde5b 100644 --- a/GPy/core/parameterised.py +++ b/GPy/core/parameterised.py @@ -87,7 +87,7 @@ class parameterised(object): Returns ------- - the indices of self.get_param_names which match the regular expression. + the indices of self._get_param_names which match the regular expression. Notes ----- @@ -96,9 +96,9 @@ class parameterised(object): if type(expr) is str: expr = re.compile(expr) - return np.nonzero([expr.search(name) for name in self.get_param_names()])[0] + return np.nonzero([expr.search(name) for name in self._get_param_names()])[0] elif type(expr) is re._pattern_type: - return np.nonzero([expr.search(name) for name in self.get_param_names()])[0] + return np.nonzero([expr.search(name) for name in self._get_param_names()])[0] else: return expr @@ -115,11 +115,11 @@ class parameterised(object): assert not np.any(matches[:,None]==self.all_constrained_indices()), "Some indices are already constrained" self.constrained_positive_indices = np.hstack((self.constrained_positive_indices, matches)) #check to ensure constraint is in place - x = self.get_param() + x = self._get_params() for i,xx in enumerate(x): if (xx<0) & (i in matches): x[i] = -xx - self.set_param(x) + self._set_params(x) def unconstrain(self,which): @@ -163,11 +163,11 @@ class parameterised(object): assert not np.any(matches[:,None]==self.all_constrained_indices()), "Some indices are already constrained" self.constrained_negative_indices = np.hstack((self.constrained_negative_indices, matches)) #check to ensure constraint is in place - x = self.get_param() + x = self._get_params() for i,xx in enumerate(x): if (xx>0.) and (i in matches): x[i] = -xx - self.set_param(x) + self._set_params(x) @@ -187,11 +187,11 @@ class parameterised(object): self.constrained_bounded_uppers.append(upper) self.constrained_bounded_lowers.append(lower) #check to ensure constraint is in place - x = self.get_param() + x = self._get_params() for i,xx in enumerate(x): if ((xx<=lower)|(xx>=upper)) & (i in matches): x[i] = sigmoid(xx)*(upper-lower) + lower - self.set_param(x) + self._set_params(x) def constrain_fixed(self, which, value = None): @@ -213,14 +213,14 @@ class parameterised(object): if value != None: self.constrained_fixed_values.append(value) else: - self.constrained_fixed_values.append(self.get_param()[self.constrained_fixed_indices[-1]]) + self.constrained_fixed_values.append(self._get_params()[self.constrained_fixed_indices[-1]]) #self.constrained_fixed_values.append(value) self.expand_param(self.extract_param()) def extract_param(self): - """use self.get_param to get the 'true' parameters of the model, which are then tied, constrained and fixed""" - x = self.get_param() + """use self._get_params to get the 'true' parameters of the model, which are then tied, constrained and fixed""" + x = self._get_params() x[self.constrained_positive_indices] = np.log(x[self.constrained_positive_indices]) x[self.constrained_negative_indices] = np.log(-x[self.constrained_negative_indices]) [np.put(x,i,np.log(np.clip(x[i]-l,1e-10,np.inf)/np.clip(h-x[i],1e-10,np.inf))) for i,l,h in zip(self.constrained_bounded_indices, self.constrained_bounded_lowers, self.constrained_bounded_uppers)] @@ -233,7 +233,7 @@ class parameterised(object): def expand_param(self,x): - """ takes the vector x, which is then modified (by untying, reparameterising or inserting fixed values), and then call self.set_param""" + """ takes the vector x, which is then modified (by untying, reparameterising or inserting fixed values), and then call self._set_params""" #work out how many places are fixed, and where they are. tricky logic! Nfix_places = 0. @@ -257,14 +257,14 @@ class parameterised(object): xx[self.constrained_positive_indices] = np.exp(xx[self.constrained_positive_indices]) xx[self.constrained_negative_indices] = -np.exp(xx[self.constrained_negative_indices]) [np.put(xx,i,low+sigmoid(xx[i])*(high-low)) for i,low,high in zip(self.constrained_bounded_indices, self.constrained_bounded_lowers, self.constrained_bounded_uppers)] - self.set_param(xx) + self._set_params(xx) def extract_param_names(self): """ Returns the parameter names as propagated after constraining, tying or fixing, i.e. a list of the same length as extract_param() """ - n = self.get_param_names() + n = self._get_param_names() #remove/concatenate the tied parameter names if len(self.tied_indices): @@ -294,13 +294,13 @@ class parameterised(object): """ Return a string describing the parameter names and their ties and constraints """ - names = self.get_param_names() + names = self._get_param_names() N = len(names) if not N: return "This object has no free parameters." header = ['Name','Value','Constraints','Ties'] - values = self.get_param() #map(str,self.get_param()) + values = self._get_params() #map(str,self._get_params()) #sort out the constraints constraints = ['']*len(names) for i in self.constrained_positive_indices: diff --git a/GPy/examples/GPLVM_demo.py b/GPy/examples/GPLVM_demo.py deleted file mode 100644 index 651af6be..00000000 --- a/GPy/examples/GPLVM_demo.py +++ /dev/null @@ -1,28 +0,0 @@ -# Copyright (c) 2012, GPy authors (see AUTHORS.txt). -# Licensed under the BSD 3-clause license (see LICENSE.txt) - - -import numpy as np -import pylab as pb -import GPy -np.random.seed(1) -print "GPLVM with RBF kernel" - -N = 100 -Q = 1 -D = 2 -X = np.random.rand(N, Q) -k = GPy.kern.rbf(Q, 1.0, 2.0) + GPy.kern.white(Q, 0.00001) -K = k.K(X) -Y = np.random.multivariate_normal(np.zeros(N),K,D).T - -m = GPy.models.GPLVM(Y, Q) -m.constrain_positive('(rbf|bias|white)') - -pb.figure() -m.plot() -pb.title('PCA initialisation') -pb.figure() -m.optimize(messages = 1) -m.plot() -pb.title('After optimisation') diff --git a/GPy/examples/GP_regression_demo.py b/GPy/examples/GP_regression_demo.py deleted file mode 100644 index 0fe2fabd..00000000 --- a/GPy/examples/GP_regression_demo.py +++ /dev/null @@ -1,51 +0,0 @@ -# Copyright (c) 2012, GPy authors (see AUTHORS.txt). -# Licensed under the BSD 3-clause license (see LICENSE.txt) - - -""" -Simple Gaussian Processes regression with an RBF kernel -""" -import pylab as pb -import numpy as np -import GPy -pb.ion() -pb.close('all') - - -###################################### -## 1 dimensional example - -# sample inputs and outputs -X = np.random.uniform(-3.,3.,(20,1)) -Y = np.sin(X)+np.random.randn(20,1)*0.05 - -# create simple GP model -m = GPy.models.GP_regression(X,Y) - -# contrain all parameters to be positive -m.constrain_positive('') - -# optimize and plot -m.optimize('tnc', max_f_eval = 1000) -m.plot() -print(m) - -###################################### -## 2 dimensional example - -# sample inputs and outputs -X = np.random.uniform(-3.,3.,(40,2)) -Y = np.sin(X[:,0:1]) * np.sin(X[:,1:2])+np.random.randn(40,1)*0.05 - -# create simple GP model -m = GPy.models.GP_regression(X,Y) - -# contrain all parameters to be positive -m.constrain_positive('') -# optimize and plot -pb.figure() -m.optimize('tnc', max_f_eval = 1000) -m.plot() -print(m) - - diff --git a/GPy/examples/GP_regression_kern_demo.py b/GPy/examples/GP_regression_kern_demo.py deleted file mode 100644 index 8f0b9226..00000000 --- a/GPy/examples/GP_regression_kern_demo.py +++ /dev/null @@ -1,33 +0,0 @@ -# Copyright (c) 2012, GPy authors (see AUTHORS.txt). -# Licensed under the BSD 3-clause license (see LICENSE.txt) - - -""" -Simple one-dimensional Gaussian Processes with assorted kernel functions -""" -import pylab as pb -import numpy as np -import GPy - -# sample inputs and outputs -D = 1 -X = np.random.randn(10,D)*2 -X = np.linspace(-1.5,1.5,5)[:,None] -X = np.append(X,[[5]],0) -Y = np.sin(np.pi*X/2) #+np.random.randn(X.shape[0],1)*0.05 - -models = [GPy.models.GP_regression(X,Y, k) for k in (GPy.kern.rbf(D), GPy.kern.Matern52(D), GPy.kern.Matern32(D), GPy.kern.exponential(D), GPy.kern.linear(D) + GPy.kern.white(D), GPy.kern.bias(D) + GPy.kern.white(D))] - -pb.figure(figsize=(12,8)) -for i,m in enumerate(models): - m.constrain_positive('') - m.optimize() - pb.subplot(3,2,i+1) - m.plot() - #pb.title(m.kern.parts[0].name) - -GPy.util.plot.align_subplots(3,2,(-3,6),(-2.5,2.5)) - -pb.show() - - diff --git a/GPy/examples/classification.py b/GPy/examples/classification.py index eb7a887e..989ed08a 100644 --- a/GPy/examples/classification.py +++ b/GPy/examples/classification.py @@ -8,8 +8,6 @@ Simple Gaussian Processes classification import pylab as pb import numpy as np import GPy -pb.ion() -pb.close('all') default_seed=10000 ###################################### @@ -27,7 +25,7 @@ def crescent_data(model_type='Full', inducing=10, seed=default_seed): likelihood = GPy.inference.likelihoods.probit(data['Y']) if model_type=='Full': - m = GPy.models.simple_GP_EP(data['X'],likelihood) + m = GPy.models.GP_EP(data['X'],likelihood) else: # create sparse GP EP model m = GPy.models.sparse_GP_EP(data['X'],likelihood=likelihood,inducing=inducing,ep_proxy=model_type) @@ -49,7 +47,7 @@ def oil(): likelihood = GPy.inference.likelihoods.probit(data['Y'][:, 0:1]) # create simple GP model - m = GPy.models.simple_GP_EP(data['X'],likelihood) + m = GPy.models.GP_EP(data['X'],likelihood) # contrain all parameters to be positive m.constrain_positive('') diff --git a/GPy/examples/oil_flow_demo.py b/GPy/examples/oil_flow_demo.py deleted file mode 100644 index eee74461..00000000 --- a/GPy/examples/oil_flow_demo.py +++ /dev/null @@ -1,53 +0,0 @@ -# Copyright (c) 2012, GPy authors (see AUTHORS.txt). -# Licensed under the BSD 3-clause license (see LICENSE.txt) - - -import cPickle as pickle -import numpy as np -import pylab as pb -import GPy -import pylab as plt -np.random.seed(1) - -def plot_oil(X, theta, labels, label): - plt.figure() - X = X[:,np.argsort(theta)[:2]] - flow_type = (X[labels[:,0]==1]) - plt.plot(flow_type[:,0], flow_type[:,1], 'rx') - flow_type = (X[labels[:,1]==1]) - plt.plot(flow_type[:,0], flow_type[:,1], 'gx') - flow_type = (X[labels[:,2]==1]) - plt.plot(flow_type[:,0], flow_type[:,1], 'bx') - plt.title(label) - -data = pickle.load(open('../util/datasets/oil_flow_3classes.pickle', 'r')) - -Y = data['DataTrn'] -N, D = Y.shape -selected = np.random.permutation(N)[:200] -labels = data['DataTrnLbls'][selected] -Y = Y[selected] -N, D = Y.shape -Y -= Y.mean(axis=0) -Y /= Y.std(axis=0) - -Q = 2 -m1 = GPy.models.sparse_GPLVM(Y, Q, M = 15) -m1.constrain_positive('(rbf|bias|noise)') -m1.constrain_bounded('white', 1e-6, 1.0) - -plot_oil(m1.X, np.array([1,1]), labels, 'PCA initialization') -# m.optimize(messages = True) -m1.optimize('bfgs', messages = True) -plot_oil(m1.X, np.array([1,1]), labels, 'sparse GPLVM') -# pb.figure() -# m.plot() -# pb.title('PCA initialisation') -# pb.figure() -# m.optimize(messages = 1) -# m.plot() -# pb.title('After optimisation') -m = GPy.models.GPLVM(Y, Q) -m.constrain_positive('(white|rbf|bias|noise)') -m.optimize() -plot_oil(m.X, np.array([1,1]), labels, 'GPLVM') diff --git a/GPy/examples/regression.py b/GPy/examples/regression.py index d61c8b10..79763504 100644 --- a/GPy/examples/regression.py +++ b/GPy/examples/regression.py @@ -8,8 +8,6 @@ Gaussian Processes regression examples import pylab as pb import numpy as np import GPy -pb.ion() -pb.close('all') def toy_rbf_1d(): @@ -48,6 +46,10 @@ def rogers_girolami_olympics(): print(m) return m +def della_gatta_TRP63_gene_expression(number=942): + """Run a standard Gaussian process regression on the della Gatta et al TRP63 Gene Expression data set for a given gene number.""" + + def toy_rbf_1d_50(): """Run a simple demonstration of a standard Gaussian process fitting it to data sampled from an RBF covariance.""" data = GPy.util.datasets.toy_rbf_1d_50() @@ -81,3 +83,94 @@ def silhouette(): print(m) return m + + +def multiple_optima(gene_number=937,resolution=80, model_restarts=10, seed=10000): + """Show an example of a multimodal error surface for Gaussian process regression. Gene 939 has bimodal behaviour where the noisey mode is higher.""" + + # Contour over a range of length scales and signal/noise ratios. + length_scales = np.linspace(0.1, 60., resolution) + log_SNRs = np.linspace(-3., 4., resolution) + + data = GPy.util.datasets.della_gatta_TRP63_gene_expression(gene_number) + # Sub sample the data to ensure multiple optima + #data['Y'] = data['Y'][0::2, :] + #data['X'] = data['X'][0::2, :] + + # Remove the mean (no bias kernel to ensure signal/noise is in RBF/white) + data['Y'] = data['Y'] - np.mean(data['Y']) + + lls = GPy.examples.regression.contour_data(data, length_scales, log_SNRs, GPy.kern.rbf) + pb.contour(length_scales, log_SNRs, np.exp(lls), 20) + ax = pb.gca() + pb.xlabel('length scale') + pb.ylabel('log_10 SNR') + + xlim = ax.get_xlim() + ylim = ax.get_ylim() + + # Now run a few optimizations + models = [] + optim_point_x = np.empty(2) + optim_point_y = np.empty(2) + np.random.seed(seed=seed) + for i in range(0, model_restarts): + kern = GPy.kern.rbf(1, variance=np.random.exponential(1.), lengthscale=np.random.exponential(50.)) + GPy.kern.white(1,variance=np.random.exponential(1.)) + + m = GPy.models.GP_regression(data['X'],data['Y'], kernel=kern) + params = m._get_params() + optim_point_x[0] = params[1] + optim_point_y[0] = np.log10(params[0]) - np.log10(params[2]); + + # contrain all parameters to be positive + m.constrain_positive('') + + # optimize + m.optimize(xtol=1e-6,ftol=1e-6) + + params = m._get_params() + optim_point_x[1] = params[1] + optim_point_y[1] = np.log10(params[0]) - np.log10(params[2]); + print(m) + + pb.arrow(optim_point_x[0], optim_point_y[0], optim_point_x[1]-optim_point_x[0], optim_point_y[1]-optim_point_y[0], label=str(i), head_length=1, head_width=0.5, fc='k', ec='k') + models.append(m) + + ax.set_xlim(xlim) + ax.set_ylim(ylim) + return (models, lls) + +def contour_data(data, length_scales, log_SNRs, signal_kernel_call=GPy.kern.rbf): + """Evaluate the GP objective function for a given data set for a range of signal to noise ratios and a range of lengthscales. + + :data_set: A data set from the utils.datasets director. + :length_scales: a list of length scales to explore for the contour plot. + :log_SNRs: a list of base 10 logarithm signal to noise ratios to explore for the contour plot. + :signal_kernel: a kernel to use for the 'signal' portion of the data.""" + + lls = [] + total_var = np.var(data['Y']) + for log_SNR in log_SNRs: + SNR = 10**log_SNR + length_scale_lls = [] + for length_scale in length_scales: + noise_var = 1. + signal_var = SNR + noise_var = noise_var/(noise_var + signal_var)*total_var + signal_var = signal_var/(noise_var + signal_var)*total_var + + signal_kernel = signal_kernel_call(1, variance=signal_var, lengthscale=length_scale) + noise_kernel = GPy.kern.white(1, variance=noise_var) + kernel = signal_kernel + noise_kernel + K = kernel.K(data['X']) + total_var = (np.dot(np.dot(data['Y'].T,GPy.util.linalg.pdinv(K)[0]), data['Y'])/data['Y'].shape[0])[0,0] + noise_var *= total_var + signal_var *= total_var + + kernel = signal_kernel_call(1, variance=signal_var, lengthscale=length_scale) + GPy.kern.white(1, variance=noise_var) + + model = GPy.models.GP_regression(data['X'], data['Y'], kernel=kernel) + model.constrain_positive('') + length_scale_lls.append(model.log_likelihood()) + lls.append(length_scale_lls) + return np.array(lls) diff --git a/GPy/examples/warped_GP_demo.py b/GPy/examples/warped_GP_demo.py deleted file mode 100644 index 3b75694a..00000000 --- a/GPy/examples/warped_GP_demo.py +++ /dev/null @@ -1,45 +0,0 @@ -# Copyright (c) 2012, GPy authors (see AUTHORS.txt). -# Licensed under the BSD 3-clause license (see LICENSE.txt) - - -import numpy as np -import scipy as sp -import pdb, sys, pickle -import matplotlib.pylab as plt -import GPy -np.random.seed(1) - -N = 100 -# sample inputs and outputs -X = np.random.uniform(-np.pi,np.pi,(N,1)) -Y = np.sin(X)+np.random.randn(N,1)*0.05 -# Y += np.abs(Y.min()) + 0.5 -Z = np.exp(Y)# Y**(1/3.0) - -# rescaling targets? -Zmax = Z.max() -Zmin = Z.min() -Z = (Z-Zmin)/(Zmax-Zmin) - 0.5 - -m = GPy.models.warpedGP(X, Z, warping_terms = 2) -m.constrain_positive('(tanh_a|tanh_b|rbf|white|bias)') -m.randomize() -plt.figure() -plt.xlabel('predicted f(Z)') -plt.ylabel('actual f(Z)') -plt.plot(m.Y, Y, 'o', alpha = 0.5, label = 'before training') -m.optimize(messages = True) -plt.plot(m.Y, Y, 'o', alpha = 0.5, label = 'after training') -plt.legend(loc = 0) -m.plot_warping() -plt.figure() -plt.title('warped GP fit') -m.plot() - -m1 = GPy.models.GP_regression(X, Z) -m1.constrain_positive('(rbf|white|bias)') -m1.randomize() -m1.optimize(messages = True) -plt.figure() -plt.title('GP fit') -m1.plot() diff --git a/GPy/inference/samplers.py b/GPy/inference/samplers.py index 1216f1eb..da159c23 100644 --- a/GPy/inference/samplers.py +++ b/GPy/inference/samplers.py @@ -73,12 +73,12 @@ class Metropolis_Hastings: def predict(self,function,args): """Make a prediction for the function, to which we will pass the additional arguments""" - param = self.model.get_param() + param = self.model._get_params() fs = [] for p in self.chain: - self.model.set_param(p) + self.model._set_params(p) fs.append(function(*args)) - self.model.set_param(param)# reset model to starting state + self.model._set_params(param)# reset model to starting state return fs diff --git a/GPy/kern/Brownian.py b/GPy/kern/Brownian.py index 07d5fbd6..fd19137c 100644 --- a/GPy/kern/Brownian.py +++ b/GPy/kern/Brownian.py @@ -23,16 +23,16 @@ class Brownian(kernpart): assert self.D==1, "Brownian motion in 1D only" self.Nparam = 1. self.name = 'Brownian' - self.set_param(np.array([variance]).flatten()) + self._set_params(np.array([variance]).flatten()) - def get_param(self): + def _get_params(self): return self.variance - def set_param(self,x): + def _set_params(self,x): assert x.shape==(1,) self.variance = x - def get_param_names(self): + def _get_param_names(self): return ['variance'] def K(self,X,X2,target): diff --git a/GPy/kern/Matern32.py b/GPy/kern/Matern32.py index 377b687f..6517ac2c 100644 --- a/GPy/kern/Matern32.py +++ b/GPy/kern/Matern32.py @@ -34,19 +34,19 @@ class Matern32(kernpart): lengthscales = np.ones(self.D) self.Nparam = self.D + 1 self.name = 'Mat32' - self.set_param(np.hstack((variance,lengthscales))) + self._set_params(np.hstack((variance,lengthscales))) - def get_param(self): + def _get_params(self): """return the value of the parameters.""" return np.hstack((self.variance,self.lengthscales)) - def set_param(self,x): + def _set_params(self,x): """set the value of the parameters.""" assert x.size==(self.D+1) self.variance = x[0] self.lengthscales = x[1:] - def get_param_names(self): + def _get_param_names(self): """return parameter names.""" if self.D==1: return ['variance','lengthscale'] diff --git a/GPy/kern/Matern52.py b/GPy/kern/Matern52.py index 172c0117..056da32d 100644 --- a/GPy/kern/Matern52.py +++ b/GPy/kern/Matern52.py @@ -32,19 +32,19 @@ class Matern52(kernpart): lengthscales = np.ones(self.D) self.Nparam = self.D + 1 self.name = 'Mat52' - self.set_param(np.hstack((variance,lengthscales))) + self._set_params(np.hstack((variance,lengthscales))) - def get_param(self): + def _get_params(self): """return the value of the parameters.""" return np.hstack((self.variance,self.lengthscales)) - def set_param(self,x): + def _set_params(self,x): """set the value of the parameters.""" assert x.size==(self.D+1) self.variance = x[0] self.lengthscales = x[1:] - def get_param_names(self): + def _get_param_names(self): """return parameter names.""" if self.D==1: return ['variance','lengthscale'] diff --git a/GPy/kern/bias.py b/GPy/kern/bias.py index a89f56c5..0a6872a5 100644 --- a/GPy/kern/bias.py +++ b/GPy/kern/bias.py @@ -17,16 +17,16 @@ class bias(kernpart): self.D = D self.Nparam = 1 self.name = 'bias' - self.set_param(np.array([variance]).flatten()) + self._set_params(np.array([variance]).flatten()) - def get_param(self): + def _get_params(self): return self.variance - def set_param(self,x): + def _set_params(self,x): assert x.shape==(1,) self.variance = x - def get_param_names(self): + def _get_param_names(self): return ['variance'] def K(self,X,X2,target): diff --git a/GPy/kern/exponential.py b/GPy/kern/exponential.py index 8ad5b813..2df6a958 100644 --- a/GPy/kern/exponential.py +++ b/GPy/kern/exponential.py @@ -32,19 +32,19 @@ class exponential(kernpart): lengthscales = np.ones(self.D) self.Nparam = self.D + 1 self.name = 'exp' - self.set_param(np.hstack((variance,lengthscales))) + self._set_params(np.hstack((variance,lengthscales))) - def get_param(self): + def _get_params(self): """return the value of the parameters.""" return np.hstack((self.variance,self.lengthscales)) - def set_param(self,x): + def _set_params(self,x): """set the value of the parameters.""" assert x.size==(self.D+1) self.variance = x[0] self.lengthscales = x[1:] - def get_param_names(self): + def _get_param_names(self): """return parameter names.""" if self.D==1: return ['variance','lengthscale'] diff --git a/GPy/kern/finite_dimensional.py b/GPy/kern/finite_dimensional.py index 98c99628..36afd1dc 100644 --- a/GPy/kern/finite_dimensional.py +++ b/GPy/kern/finite_dimensional.py @@ -27,15 +27,15 @@ class finite_dimensional(kernpart): weights = np.ones(self.n) self.Nparam = self.n + 1 self.name = 'finite_dim' - self.set_param(np.hstack((variance,weights))) + self._set_params(np.hstack((variance,weights))) - def get_param(self): + def _get_params(self): return np.hstack((self.variance,self.weights)) - def set_param(self,x): + def _set_params(self,x): assert x.size == (self.Nparam) self.variance = x[0] self.weights = x[1:] - def get_param_names(self): + def _get_param_names(self): if self.n==1: return ['variance','weight'] else: diff --git a/GPy/kern/kern.py b/GPy/kern/kern.py index 3ba7d97b..5f259f55 100644 --- a/GPy/kern/kern.py +++ b/GPy/kern/kern.py @@ -133,20 +133,20 @@ class kern(parameterised): newkern.tied_indices = self.tied_indices + [self.Nparam + x for x in other.tied_indices] return newkern - def get_param(self): - return np.hstack([p.get_param() for p in self.parts]) + def _get_params(self): + return np.hstack([p._get_params() for p in self.parts]) - def set_param(self,x): - [p.set_param(x[s]) for p, s in zip(self.parts, self.param_slices)] + def _set_params(self,x): + [p._set_params(x[s]) for p, s in zip(self.parts, self.param_slices)] - def get_param_names(self): + def _get_param_names(self): #this is a bit nasty: we wat to distinguish between parts with the same name by appending a count part_names = np.array([k.name for k in self.parts],dtype=np.str) counts = [np.sum(part_names==ni) for i, ni in enumerate(part_names)] cum_counts = [np.sum(part_names[i:]==ni) for i, ni in enumerate(part_names)] names = [name+'_'+str(cum_count) if count>1 else name for name,count,cum_count in zip(part_names,counts,cum_counts)] - return sum([[name+'_'+n for n in k.get_param_names()] for name,k in zip(names,self.parts)],[]) + return sum([[name+'_'+n for n in k._get_param_names()] for name,k in zip(names,self.parts)],[]) def K(self,X,X2=None,slices1=None,slices2=None): assert X.shape[1]==self.D diff --git a/GPy/kern/kernpart.py b/GPy/kern/kernpart.py index d06749b8..3a5486de 100644 --- a/GPy/kern/kernpart.py +++ b/GPy/kern/kernpart.py @@ -16,11 +16,11 @@ class kernpart(object): self.Nparam = 1 self.name = 'unnamed' - def get_param(self): + def _get_params(self): raise NotImplementedError - def set_param(self,x): + def _set_params(self,x): raise NotImplementedError - def get_param_names(self): + def _get_param_names(self): raise NotImplementedError def K(self,X,X2,target): raise NotImplementedError diff --git a/GPy/kern/linear.py b/GPy/kern/linear.py index 7246244e..14c0ece3 100644 --- a/GPy/kern/linear.py +++ b/GPy/kern/linear.py @@ -20,16 +20,16 @@ class linear(kernpart): variance = 1.0 self.Nparam = 1 self.name = 'linear' - self.set_param(variance) + self._set_params(variance) self._Xcache, self._X2cache = np.empty(shape=(2,)) - def get_param(self): + def _get_params(self): return self.variance - def set_param(self,x): + def _set_params(self,x): self.variance = x - def get_param_names(self): + def _get_param_names(self): return ['variance'] def K(self,X,X2,target): diff --git a/GPy/kern/linear_ARD.py b/GPy/kern/linear_ARD.py index 149bfb4d..b9112044 100644 --- a/GPy/kern/linear_ARD.py +++ b/GPy/kern/linear_ARD.py @@ -23,16 +23,16 @@ class linear_ARD(kernpart): variances = np.ones(self.D) self.Nparam = int(self.D) self.name = 'linear' - self.set_param(variances) + self._set_params(variances) - def get_param(self): + def _get_params(self): return self.variances - def set_param(self,x): + def _set_params(self,x): assert x.size==(self.Nparam) self.variances = x - def get_param_names(self): + def _get_param_names(self): if self.D==1: return ['variance'] else: diff --git a/GPy/kern/rbf.py b/GPy/kern/rbf.py index 9e2bb509..c929369f 100644 --- a/GPy/kern/rbf.py +++ b/GPy/kern/rbf.py @@ -30,23 +30,23 @@ class rbf(kernpart): self.D = D self.Nparam = 2 self.name = 'rbf' - self.set_param(np.hstack((variance,lengthscale))) + self._set_params(np.hstack((variance,lengthscale))) #initialize cache self._Z, self._mu, self._S = np.empty(shape=(3,1)) self._X, self._X2, self._params = np.empty(shape=(3,1)) - def get_param(self): + def _get_params(self): return np.hstack((self.variance,self.lengthscale)) - def set_param(self,x): + def _set_params(self,x): self.variance, self.lengthscale = x self.lengthscale2 = np.square(self.lengthscale) #reset cached results self._X, self._X2, self._params = np.empty(shape=(3,1)) self._Z, self._mu, self._S = np.empty(shape=(3,1)) # cached versions of Z,mu,S - def get_param_names(self): + def _get_param_names(self): return ['variance','lengthscale'] def K(self,X,X2,target): diff --git a/GPy/kern/rbf_ARD.py b/GPy/kern/rbf_ARD.py index 732be590..12924b3f 100644 --- a/GPy/kern/rbf_ARD.py +++ b/GPy/kern/rbf_ARD.py @@ -22,16 +22,16 @@ class rbf_ARD(kernpart): lengthscales = np.ones(self.D) self.Nparam = self.D + 1 self.name = 'rbf_ARD' - self.set_param(np.hstack((variance,lengthscales))) + self._set_params(np.hstack((variance,lengthscales))) #initialize cache self._Z, self._mu, self._S = np.empty(shape=(3,1)) self._X, self._X2, self._params = np.empty(shape=(3,1)) - def get_param(self): + def _get_params(self): return np.hstack((self.variance,self.lengthscales)) - def set_param(self,x): + def _set_params(self,x): assert x.size==(self.D+1) self.variance = x[0] self.lengthscales = x[1:] @@ -39,7 +39,7 @@ class rbf_ARD(kernpart): #reset cached results self._Z, self._mu, self._S = np.empty(shape=(3,1)) # cached versions of Z,mu,S - def get_param_names(self): + def _get_param_names(self): if self.D==1: return ['variance','lengthscale'] else: @@ -135,8 +135,8 @@ class rbf_ARD(kernpart): if X2 is None: X2 = X self._K_dist = X[:,None,:]-X2[None,:,:] # this can be computationally heavy self._params = np.empty(shape=(1,0))#ensure the next section gets called - if not np.all(self._params == self.get_param()): - self._params == self.get_param() + if not np.all(self._params == self._get_params()): + self._params == self._get_params() self._K_dist2 = np.square(self._K_dist/self.lengthscales) self._K_exponent = -0.5*self._K_dist2.sum(-1) self._K_dvar = np.exp(-0.5*self._K_dist2.sum(-1)) @@ -189,7 +189,7 @@ if __name__=='__main__': from checkgrad import checkgrad def k_theta_test(param,k): - k.set_param(param) + k._set_params(param) K = k.K(Z) dK_dtheta = k.dK_dtheta(Z) f = np.sum(K) @@ -216,7 +216,7 @@ if __name__=='__main__': checkgrad(psi1_S_test,np.random.rand(N*Q),args=(k,)) def psi1_theta_test(theta,k): - k.set_param(theta) + k._set_params(theta) f = np.sum(k.psi1(Z,mu,S)) df = np.array([np.sum(grad) for grad in k.dpsi1_dtheta(Z,mu,S)]) return f,df @@ -241,7 +241,7 @@ if __name__=='__main__': checkgrad(psi2_S_test,np.random.rand(N*Q),args=(k,)) def psi2_theta_test(theta,k): - k.set_param(theta) + k._set_params(theta) f = np.sum(k.psi2(Z,mu,S)) df = np.array([np.sum(grad) for grad in k.dpsi2_dtheta(Z,mu,S)]) return f,df diff --git a/GPy/kern/spline.py b/GPy/kern/spline.py index 44033eea..030b2f02 100644 --- a/GPy/kern/spline.py +++ b/GPy/kern/spline.py @@ -25,15 +25,15 @@ class spline(kernpart): assert self.D==1 self.Nparam = 1 self.name = 'spline' - self.set_param(np.squeeze(variance)) + self._set_params(np.squeeze(variance)) - def get_param(self): + def _get_params(self): return self.variance - def set_param(self,x): + def _set_params(self,x): self.variance = x - def get_param_names(self): + def _get_param_names(self): return ['variance'] def K(self,X,X2,target): diff --git a/GPy/kern/sympykern.py b/GPy/kern/sympykern.py index d9f89f5b..db3cc976 100644 --- a/GPy/kern/sympykern.py +++ b/GPy/kern/sympykern.py @@ -44,7 +44,7 @@ class spkern(kernpart): if param is None: param = np.ones(self.Nparam) assert param.size==self.Nparam - self.set_param(param) + self._set_params(param) #Differentiate! self._sp_dk_dtheta = [sp.diff(k,theta).simplify() for theta in self._sp_theta] @@ -247,12 +247,12 @@ class spkern(kernpart): Z = X weave.inline(self._dKdiag_dX_code,arg_names=['target','X','Z','param','partial'],**self.weave_kwargs) - def set_param(self,param): + def _set_params(self,param): #print param.flags['C_CONTIGUOUS'] self._param = param.copy() - def get_param(self): + def _get_params(self): return self._param - def get_param_names(self): + def _get_param_names(self): return [x.name for x in self._sp_theta] diff --git a/GPy/kern/white.py b/GPy/kern/white.py index 587a2b4a..b3b00c48 100644 --- a/GPy/kern/white.py +++ b/GPy/kern/white.py @@ -17,16 +17,16 @@ class white(kernpart): self.D = D self.Nparam = 1 self.name = 'white' - self.set_param(np.array([variance]).flatten()) + self._set_params(np.array([variance]).flatten()) - def get_param(self): + def _get_params(self): return self.variance - def set_param(self,x): + def _set_params(self,x): assert x.shape==(1,) self.variance = x - def get_param_names(self): + def _get_param_names(self): return ['variance'] def K(self,X,X2,target): diff --git a/GPy/models/GPLVM.py b/GPy/models/GPLVM.py index 44147b73..0a168f25 100644 --- a/GPy/models/GPLVM.py +++ b/GPy/models/GPLVM.py @@ -33,16 +33,16 @@ class GPLVM(GP_regression): else: return np.random.randn(Y.shape[0], Q) - def get_param_names(self): + def _get_param_names(self): return (sum([['X_%i_%i'%(n,q) for n in range(self.N)] for q in range(self.Q)],[]) + self.kern.extract_param_names()) - def get_param(self): + def _get_params(self): return np.hstack((self.X.flatten(), self.kern.extract_param())) - def set_param(self,x): + def _set_params(self,x): self.X = x[:self.X.size].reshape(self.N,self.Q).copy() - GP_regression.set_param(self, x[self.X.size:]) + GP_regression._set_params(self, x[self.X.size:]) def log_likelihood_gradients(self): dL_dK = self.dL_dK() diff --git a/GPy/models/GP_EP.py b/GPy/models/GP_EP.py index bb582674..5cb2ba45 100644 --- a/GPy/models/GP_EP.py +++ b/GPy/models/GP_EP.py @@ -41,13 +41,13 @@ class GP_EP(model): self.K = self.kernel.K(self.X) model.__init__(self) - def set_param(self,p): + def _set_params(self,p): self.kernel.expand_param(p) - def get_param(self): + def _get_params(self): return self.kernel.extract_param() - def get_param_names(self): + def _get_param_names(self): return self.kernel.extract_param_names() def approximate_likelihood(self): @@ -138,7 +138,7 @@ class GP_EP(model): """ self.epsilon_em = epsilon log_likelihood_change = self.epsilon_em + 1. - self.parameters_path = [self.kernel.get_param()] + self.parameters_path = [self.kernel._get_params()] self.approximate_likelihood() self.site_approximations_path = [[self.ep_approx.tau_tilde,self.ep_approx.v_tilde]] self.log_likelihood_path = [self.log_likelihood()] @@ -155,6 +155,6 @@ class GP_EP(model): else: self.approximate_likelihood() self.log_likelihood_path.append(self.log_likelihood()) - self.parameters_path.append(self.kernel.get_param()) + self.parameters_path.append(self.kernel._get_params()) self.site_approximations_path.append([self.ep_approx.tau_tilde,self.ep_approx.v_tilde]) iteration += 1 diff --git a/GPy/models/GP_regression.py b/GPy/models/GP_regression.py index ee2bb448..da657950 100644 --- a/GPy/models/GP_regression.py +++ b/GPy/models/GP_regression.py @@ -70,15 +70,15 @@ class GP_regression(model): model.__init__(self) - def set_param(self,p): + def _set_params(self,p): self.kern.expand_param(p) self.K = self.kern.K(self.X,slices1=self.Xslices) self.Ki, self.L, self.Li, self.K_logdet = pdinv(self.K) - def get_param(self): + def _get_params(self): return self.kern.extract_param() - def get_param_names(self): + def _get_param_names(self): return self.kern.extract_param_names() def _model_fit_term(self): diff --git a/GPy/models/generalized_FITC.py b/GPy/models/generalized_FITC.py index 4ef36f0e..40f18ff9 100644 --- a/GPy/models/generalized_FITC.py +++ b/GPy/models/generalized_FITC.py @@ -42,14 +42,14 @@ class generalized_FITC(model): self.jitter = 1e-12 model.__init__(self) - def set_param(self,p): + def _set_params(self,p): self.kernel.expand_param(p[0:-self.Z.size]) self.Z = p[-self.Z.size:].reshape(self.M,self.D) - def get_param(self): + def _get_params(self): return np.hstack([self.kernel.extract_param(),self.Z.flatten()]) - def get_param_names(self): + def _get_param_names(self): return self.kernel.extract_param_names()+['iip_%i'%i for i in range(self.Z.size)] def approximate_likelihood(self): @@ -214,7 +214,7 @@ class generalized_FITC(model): """ self.epsilon_em = epsilon log_likelihood_change = self.epsilon_em + 1. - self.parameters_path = [self.kernel.get_param()] + self.parameters_path = [self.kernel._get_params()] self.approximate_likelihood() self.site_approximations_path = [[self.ep_approx.tau_tilde,self.ep_approx.v_tilde]] self.inducing_inputs_path = [self.Z] @@ -235,7 +235,7 @@ class generalized_FITC(model): else: self.approximate_likelihood() self.log_likelihood_path.append(self.log_likelihood()) - self.parameters_path.append(self.kernel.get_param()) + self.parameters_path.append(self.kernel._get_params()) self.site_approximations_path.append([self.ep_approx.tau_tilde,self.ep_approx.v_tilde]) self.inducing_inputs_path.append(self.Z) iteration += 1 diff --git a/GPy/models/sparse_GPLVM.py b/GPy/models/sparse_GPLVM.py index c5125c85..1c6a03cf 100644 --- a/GPy/models/sparse_GPLVM.py +++ b/GPy/models/sparse_GPLVM.py @@ -27,16 +27,16 @@ class sparse_GPLVM(sparse_GP_regression, GPLVM): X = self.initialise_latent(init, Q, Y) sparse_GP_regression.__init__(self, X, Y, **kwargs) - def get_param_names(self): + def _get_param_names(self): return (sum([['X_%i_%i'%(n,q) for n in range(self.N)] for q in range(self.Q)],[]) - + sparse_GP_regression.get_param_names(self)) + + sparse_GP_regression._get_param_names(self)) - def get_param(self): - return np.hstack((self.X.flatten(), sparse_GP_regression.get_param(self))) + def _get_params(self): + return np.hstack((self.X.flatten(), sparse_GP_regression._get_params(self))) - def set_param(self,x): + def _set_params(self,x): self.X = x[:self.X.size].reshape(self.N,self.Q).copy() - sparse_GP_regression.set_param(self, x[self.X.size:]) + sparse_GP_regression._set_params(self, x[self.X.size:]) def log_likelihood(self): return sparse_GP_regression.log_likelihood(self) diff --git a/GPy/models/sparse_GP_regression.py b/GPy/models/sparse_GP_regression.py index fe5f7cc1..fc04d5ed 100644 --- a/GPy/models/sparse_GP_regression.py +++ b/GPy/models/sparse_GP_regression.py @@ -59,10 +59,10 @@ class sparse_GP_regression(GP_regression): if self.has_uncertain_inputs: self.X_uncertainty /= np.square(self._Xstd) - def set_param(self, p): + def _set_params(self, p): self.Z = p[:self.M*self.Q].reshape(self.M, self.Q) self.beta = p[self.M*self.Q] - self.kern.set_param(p[self.Z.size + 1:]) + self.kern._set_params(p[self.Z.size + 1:]) self.beta2 = self.beta**2 self._compute_kernel_matrices() self._computations() @@ -106,10 +106,10 @@ class sparse_GP_regression(GP_regression): self.dL_dKmm += -0.5 * self.D * (- self.LBL_inv - 2.*self.beta*mdot(self.LBL_inv, self.psi2, self.Kmmi) + self.Kmmi) # dC self.dL_dKmm += np.dot(np.dot(self.G,self.beta*self.psi2) - np.dot(self.LBL_inv, self.psi1VVpsi1), self.Kmmi) + 0.5*self.G # dE - def get_param(self): + def _get_params(self): return np.hstack([self.Z.flatten(),self.beta,self.kern.extract_param()]) - def get_param_names(self): + def _get_param_names(self): return sum([['iip_%i_%i'%(i,j) for i in range(self.Z.shape[0])] for j in range(self.Z.shape[1])],[]) + ['noise_precision']+self.kern.extract_param_names() def log_likelihood(self): diff --git a/GPy/models/warped_GP.py b/GPy/models/warped_GP.py index bf5af21f..ea3c44cc 100644 --- a/GPy/models/warped_GP.py +++ b/GPy/models/warped_GP.py @@ -13,7 +13,7 @@ from GP_regression import GP_regression class warpedGP(GP_regression): """ - TODO: fucking docstrings! + TODO: fecking docstrings! @nfusi: I'#ve hacked a little on this, but no guarantees. J. """ @@ -30,17 +30,17 @@ class warpedGP(GP_regression): self.transform_data() GP_regression.__init__(self, X, self.Y, **kwargs) - def set_param(self, x): + def _set_params(self, x): self.warping_params = x[:self.warping_function.num_parameters].reshape(self.warp_params_shape).copy() self.transform_data() - GP_regression.set_param(self, x[self.warping_function.num_parameters:].copy()) + GP_regression._set_params(self, x[self.warping_function.num_parameters:].copy()) - def get_param(self): - return np.hstack((self.warping_params.flatten().copy(), GP_regression.get_param(self).copy())) + def _get_params(self): + return np.hstack((self.warping_params.flatten().copy(), GP_regression._get_params(self).copy())) - def get_param_names(self): - warping_names = self.warping_function.get_param_names() - param_names = GP_regression.get_param_names(self) + def _get_param_names(self): + warping_names = self.warping_function._get_param_names() + param_names = GP_regression._get_param_names(self) return warping_names + param_names def transform_data(self): @@ -81,7 +81,7 @@ class warpedGP(GP_regression): def predict(self, X, in_unwarped_space = False, **kwargs): mu, var = GP_regression.predict(self, X, **kwargs) - # The plot() function calls set_param() before calling predict() + # The plot() function calls _set_params() before calling predict() # this is causing the observations to be plotted in the transformed # space (where Y lives), making the plot looks very wrong # if the predictions are made in the untransformed space diff --git a/GPy/util/warping_functions.py b/GPy/util/warping_functions.py index 5a6dc6b3..a87deb5b 100644 --- a/GPy/util/warping_functions.py +++ b/GPy/util/warping_functions.py @@ -33,7 +33,7 @@ class WarpingFunction(object): """inverse function transformation""" raise NotImplementedError - def get_param_names(self): + def _get_param_names(self): raise NotImplementedError def plot(self, psi, xmin, xmax): @@ -151,7 +151,7 @@ class TanhWarpingFunction(WarpingFunction): return gradients - def get_param_names(self): + def _get_param_names(self): variables = ['a', 'b', 'c'] names = sum([['warp_tanh_%s_t%i' % (variables[n],q) for n in range(3)] for q in range(self.n_terms)],[]) return names From 377cc383be80852416a630d35d48e3c860cd7425 Mon Sep 17 00:00:00 2001 From: Alan Saul Date: Fri, 18 Jan 2013 12:39:54 +0000 Subject: [PATCH 06/36] Added apt-get scipy installation for travis, need to ensure version --- .travis.yml | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/.travis.yml b/.travis.yml index 921cf083..af20bec4 100644 --- a/.travis.yml +++ b/.travis.yml @@ -3,8 +3,7 @@ python: - "2.7" # command to install dependencies, e.g. pip install -r requirements.txt --use-mirrors install: - - pip install "scipy>=0.10.1" - - pip install "numpy>=1.6.2" + - sudo apt-get install python-scipy - pip install sphinx - pip install . --use-mirrors # command to run tests, e.g. python setup.py test From 1a795227dff689f0b3cee9de0ffca365074319a7 Mon Sep 17 00:00:00 2001 From: Alan Saul Date: Fri, 18 Jan 2013 13:07:14 +0000 Subject: [PATCH 07/36] Removed latent force model fortran code from setup.py from github code --- setup.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/setup.py b/setup.py index 076af82c..497429d0 100644 --- a/setup.py +++ b/setup.py @@ -24,8 +24,8 @@ setup(name = 'GPy', package_data = {'GPy': ['GPy/examples']}, py_modules = ['GPy.__init__'], long_description=read('README.md'), - ext_modules = [Extension(name = 'GPy.kern.lfmUpsilonf2py', - sources = ['GPy/kern/src/lfmUpsilonf2py.f90'])], + #ext_modules = [Extension(name = 'GPy.kern.lfmUpsilonf2py', + # sources = ['GPy/kern/src/lfmUpsilonf2py.f90'])], install_requires=['numpy>=1.6', 'scipy>=0.9','matplotlib>=1.1'], setup_requires=['sphinx'], cmdclass = {'build_sphinx': BuildDoc}, From f40b31bea72b4864753baccbc61754d705c20d60 Mon Sep 17 00:00:00 2001 From: Alan Saul Date: Fri, 18 Jan 2013 13:33:49 +0000 Subject: [PATCH 08/36] Removed version specification of scipy --- setup.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/setup.py b/setup.py index 497429d0..ee44dde0 100644 --- a/setup.py +++ b/setup.py @@ -26,7 +26,7 @@ setup(name = 'GPy', long_description=read('README.md'), #ext_modules = [Extension(name = 'GPy.kern.lfmUpsilonf2py', # sources = ['GPy/kern/src/lfmUpsilonf2py.f90'])], - install_requires=['numpy>=1.6', 'scipy>=0.9','matplotlib>=1.1'], + install_requires=['numpy>=1.6', 'scipy','matplotlib>=1.1'], setup_requires=['sphinx'], cmdclass = {'build_sphinx': BuildDoc}, classifiers=[ From 1674bc529bc49949fc39fe992ad5c487f942251d Mon Sep 17 00:00:00 2001 From: Neil Lawrence Date: Fri, 18 Jan 2013 13:37:17 +0000 Subject: [PATCH 09/36] expand_param and extract_param replaced with set_params_transformed and get_params_transformed --- GPy/core/model.py | 56 +++++++++++++++--------------- GPy/core/parameterised.py | 12 +++---- GPy/inference/samplers.py | 8 ++--- GPy/models/GPLVM.py | 4 +-- GPy/models/GP_EP.py | 10 +++--- GPy/models/GP_regression.py | 6 ++-- GPy/models/generalized_FITC.py | 8 ++--- GPy/models/sparse_GP_regression.py | 4 +-- 8 files changed, 54 insertions(+), 54 deletions(-) diff --git a/GPy/core/model.py b/GPy/core/model.py index d3c6e582..2bfb02a5 100644 --- a/GPy/core/model.py +++ b/GPy/core/model.py @@ -115,7 +115,7 @@ class model(parameterised): [np.put(ret,i,p.lnpdf_grad(xx)) for i,(p,xx) in enumerate(zip(self.priors,x)) if not p is None] return ret - def extract_gradients(self): + def _log_likelihood_gradients_transformed(self): """ Use self.log_likelihood_gradients and self.prior_gradients to get the gradients of the model. Adjust the gradient for constraints and ties, return. @@ -138,14 +138,14 @@ class model(parameterised): Make this draw from the prior if one exists, else draw from N(0,1) """ #first take care of all parameters (from N(0,1)) - x = self.extract_param() + x = self._get_params_transformed() x = np.random.randn(x.size) - self.expand_param(x) + self._set_params_transformed(x) #now draw from prior where possible x = self._get_params() [np.put(x,i,p.rvs(1)) for i,p in enumerate(self.priors) if not p is None] self._set_params(x) - self.expand_param(self.extract_param())#makes sure all of the tied parameters get the same init (since there's only one prior object...) + self._set_params_transformed(self._get_params_transformed())#makes sure all of the tied parameters get the same init (since there's only one prior object...) def optimize_restarts(self, Nrestarts=10, robust=False, verbose=True, **kwargs): @@ -165,7 +165,7 @@ class model(parameterised): :verbose: whether to show informations about the current restart """ - initial_parameters = self.extract_param() + initial_parameters = self._get_params_transformed() for i in range(Nrestarts): try: self.randomize() @@ -182,9 +182,9 @@ class model(parameterised): raise e if len(self.optimization_runs): i = np.argmax([o.f_opt for o in self.optimization_runs]) - self.expand_param(self.optimization_runs[i].x_opt) + self._set_params_transformed(self.optimization_runs[i].x_opt) else: - self.expand_param(initial_parameters) + self._set_params_transformed(initial_parameters) def ensure_default_constraints(self,warn=False): """ @@ -214,24 +214,24 @@ class model(parameterised): optimizer = self.preferred_optimizer def f(x): - self.expand_param(x) + self._set_params_transformed(x) return -self.log_likelihood()-self.log_prior() def fp(x): - self.expand_param(x) - return -self.extract_gradients() + self._set_params_transformed(x) + return -self._log_likelihood_gradients_transformed() def f_fp(x): - self.expand_param(x) - return -self.log_likelihood()-self.log_prior(),-self.extract_gradients() + self._set_params_transformed(x) + return -self.log_likelihood()-self.log_prior(),-self._log_likelihood_gradients_transformed() if start == None: - start = self.extract_param() + start = self._get_params_transformed() optimizer = optimization.get_optimizer(optimizer) opt = optimizer(start, model = self, **kwargs) opt.run(f_fp=f_fp, f=f, fp=fp) self.optimization_runs.append(opt) - self.expand_param(opt.x_opt) + self._set_params_transformed(opt.x_opt) def optimize_SGD(self, momentum = 0.1, learning_rate = 0.01, iterations = 20, **kwargs): # assert self.Y.shape[1] > 1, "SGD only works with D > 1" @@ -292,18 +292,18 @@ class model(parameterised): If the overall gradient fails, invividual components are tested. """ - x = self.extract_param().copy() + x = self._get_params_transformed().copy() #choose a random direction to step in: dx = step*np.sign(np.random.uniform(-1,1,x.size)) #evaulate around the point x - self.expand_param(x+dx) - f1,g1 = self.log_likelihood() + self.log_prior(), self.extract_gradients() - self.expand_param(x-dx) - f2,g2 = self.log_likelihood() + self.log_prior(), self.extract_gradients() - self.expand_param(x) - gradient = self.extract_gradients() + self._set_params_transformed(x+dx) + f1,g1 = self.log_likelihood() + self.log_prior(), self._log_likelihood_gradients_transformed() + self._set_params_transformed(x-dx) + f2,g2 = self.log_likelihood() + self.log_prior(), self._log_likelihood_gradients_transformed() + self._set_params_transformed(x) + gradient = self._log_likelihood_gradients_transformed() numerical_gradient = (f1-f2)/(2*dx) ratio = (f1-f2)/(2*np.dot(dx,gradient)) @@ -319,7 +319,7 @@ class model(parameterised): print "Global check failed. Testing individual gradients\n" try: - names = self.extract_param_names() + names = self._get_param_names_transformed() except NotImplementedError: names = ['Variable %i'%i for i in range(len(x))] @@ -338,13 +338,13 @@ class model(parameterised): for i in range(len(x)): xx = x.copy() xx[i] += step - self.expand_param(xx) - f1,g1 = self.log_likelihood() + self.log_prior(), self.extract_gradients()[i] + self._set_params_transformed(xx) + f1,g1 = self.log_likelihood() + self.log_prior(), self._log_likelihood_gradients_transformed()[i] xx[i] -= 2.*step - self.expand_param(xx) - f2,g2 = self.log_likelihood() + self.log_prior(), self.extract_gradients()[i] - self.expand_param(x) - gradient = self.extract_gradients()[i] + self._set_params_transformed(xx) + f2,g2 = self.log_likelihood() + self.log_prior(), self._log_likelihood_gradients_transformed()[i] + self._set_params_transformed(x) + gradient = self._log_likelihood_gradients_transformed()[i] numerical_gradient = (f1-f2)/(2*step) diff --git a/GPy/core/parameterised.py b/GPy/core/parameterised.py index 738bde5b..81b6aa8a 100644 --- a/GPy/core/parameterised.py +++ b/GPy/core/parameterised.py @@ -66,7 +66,7 @@ class parameterised(object): if hasattr(self,'prior'): pass - self.expand_param(self.extract_param())# sets tied parameters to single value + self._set_params_transformed(self._get_params_transformed())# sets tied parameters to single value def untie_everything(self): """Unties all parameters by setting tied_indices to an empty list.""" @@ -216,9 +216,9 @@ class parameterised(object): self.constrained_fixed_values.append(self._get_params()[self.constrained_fixed_indices[-1]]) #self.constrained_fixed_values.append(value) - self.expand_param(self.extract_param()) + self._set_params_transformed(self._get_params_transformed()) - def extract_param(self): + def _get_params_transformed(self): """use self._get_params to get the 'true' parameters of the model, which are then tied, constrained and fixed""" x = self._get_params() x[self.constrained_positive_indices] = np.log(x[self.constrained_positive_indices]) @@ -232,7 +232,7 @@ class parameterised(object): return x - def expand_param(self,x): + def _set_params_transformed(self,x): """ takes the vector x, which is then modified (by untying, reparameterising or inserting fixed values), and then call self._set_params""" #work out how many places are fixed, and where they are. tricky logic! @@ -259,10 +259,10 @@ class parameterised(object): [np.put(xx,i,low+sigmoid(xx[i])*(high-low)) for i,low,high in zip(self.constrained_bounded_indices, self.constrained_bounded_lowers, self.constrained_bounded_uppers)] self._set_params(xx) - def extract_param_names(self): + def _get_param_names_transformed(self): """ Returns the parameter names as propagated after constraining, - tying or fixing, i.e. a list of the same length as extract_param() + tying or fixing, i.e. a list of the same length as _get_params_transformed() """ n = self._get_param_names() diff --git a/GPy/inference/samplers.py b/GPy/inference/samplers.py index da159c23..c2b47bce 100644 --- a/GPy/inference/samplers.py +++ b/GPy/inference/samplers.py @@ -17,7 +17,7 @@ class Metropolis_Hastings: def __init__(self,model,cov=None): """Metropolis Hastings, with tunings according to Gelman et al. """ self.model = model - current = self.model.extract_param() + current = self.model._get_params_transformed() self.D = current.size self.chains = [] if cov is None: @@ -32,19 +32,19 @@ class Metropolis_Hastings: if start is None: self.model.randomize() else: - self.model.expand_param(start) + self.model._set_params_transformed(start) def sample(self, Ntotal, Nburn, Nthin, tune=True, tune_throughout=False, tune_interval=400): - current = self.model.extract_param() + current = self.model._get_params_transformed() fcurrent = self.model.log_likelihood() + self.model.log_prior() accepted = np.zeros(Ntotal,dtype=np.bool) for it in range(Ntotal): print "sample %d of %d\r"%(it,Ntotal), sys.stdout.flush() prop = np.random.multivariate_normal(current, self.cov*self.scale*self.scale) - self.model.expand_param(prop) + self.model._set_params_transformed(prop) fprop = self.model.log_likelihood() + self.model.log_prior() if fprop>fcurrent:#sample accepted, going 'uphill' diff --git a/GPy/models/GPLVM.py b/GPy/models/GPLVM.py index 0a168f25..f4723396 100644 --- a/GPy/models/GPLVM.py +++ b/GPy/models/GPLVM.py @@ -35,10 +35,10 @@ class GPLVM(GP_regression): def _get_param_names(self): return (sum([['X_%i_%i'%(n,q) for n in range(self.N)] for q in range(self.Q)],[]) - + self.kern.extract_param_names()) + + self.kern._get_param_names_transformed()) def _get_params(self): - return np.hstack((self.X.flatten(), self.kern.extract_param())) + return np.hstack((self.X.flatten(), self.kern._get_params_transformed())) def _set_params(self,x): self.X = x[:self.X.size].reshape(self.N,self.Q).copy() diff --git a/GPy/models/GP_EP.py b/GPy/models/GP_EP.py index 5cb2ba45..6f5aad6e 100644 --- a/GPy/models/GP_EP.py +++ b/GPy/models/GP_EP.py @@ -42,13 +42,13 @@ class GP_EP(model): model.__init__(self) def _set_params(self,p): - self.kernel.expand_param(p) + self.kernel._set_params_transformed(p) def _get_params(self): - return self.kernel.extract_param() + return self.kernel._get_params_transformed() def _get_param_names(self): - return self.kernel.extract_param_names() + return self.kernel._get_param_names_transformed() def approximate_likelihood(self): self.ep_approx = Full(self.K,self.likelihood,epsilon=self.epsilon_ep,powerep=[self.eta,self.delta]) @@ -150,8 +150,8 @@ class GP_EP(model): log_likelihood_change = log_likelihood_new - self.log_likelihood_path[-1] if log_likelihood_change < 0: print 'log_likelihood decrement' - self.kernel.expand_param(self.parameters_path[-1]) - self.kernM.expand_param(self.parameters_path[-1]) + self.kernel._set_params_transformed(self.parameters_path[-1]) + self.kernM._set_params_transformed(self.parameters_path[-1]) else: self.approximate_likelihood() self.log_likelihood_path.append(self.log_likelihood()) diff --git a/GPy/models/GP_regression.py b/GPy/models/GP_regression.py index da657950..f34802bf 100644 --- a/GPy/models/GP_regression.py +++ b/GPy/models/GP_regression.py @@ -71,15 +71,15 @@ class GP_regression(model): model.__init__(self) def _set_params(self,p): - self.kern.expand_param(p) + self.kern._set_params_transformed(p) self.K = self.kern.K(self.X,slices1=self.Xslices) self.Ki, self.L, self.Li, self.K_logdet = pdinv(self.K) def _get_params(self): - return self.kern.extract_param() + return self.kern._get_params_transformed() def _get_param_names(self): - return self.kern.extract_param_names() + return self.kern._get_params_names_transformed() def _model_fit_term(self): """ diff --git a/GPy/models/generalized_FITC.py b/GPy/models/generalized_FITC.py index 40f18ff9..a20666d0 100644 --- a/GPy/models/generalized_FITC.py +++ b/GPy/models/generalized_FITC.py @@ -43,14 +43,14 @@ class generalized_FITC(model): model.__init__(self) def _set_params(self,p): - self.kernel.expand_param(p[0:-self.Z.size]) + self.kernel._set_params_transformed(p[0:-self.Z.size]) self.Z = p[-self.Z.size:].reshape(self.M,self.D) def _get_params(self): - return np.hstack([self.kernel.extract_param(),self.Z.flatten()]) + return np.hstack([self.kernel._get_params_transformed(),self.Z.flatten()]) def _get_param_names(self): - return self.kernel.extract_param_names()+['iip_%i'%i for i in range(self.Z.size)] + return self.kernel._get_param_names_transformed()+['iip_%i'%i for i in range(self.Z.size)] def approximate_likelihood(self): self.Kmm = self.kernel.K(self.Z) @@ -227,7 +227,7 @@ class generalized_FITC(model): log_likelihood_change = log_likelihood_new - self.log_likelihood_path[-1] if log_likelihood_change < 0: print 'log_likelihood decrement' - self.kernel.expand_param(self.parameters_path[-1]) + self.kernel._set_params_transformed(self.parameters_path[-1]) self.kernM = self.kernel.copy() slef.kernM.expand_X(self.iducing_inputs_path[-1]) self.__init__(self.kernel,self.likelihood,kernM=self.kernM,powerep=[self.eta,self.delta],epsilon_ep = self.epsilon_ep, epsilon_em = self.epsilon_em) diff --git a/GPy/models/sparse_GP_regression.py b/GPy/models/sparse_GP_regression.py index fc04d5ed..59e6ebbb 100644 --- a/GPy/models/sparse_GP_regression.py +++ b/GPy/models/sparse_GP_regression.py @@ -107,10 +107,10 @@ class sparse_GP_regression(GP_regression): self.dL_dKmm += np.dot(np.dot(self.G,self.beta*self.psi2) - np.dot(self.LBL_inv, self.psi1VVpsi1), self.Kmmi) + 0.5*self.G # dE def _get_params(self): - return np.hstack([self.Z.flatten(),self.beta,self.kern.extract_param()]) + return np.hstack([self.Z.flatten(),self.beta,self.kern._get_params_transformed()]) def _get_param_names(self): - return sum([['iip_%i_%i'%(i,j) for i in range(self.Z.shape[0])] for j in range(self.Z.shape[1])],[]) + ['noise_precision']+self.kern.extract_param_names() + return sum([['iip_%i_%i'%(i,j) for i in range(self.Z.shape[0])] for j in range(self.Z.shape[1])],[]) + ['noise_precision']+self.kern._get_param_names_transformed() def log_likelihood(self): """ From 333d8d3e5ffd05aeb4e99602c5512b5e17cd627e Mon Sep 17 00:00:00 2001 From: Neil Lawrence Date: Fri, 18 Jan 2013 13:39:51 +0000 Subject: [PATCH 10/36] Fix error introduced into GP_regression when doing name changes. --- GPy/models/GP_regression.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/GPy/models/GP_regression.py b/GPy/models/GP_regression.py index f34802bf..91074269 100644 --- a/GPy/models/GP_regression.py +++ b/GPy/models/GP_regression.py @@ -79,7 +79,7 @@ class GP_regression(model): return self.kern._get_params_transformed() def _get_param_names(self): - return self.kern._get_params_names_transformed() + return self.kern._get_param_names_transformed() def _model_fit_term(self): """ From b43db3f8da8add0ef7fcd2d1a49b96412951b2e7 Mon Sep 17 00:00:00 2001 From: Neil Lawrence Date: Fri, 18 Jan 2013 13:47:37 +0000 Subject: [PATCH 11/36] Modified log_like_gradients to make it _log_like_gradients and moved extract_gradient to _log_like_gradients_transformed. --- GPy/core/model.py | 6 +++--- GPy/inference/likelihoods.py | 2 +- GPy/models/GPLVM.py | 2 +- GPy/models/GP_EP.py | 2 +- GPy/models/GP_regression.py | 2 +- GPy/models/generalized_FITC.py | 2 +- GPy/models/sparse_GPLVM.py | 4 ++-- GPy/models/sparse_GP_regression.py | 2 +- GPy/models/warped_GP.py | 4 ++-- 9 files changed, 13 insertions(+), 13 deletions(-) diff --git a/GPy/core/model.py b/GPy/core/model.py index 2bfb02a5..063eaf7d 100644 --- a/GPy/core/model.py +++ b/GPy/core/model.py @@ -25,7 +25,7 @@ class model(parameterised): raise NotImplementedError, "this needs to be implemented to utilise the model class" def log_likelihood(self): raise NotImplementedError, "this needs to be implemented to utilise the model class" - def log_likelihood_gradients(self): + def _log_likelihood_gradients(self): raise NotImplementedError, "this needs to be implemented to utilise the model class" def set_prior(self,which,what): @@ -108,7 +108,7 @@ class model(parameterised): """evaluate the prior""" return np.sum([p.lnpdf(x) for p, x in zip(self.priors,self._get_params()) if p is not None]) - def log_prior_gradients(self): + def _log_prior_gradients(self): """evaluate the gradients of the priors""" x = self._get_params() ret = np.zeros(x.size) @@ -120,7 +120,7 @@ class model(parameterised): Use self.log_likelihood_gradients and self.prior_gradients to get the gradients of the model. Adjust the gradient for constraints and ties, return. """ - g = self.log_likelihood_gradients() + self.log_prior_gradients() + g = self._log_likelihood_gradients() + self._log_prior_gradients() x = self._get_params() g[self.constrained_positive_indices] = g[self.constrained_positive_indices]*x[self.constrained_positive_indices] g[self.constrained_negative_indices] = g[self.constrained_negative_indices]*x[self.constrained_negative_indices] diff --git a/GPy/inference/likelihoods.py b/GPy/inference/likelihoods.py index 6f5e81fa..53dd4248 100644 --- a/GPy/inference/likelihoods.py +++ b/GPy/inference/likelihoods.py @@ -99,6 +99,6 @@ class probit(likelihood): def predictive_mean(self,mu,variance): return stats.norm.cdf(mu/np.sqrt(1+variance)) - def log_likelihood_gradients(): + def _log_likelihood_gradients(): raise NotImplementedError diff --git a/GPy/models/GPLVM.py b/GPy/models/GPLVM.py index f4723396..3749ee32 100644 --- a/GPy/models/GPLVM.py +++ b/GPy/models/GPLVM.py @@ -44,7 +44,7 @@ class GPLVM(GP_regression): self.X = x[:self.X.size].reshape(self.N,self.Q).copy() GP_regression._set_params(self, x[self.X.size:]) - def log_likelihood_gradients(self): + def _log_likelihood_gradients(self): dL_dK = self.dL_dK() dL_dtheta = self.kern.dK_dtheta(dL_dK,self.X) diff --git a/GPy/models/GP_EP.py b/GPy/models/GP_EP.py index 6f5aad6e..51d69d0a 100644 --- a/GPy/models/GP_EP.py +++ b/GPy/models/GP_EP.py @@ -78,7 +78,7 @@ class GP_EP(model): L3 = sum(np.log(self.ep_approx.Z_hat)) return L1 + L2A + L2B + L3 - def log_likelihood_gradients(self): + def _log_likelihood_gradients(self): dK_dp = self.kernel.dK_dtheta(self.X) self.dK_dp = dK_dp aux1,info_1 = linalg.flapack.dtrtrs(self.L,np.dot(self.Sroot_tilde_K,self.ep_approx.v_tilde),lower=1) diff --git a/GPy/models/GP_regression.py b/GPy/models/GP_regression.py index 91074269..eee0fb58 100644 --- a/GPy/models/GP_regression.py +++ b/GPy/models/GP_regression.py @@ -103,7 +103,7 @@ class GP_regression(model): return dL_dK - def log_likelihood_gradients(self): + def _log_likelihood_gradients(self): return self.kern.dK_dtheta(partial=self.dL_dK(),X=self.X) def predict(self,Xnew, slices=None, full_cov=False): diff --git a/GPy/models/generalized_FITC.py b/GPy/models/generalized_FITC.py index a20666d0..d8e9c23d 100644 --- a/GPy/models/generalized_FITC.py +++ b/GPy/models/generalized_FITC.py @@ -99,7 +99,7 @@ class generalized_FITC(model): E = .5*np.sum((self.ep_approx.v_/self.ep_approx.tau_ - self.mu_tilde.flatten())**2/(1./self.ep_approx.tau_ + 1./self.ep_approx.tau_tilde)) return A + B + C + D + E - def log_likelihood_gradients(self): + def _log_likelihood_gradients(self): dKmm_dtheta = self.kernel.dK_dtheta(self.Z) dKnn_dtheta = self.kernel.dK_dtheta(self.X) dKmn_dtheta = self.kernel.dK_dtheta(self.Z,self.X) diff --git a/GPy/models/sparse_GPLVM.py b/GPy/models/sparse_GPLVM.py index 1c6a03cf..84a04bfd 100644 --- a/GPy/models/sparse_GPLVM.py +++ b/GPy/models/sparse_GPLVM.py @@ -49,8 +49,8 @@ class sparse_GPLVM(sparse_GP_regression, GPLVM): return dL_dX - def log_likelihood_gradients(self): - return np.hstack((self.dL_dX().flatten(), sparse_GP_regression.log_likelihood_gradients(self))) + def _log_likelihood_gradients(self): + return np.hstack((self.dL_dX().flatten(), sparse_GP_regression._log_likelihood_gradients(self))) def plot(self): GPLVM.plot(self) diff --git a/GPy/models/sparse_GP_regression.py b/GPy/models/sparse_GP_regression.py index 59e6ebbb..870828ca 100644 --- a/GPy/models/sparse_GP_regression.py +++ b/GPy/models/sparse_GP_regression.py @@ -168,7 +168,7 @@ class sparse_GP_regression(GP_regression): dL_dZ += self.kern.dK_dX(dL_dpsi1,self.Z,self.X) return dL_dZ - def log_likelihood_gradients(self): + def _log_likelihood_gradients(self): return np.hstack([self.dL_dZ().flatten(), self.dL_dbeta(), self.dL_dtheta()]) def _raw_predict(self, Xnew, slices, full_cov=False): diff --git a/GPy/models/warped_GP.py b/GPy/models/warped_GP.py index ea3c44cc..ff6267d2 100644 --- a/GPy/models/warped_GP.py +++ b/GPy/models/warped_GP.py @@ -59,8 +59,8 @@ class warpedGP(GP_regression): jacobian = self.warping_function.fgrad_y(self.Z, self.warping_params) return ll + np.log(jacobian).sum() - def log_likelihood_gradients(self): - ll_grads = GP_regression.log_likelihood_gradients(self) + def _log_likelihood_gradients(self): + ll_grads = GP_regression._log_likelihood_gradients(self) alpha = np.dot(self.Ki, self.Y.flatten()) warping_grads = self.warping_function_gradients(alpha) return np.hstack((warping_grads.flatten(), ll_grads.flatten())) From 68d7e2364862b83213a5824ae51b49bd83780fa7 Mon Sep 17 00:00:00 2001 From: Nicolas Date: Fri, 18 Jan 2013 13:58:41 +0000 Subject: [PATCH 12/36] rbf kernel now has an ARD flag --- GPy/examples/GP_regression_demo.py | 12 ++++-- GPy/examples/sparse_GPLVM_demo.py | 4 +- GPy/kern/constructors.py | 4 +- GPy/kern/rbf.py | 65 ++++++++++++++++++++---------- 4 files changed, 57 insertions(+), 28 deletions(-) diff --git a/GPy/examples/GP_regression_demo.py b/GPy/examples/GP_regression_demo.py index 0fe2fabd..96bd6dc3 100644 --- a/GPy/examples/GP_regression_demo.py +++ b/GPy/examples/GP_regression_demo.py @@ -1,7 +1,6 @@ # Copyright (c) 2012, GPy authors (see AUTHORS.txt). # Licensed under the BSD 3-clause license (see LICENSE.txt) - """ Simple Gaussian Processes regression with an RBF kernel """ @@ -19,8 +18,11 @@ pb.close('all') X = np.random.uniform(-3.,3.,(20,1)) Y = np.sin(X)+np.random.randn(20,1)*0.05 +# define kernel +ker = GPy.kern.rbf(1,ARD=False) + GPy.kern.white(1) + # create simple GP model -m = GPy.models.GP_regression(X,Y) +m = GPy.models.GP_regression(X,Y,ker) # contrain all parameters to be positive m.constrain_positive('') @@ -30,6 +32,7 @@ m.optimize('tnc', max_f_eval = 1000) m.plot() print(m) + ###################################### ## 2 dimensional example @@ -37,8 +40,11 @@ print(m) X = np.random.uniform(-3.,3.,(40,2)) Y = np.sin(X[:,0:1]) * np.sin(X[:,1:2])+np.random.randn(40,1)*0.05 +# define kernel +ker = GPy.kern.rbf(2,ARD=True) + GPy.kern.white(2) + # create simple GP model -m = GPy.models.GP_regression(X,Y) +m = GPy.models.GP_regression(X,Y,ker) # contrain all parameters to be positive m.constrain_positive('') diff --git a/GPy/examples/sparse_GPLVM_demo.py b/GPy/examples/sparse_GPLVM_demo.py index 6ca6c941..3f1969fe 100644 --- a/GPy/examples/sparse_GPLVM_demo.py +++ b/GPy/examples/sparse_GPLVM_demo.py @@ -10,11 +10,11 @@ print "sparse GPLVM with RBF kernel" N = 100 M = 4 -Q = 1 +Q = 2 D = 2 #generate GPLVM-like data X = np.random.rand(N, Q) -k = GPy.kern.rbf(Q, 1.0, 2.0) + GPy.kern.white(Q, 0.00001) +k = GPy.kern.rbf(Q,1.,2*np.ones((1,))) + GPy.kern.white(Q, 0.00001) K = k.K(X) Y = np.random.multivariate_normal(np.zeros(N),K,D).T diff --git a/GPy/kern/constructors.py b/GPy/kern/constructors.py index 0ddc09e3..e1304f11 100644 --- a/GPy/kern/constructors.py +++ b/GPy/kern/constructors.py @@ -22,7 +22,7 @@ from Brownian import Brownian as Brownianpart #using meta-classes to make the objects construct properly wthout them. -def rbf(D,variance=1., lengthscale=1.): +def rbf(D,variance=1., lengthscale=None,ARD=False): """ Construct an RBF kernel @@ -33,7 +33,7 @@ def rbf(D,variance=1., lengthscale=1.): :param lengthscale: the lengthscale of the kernel :type lengthscale: float """ - part = rbfpart(D,variance,lengthscale) + part = rbfpart(D,variance,lengthscale,ARD) return kern(D, [part]) def rbf_ARD(D,variance=1., lengthscales=None): diff --git a/GPy/kern/rbf.py b/GPy/kern/rbf.py index 9e2bb509..775bb797 100644 --- a/GPy/kern/rbf.py +++ b/GPy/kern/rbf.py @@ -20,17 +20,33 @@ class rbf(kernpart): :type D: int :param variance: the variance of the kernel :type variance: float - :param lengthscale: the lengthscale of the kernel - :type lengthscale: float + :param lengthscale: the vector of lengthscale of the kernel + :type lengthscale: np.ndarray + :param ARD: Auto Relevance Determination. If equal to "False", the kernel is isotropic (ie. one single lengthscale parameter \ell), otherwise there is one lengthscale parameter per dimension. + :type ARD: Boolean - .. Note: for rbf with different lengthscale on each dimension, see rbf_ARD """ - def __init__(self,D,variance=1.,lengthscale=1.): + def __init__(self,D,variance=1.,lengthscale=None,ARD=False): self.D = D - self.Nparam = 2 - self.name = 'rbf' - self.set_param(np.hstack((variance,lengthscale))) + self.ARD = ARD + if ARD == False: + self.Nparam = 2 + self.name = 'rbf' + if lengthscale is not None: + assert lengthscale.shape == (1,) + else: + lengthscale = np.ones(1) + + else: + self.Nparam = self.D + 1 + self.name = 'rbf_ARD' + if lengthscale is not None: + assert lengthscale.shape == (self.D,) + else: + lengthscale = np.ones(self.D) + + self.set_param(np.hstack((variance,lengthscale))) #initialize cache self._Z, self._mu, self._S = np.empty(shape=(3,1)) @@ -40,14 +56,19 @@ class rbf(kernpart): return np.hstack((self.variance,self.lengthscale)) def set_param(self,x): - self.variance, self.lengthscale = x + assert x.size==(self.Nparam) + self.variance = x[0] + self.lengthscale = x[1:] self.lengthscale2 = np.square(self.lengthscale) #reset cached results self._X, self._X2, self._params = np.empty(shape=(3,1)) self._Z, self._mu, self._S = np.empty(shape=(3,1)) # cached versions of Z,mu,S def get_param_names(self): - return ['variance','lengthscale'] + if self.Nparam == 2: + return ['variance','lengthscale'] + else: + return ['variance']+['lengthscale_%i'%i for i in range(self.lengthscale.size)] def K(self,X,X2,target): if X2 is None: @@ -61,7 +82,12 @@ class rbf(kernpart): def dK_dtheta(self,partial,X,X2,target): self._K_computations(X,X2) target[0] += np.sum(self._K_dvar*partial) - target[1] += np.sum(self._K_dvar*self.variance*self._K_dist2/self.lengthscale*partial) + if self.ARD == True: + dl = self._K_dvar[:,:,None]*self.variance*self._K_dist2/self.lengthscale + target[1:] += (dl*partial[:,:,None]).sum(0).sum(0) + else: + target[1] += np.sum(self._K_dvar*self.variance*(self._K_dist2.sum(-1))/self.lengthscale*partial) + #np.sum(self._K_dvar*self.variance*self._K_dist2/self.lengthscale*partial) def dKdiag_dtheta(self,partial,X,target): #NB: derivative of diagonal elements wrt lengthscale is 0 @@ -81,15 +107,13 @@ class rbf(kernpart): self._X = X self._X2 = X2 if X2 is None: X2 = X - XXT = np.dot(X,X2.T) - if X is X2: - self._K_dist2 = (-2.*XXT + np.diag(XXT)[:,np.newaxis] + np.diag(XXT)[np.newaxis,:])/self.lengthscale2 - else: - self._K_dist2 = (-2.*XXT + np.sum(np.square(X),1)[:,None] + np.sum(np.square(X2),1)[None,:])/self.lengthscale2 - # TODO Remove comments if this is fine. - # Commented out by Neil as doesn't seem to be used elsewhere. - #self._K_exponent = -0.5*self._K_dist2 - self._K_dvar = np.exp(-0.5*self._K_dist2) + self._K_dist = X[:,None,:]-X2[None,:,:] # this can be computationally heavy + self._params = np.empty(shape=(1,0))#ensure the next section gets called + if not np.all(self._params == self.get_param()): + self._params == self.get_param() + self._K_dist2 = np.square(self._K_dist/self.lengthscale) + #self._K_exponent = -0.5*self._K_dist2.sum(-1) #ND: commented out because seems not to be used + self._K_dvar = np.exp(-0.5*self._K_dist2.sum(-1)) def psi0(self,Z,mu,S,target): target += self.variance @@ -132,7 +156,7 @@ class rbf(kernpart): d_length = self._psi2[:,:,:,None]*(0.5*self._psi2_Zdist_sq*self._psi2_denom + 2.*self._psi2_mudist_sq + 2.*S[:,None,None,:]/self.lengthscale2)/(self.lengthscale*self._psi2_denom) d_length = d_length.sum(0) target[0] += np.sum(partial*d_var) - target[1] += np.sum(d_length*partial) + target[1:] += (d_length*partial[:,:,None]).sum(0).sum(0) def dpsi2_dZ(self,partial,Z,mu,S,target): """Returns shape N,M,M,Q""" @@ -175,4 +199,3 @@ class rbf(kernpart): self._psi2 = np.square(self.variance)*np.exp(self._psi2_exponent) # N,M,M self._Z, self._mu, self._S = Z, mu,S - From 17ec4d6275c6887b0a45ee7c7b26f1fefa2d6b35 Mon Sep 17 00:00:00 2001 From: Nicolas Date: Fri, 18 Jan 2013 14:33:52 +0000 Subject: [PATCH 13/36] change in unit_test to take into account the ARD changes in rbf --- GPy/testing/unit_tests.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/GPy/testing/unit_tests.py b/GPy/testing/unit_tests.py index ff9aba0e..d2ef87f7 100644 --- a/GPy/testing/unit_tests.py +++ b/GPy/testing/unit_tests.py @@ -121,7 +121,7 @@ class GradientTests(unittest.TestCase): """ Testing GPLVM with rbf + bias and white kernel """ N, Q, D = 50, 1, 2 X = np.random.rand(N, Q) - k = GPy.kern.rbf(Q, 0.5, 0.9) + GPy.kern.bias(Q, 0.1) + GPy.kern.white(Q, 0.05) + k = GPy.kern.rbf(Q, 0.5, 0.9*np.ones((1,))) + GPy.kern.bias(Q, 0.1) + GPy.kern.white(Q, 0.05) K = k.K(X) Y = np.random.multivariate_normal(np.zeros(N),K,D).T m = GPy.models.GPLVM(Y, Q, kernel = k) From 11d088cf903191254b53a47a6647e5512a8511ec Mon Sep 17 00:00:00 2001 From: Nicolas Date: Fri, 18 Jan 2013 15:14:23 +0000 Subject: [PATCH 14/36] added ARD flag to exponential --- GPy/kern/__init__.py | 2 +- GPy/kern/constructors.py | 22 +++------------- GPy/kern/exponential.py | 54 +++++++++++++++++++++++++--------------- 3 files changed, 39 insertions(+), 39 deletions(-) diff --git a/GPy/kern/__init__.py b/GPy/kern/__init__.py index cd893bac..4a36d6d0 100644 --- a/GPy/kern/__init__.py +++ b/GPy/kern/__init__.py @@ -2,5 +2,5 @@ # Licensed under the BSD 3-clause license (see LICENSE.txt) -from constructors import rbf, Matern32, Matern52, exponential, linear, white, bias, finite_dimensional, rbf_ARD, spline, Brownian, linear_ARD, rbf_sympy, sympykern +from constructors import rbf, Matern32, Matern52, exponential, linear, white, bias, finite_dimensional, spline, Brownian, linear_ARD, rbf_sympy, sympykern from kern import kern diff --git a/GPy/kern/constructors.py b/GPy/kern/constructors.py index e1304f11..5f676d9b 100644 --- a/GPy/kern/constructors.py +++ b/GPy/kern/constructors.py @@ -36,20 +36,6 @@ def rbf(D,variance=1., lengthscale=None,ARD=False): part = rbfpart(D,variance,lengthscale,ARD) return kern(D, [part]) -def rbf_ARD(D,variance=1., lengthscales=None): - """ - Construct an RBF kernel with Automatic Relevance Determination (ARD) - - :param D: dimensionality of the kernel, obligatory - :type D: int - :param variance: the variance of the kernel - :type variance: float - :param lengthscales: the lengthscales of the kernel - :type lengthscales: None|np.ndarray - """ - part = rbf_ARD_part(D,variance,lengthscales) - return kern(D, [part]) - def linear(D,lengthscales=None): """ Construct a linear kernel. @@ -86,7 +72,7 @@ def white(D,variance=1.): part = whitepart(D,variance) return kern(D, [part]) -def exponential(D,variance=1., lengthscales=None): +def exponential(D,variance=1., lengthscale=None, ARD=False): """ Construct a exponential kernel. @@ -96,10 +82,10 @@ def exponential(D,variance=1., lengthscales=None): variance (float) lengthscales (np.ndarray) """ - part = exponentialpart(D,variance, lengthscales) + part = exponentialpart(D,variance, lengthscale, ARD) return kern(D, [part]) -def Matern32(D,variance=1., lengthscales=None): +def Matern32(D,variance=1., lengthscale=None, ARD=False): """ Construct a Matern 3/2 kernel. @@ -109,7 +95,7 @@ def Matern32(D,variance=1., lengthscales=None): variance (float) lengthscales (np.ndarray) """ - part = Matern32part(D,variance, lengthscales) + part = Matern32part(D,variance, lengthscale, ARD) return kern(D, [part]) def Matern52(D,variance=1., lengthscales=None): diff --git a/GPy/kern/exponential.py b/GPy/kern/exponential.py index 2df6a958..21bb8398 100644 --- a/GPy/kern/exponential.py +++ b/GPy/kern/exponential.py @@ -24,37 +24,46 @@ class exponential(kernpart): :rtype: kernel object """ - def __init__(self,D,variance=1.,lengthscales=None): + def __init__(self,D,variance=1.,lengthscale=None,ARD=False): self.D = D - if lengthscales is not None: - assert lengthscales.shape==(self.D,) + self.ARD = ARD + if ARD == False: + self.Nparam = 2 + self.name = 'exp' + if lengthscale is not None: + assert lengthscale.shape == (1,) + else: + lengthscale = np.ones(1) else: - lengthscales = np.ones(self.D) - self.Nparam = self.D + 1 - self.name = 'exp' - self._set_params(np.hstack((variance,lengthscales))) + self.Nparam = self.D + 1 + self.name = 'exp_ARD' + if lengthscale is not None: + assert lengthscale.shape == (self.D,) + else: + lengthscale = np.ones(self.D) + self._set_params(np.hstack((variance,lengthscale))) def _get_params(self): """return the value of the parameters.""" - return np.hstack((self.variance,self.lengthscales)) + return np.hstack((self.variance,self.lengthscale)) def _set_params(self,x): """set the value of the parameters.""" assert x.size==(self.D+1) self.variance = x[0] - self.lengthscales = x[1:] + self.lengthscale = x[1:] def _get_param_names(self): """return parameter names.""" - if self.D==1: + if self.Nparam==2: return ['variance','lengthscale'] else: - return ['variance']+['lengthscale_%i'%i for i in range(self.lengthscales.size)] + return ['variance']+['lengthscale_%i'%i for i in range(self.lengthscale.size)] def K(self,X,X2,target): """Compute the covariance matrix between X and X2.""" if X2 is None: X2 = X - dist = np.sqrt(np.sum(np.square((X[:,None,:]-X2[None,:,:])/self.lengthscales),-1)) + dist = np.sqrt(np.sum(np.square((X[:,None,:]-X2[None,:,:])/self.lengthscale),-1)) np.add(self.variance*np.exp(-dist), target,target) def Kdiag(self,X,target): @@ -64,13 +73,18 @@ class exponential(kernpart): def dK_dtheta(self,partial,X,X2,target): """derivative of the covariance matrix with respect to the parameters.""" if X2 is None: X2 = X - dist = np.sqrt(np.sum(np.square((X[:,None,:]-X2[None,:,:])/self.lengthscales),-1)) + dist = np.sqrt(np.sum(np.square((X[:,None,:]-X2[None,:,:])/self.lengthscale),-1)) invdist = 1./np.where(dist!=0.,dist,np.inf) - dist2M = np.square(X[:,None,:]-X2[None,:,:])/self.lengthscales**3 + dist2M = np.square(X[:,None,:]-X2[None,:,:])/self.lengthscale**3 dvar = np.exp(-dist) - dl = self.variance*dvar[:,:,None]*dist2M*invdist[:,:,None] target[0] += np.sum(dvar*partial) - target[1:] += (dl*partial[:,:,None]).sum(0).sum(0) + if self.ARD == True: + dl = self.variance*dvar[:,:,None]*dist2M*invdist[:,:,None] + target[1:] += (dl*partial[:,:,None]).sum(0).sum(0) + else: + dl = self.variance*dvar*dist2M.sum(-1)*invdist + target[1] += np.sum(dl*partial) + #foo def dKdiag_dtheta(self,partial,X,target): """derivative of the diagonal of the covariance matrix with respect to the parameters.""" @@ -80,8 +94,8 @@ class exponential(kernpart): def dK_dX(self,partial,X,X2,target): """derivative of the covariance matrix with respect to X.""" if X2 is None: X2 = X - dist = np.sqrt(np.sum(np.square((X[:,None,:]-X2[None,:,:])/self.lengthscales),-1))[:,:,None] - ddist_dX = (X[:,None,:]-X2[None,:,:])/self.lengthscales**2/np.where(dist!=0.,dist,np.inf) + dist = np.sqrt(np.sum(np.square((X[:,None,:]-X2[None,:,:])/self.lengthscale),-1))[:,:,None] + ddist_dX = (X[:,None,:]-X2[None,:,:])/self.lengthscale**2/np.where(dist!=0.,dist,np.inf) dK_dX = - np.transpose(self.variance*np.exp(-dist)*ddist_dX,(1,0,2)) target += np.sum(dK_dX*partial.T[:,:,None],0) @@ -101,14 +115,14 @@ class exponential(kernpart): """ assert self.D == 1 def L(x,i): - return(1./self.lengthscales*F[i](x) + F1[i](x)) + return(1./self.lengthscale*F[i](x) + F1[i](x)) n = F.shape[0] G = np.zeros((n,n)) for i in range(n): for j in range(i,n): G[i,j] = G[j,i] = integrate.quad(lambda x : L(x,i)*L(x,j),lower,upper)[0] Flower = np.array([f(lower) for f in F])[:,None] - return(self.lengthscales/2./self.variance * G + 1./self.variance * np.dot(Flower,Flower.T)) + return(self.lengthscale/2./self.variance * G + 1./self.variance * np.dot(Flower,Flower.T)) From 99034d0fb0a96dc522d6c82ee0abba12e0cf5718 Mon Sep 17 00:00:00 2001 From: Neil Lawrence Date: Fri, 18 Jan 2013 15:31:20 +0000 Subject: [PATCH 15/36] Mods to regression.py now that we have get to get parameters. Moved Youter to YYT. --- GPy/__init__.py | 1 + GPy/core/model.py | 25 +++++++++++++++++++++---- GPy/examples/regression.py | 31 +++++++++---------------------- GPy/models/GP_regression.py | 16 ++++++++-------- GPy/models/generalized_FITC.py | 4 ++-- GPy/models/warped_GP.py | 4 ++-- 6 files changed, 43 insertions(+), 38 deletions(-) diff --git a/GPy/__init__.py b/GPy/__init__.py index 6993d5c2..381d6232 100644 --- a/GPy/__init__.py +++ b/GPy/__init__.py @@ -6,5 +6,6 @@ import kern import models import inference import util +import examples #import examples TODO: discuss! from core import priors diff --git a/GPy/core/model.py b/GPy/core/model.py index 063eaf7d..77e66600 100644 --- a/GPy/core/model.py +++ b/GPy/core/model.py @@ -80,19 +80,22 @@ class model(parameterised): for w in which: self.priors[w] = what - def get(self,name): + def get(self,name, return_names=False): """ - get a model parameter by name + Get a model parameter by name. The name is applied as a regular expression and all parameters that match that regular expression are returned. """ matches = self.grep_param_names(name) if len(matches): - return self._get_params()[matches] + if return_names: + return self._get_params()[matches], np.asarray(self._get_param_names())[matches].tolist() + else: + return self._get_params()[matches] else: raise AttributeError, "no parameter matches %s"%name def set(self,name,val): """ - Set a model parameter by name + Set model parameter(s) by name. The name is provided as a regular expression. All parameters matching that regular expression are set to ghe given value. """ matches = self.grep_param_names(name) if len(matches): @@ -102,6 +105,20 @@ class model(parameterised): else: raise AttributeError, "no parameter matches %s"%name + def get_gradient(self,name, return_names=False): + """ + Get model gradient(s) by name. The name is applied as a regular expression and all parameters that match that regular expression are returned. + """ + matches = self.grep_param_names(name) + if len(matches): + if return_names: + return self._log_likelihood_gradients()[matches], np.asarray(self._get_param_names())[matches].tolist() + else: + return self._log_likelihood_gradients()[matches] + else: + raise AttributeError, "no parameter matches %s"%name + + def log_prior(self): diff --git a/GPy/examples/regression.py b/GPy/examples/regression.py index 79763504..9444e899 100644 --- a/GPy/examples/regression.py +++ b/GPy/examples/regression.py @@ -17,10 +17,8 @@ def toy_rbf_1d(): # create simple GP model m = GPy.models.GP_regression(data['X'],data['Y']) - # contrain all parameters to be positive - m.constrain_positive('') - # optimize + m.ensure_default_constraints() m.optimize() # plot @@ -35,10 +33,8 @@ def rogers_girolami_olympics(): # create simple GP model m = GPy.models.GP_regression(data['X'],data['Y']) - # contrain all parameters to be positive - m.constrain_positive('') - # optimize + m.ensure_default_constraints() m.optimize() # plot @@ -57,10 +53,8 @@ def toy_rbf_1d_50(): # create simple GP model m = GPy.models.GP_regression(data['X'],data['Y']) - # contrain all parameters to be positive - m.constrain_positive('') - # optimize + m.ensure_default_constraints() m.optimize() # plot @@ -75,10 +69,8 @@ def silhouette(): # create simple GP model m = GPy.models.GP_regression(data['X'],data['Y']) - # contrain all parameters to be positive - m.constrain_positive('') - # optimize + m.ensure_default_constraints() m.optimize() print(m) @@ -118,20 +110,15 @@ def multiple_optima(gene_number=937,resolution=80, model_restarts=10, seed=10000 kern = GPy.kern.rbf(1, variance=np.random.exponential(1.), lengthscale=np.random.exponential(50.)) + GPy.kern.white(1,variance=np.random.exponential(1.)) m = GPy.models.GP_regression(data['X'],data['Y'], kernel=kern) - params = m._get_params() - optim_point_x[0] = params[1] - optim_point_y[0] = np.log10(params[0]) - np.log10(params[2]); - - # contrain all parameters to be positive - m.constrain_positive('') + optim_point_x[0] = m.get('rbf_lengthscale') + optim_point_y[0] = np.log10(m.get('rbf_variance')) - np.log10(m.get('white_variance')); # optimize + m.ensure_default_constraints() m.optimize(xtol=1e-6,ftol=1e-6) - params = m._get_params() - optim_point_x[1] = params[1] - optim_point_y[1] = np.log10(params[0]) - np.log10(params[2]); - print(m) + optim_point_x[1] = m.get('rbf_lengthscale') + optim_point_y[1] = np.log10(m.get('rbf_variance')) - np.log10(m.get('white_variance')); pb.arrow(optim_point_x[0], optim_point_y[0], optim_point_x[1]-optim_point_x[0], optim_point_y[1]-optim_point_y[0], label=str(i), head_length=1, head_width=0.5, fc='k', ec='k') models.append(m) diff --git a/GPy/models/GP_regression.py b/GPy/models/GP_regression.py index eee0fb58..72a24307 100644 --- a/GPy/models/GP_regression.py +++ b/GPy/models/GP_regression.py @@ -63,10 +63,10 @@ class GP_regression(model): self._Ystd = np.ones((1,self.Y.shape[1])) if self.D > self.N: - # then it's more efficient to store Youter - self.Youter = np.dot(self.Y, self.Y.T) + # then it's more efficient to store YYT + self.YYT = np.dot(self.Y, self.Y.T) else: - self.Youter = None + self.YYT = None model.__init__(self) @@ -83,23 +83,23 @@ class GP_regression(model): def _model_fit_term(self): """ - Computes the model fit using Youter if it's available + Computes the model fit using YYT if it's available """ - if self.Youter is None: + if self.YYT is None: return -0.5*np.sum(np.square(np.dot(self.Li,self.Y))) else: - return -0.5*np.sum(np.multiply(self.Ki, self.Youter)) + return -0.5*np.sum(np.multiply(self.Ki, self.YYT)) def log_likelihood(self): complexity_term = -0.5*self.N*self.D*np.log(2.*np.pi) - 0.5*self.D*self.K_logdet return complexity_term + self._model_fit_term() def dL_dK(self): - if self.Youter is None: + if self.YYT is None: alpha = np.dot(self.Ki,self.Y) dL_dK = 0.5*(np.dot(alpha,alpha.T)-self.D*self.Ki) else: - dL_dK = 0.5*(mdot(self.Ki, self.Youter, self.Ki) - self.D*self.Ki) + dL_dK = 0.5*(mdot(self.Ki, self.YYT, self.Ki) - self.D*self.Ki) return dL_dK diff --git a/GPy/models/generalized_FITC.py b/GPy/models/generalized_FITC.py index d8e9c23d..a5ed8d0a 100644 --- a/GPy/models/generalized_FITC.py +++ b/GPy/models/generalized_FITC.py @@ -91,9 +91,9 @@ class generalized_FITC(model): def log_likelihood(self): self.posterior_param() - self.Youter = np.dot(self.mu_tilde,self.mu_tilde.T) + self.YYT = np.dot(self.mu_tilde,self.mu_tilde.T) A = -self.hld - B = -.5*np.sum(self.Qi*self.Youter) + B = -.5*np.sum(self.Qi*self.YYT) C = sum(np.log(self.ep_approx.Z_hat)) D = .5*np.sum(np.log(1./self.ep_approx.tau_tilde + 1./self.ep_approx.tau_)) E = .5*np.sum((self.ep_approx.v_/self.ep_approx.tau_ - self.mu_tilde.flatten())**2/(1./self.ep_approx.tau_ + 1./self.ep_approx.tau_tilde)) diff --git a/GPy/models/warped_GP.py b/GPy/models/warped_GP.py index ff6267d2..8ce80c76 100644 --- a/GPy/models/warped_GP.py +++ b/GPy/models/warped_GP.py @@ -48,9 +48,9 @@ class warpedGP(GP_regression): # this supports the 'smart' behaviour in GP_regression if self.D > self.N: - self.Youter = np.dot(self.Y, self.Y.T) + self.YYT = np.dot(self.Y, self.Y.T) else: - self.Youter = None + self.YYT = None return self.Y From 0c30048e572568bfe6bace4f73fcd95c8796658d Mon Sep 17 00:00:00 2001 From: Nicolas Date: Fri, 18 Jan 2013 15:34:06 +0000 Subject: [PATCH 16/36] More ARD flags in exponential and Matern32 --- GPy/kern/Matern32.py | 60 ++++++++++++++++++++++++++--------------- GPy/kern/exponential.py | 5 ++-- 2 files changed, 40 insertions(+), 25 deletions(-) diff --git a/GPy/kern/Matern32.py b/GPy/kern/Matern32.py index 6517ac2c..1270e3f9 100644 --- a/GPy/kern/Matern32.py +++ b/GPy/kern/Matern32.py @@ -20,43 +20,52 @@ class Matern32(kernpart): :type D: int :param variance: the variance :math:`\sigma^2` :type variance: float - :param lengthscale: the lengthscales :math:`\ell_i` + :param lengthscale: the lengthscale :math:`\ell_i` :type lengthscale: np.ndarray of size (D,) :rtype: kernel object """ - def __init__(self,D,variance=1.,lengthscales=None): + def __init__(self,D,variance=1.,lengthscale=None,ARD=False): self.D = D - if lengthscales is not None: - assert lengthscales.shape==(self.D,) + self.ARD = ARD + if ARD == False: + self.Nparam = 2 + self.name = 'Mat32' + if lengthscale is not None: + assert lengthscale.shape == (1,) + else: + lengthscale = np.ones(1) else: - lengthscales = np.ones(self.D) - self.Nparam = self.D + 1 - self.name = 'Mat32' - self._set_params(np.hstack((variance,lengthscales))) + self.Nparam = self.D + 1 + self.name = 'Mat32_ARD' + if lengthscale is not None: + assert lengthscale.shape == (self.D,) + else: + lengthscale = np.ones(self.D) + self._set_params(np.hstack((variance,lengthscale))) def _get_params(self): """return the value of the parameters.""" - return np.hstack((self.variance,self.lengthscales)) + return np.hstack((self.variance,self.lengthscale)) def _set_params(self,x): """set the value of the parameters.""" - assert x.size==(self.D+1) + assert x.size == self.Nparam self.variance = x[0] - self.lengthscales = x[1:] + self.lengthscale = x[1:] def _get_param_names(self): """return parameter names.""" - if self.D==1: + if self.Nparam == 2: return ['variance','lengthscale'] else: - return ['variance']+['lengthscale_%i'%i for i in range(self.lengthscales.size)] + return ['variance']+['lengthscale_%i'%i for i in range(self.lengthscale.size)] def K(self,X,X2,target): """Compute the covariance matrix between X and X2.""" if X2 is None: X2 = X - dist = np.sqrt(np.sum(np.square((X[:,None,:]-X2[None,:,:])/self.lengthscales),-1)) + dist = np.sqrt(np.sum(np.square((X[:,None,:]-X2[None,:,:])/self.lengthscale),-1)) np.add(self.variance*(1+np.sqrt(3.)*dist)*np.exp(-np.sqrt(3.)*dist), target,target) def Kdiag(self,X,target): @@ -66,13 +75,20 @@ class Matern32(kernpart): def dK_dtheta(self,partial,X,X2,target): """derivative of the covariance matrix with respect to the parameters.""" if X2 is None: X2 = X - dist = np.sqrt(np.sum(np.square((X[:,None,:]-X2[None,:,:])/self.lengthscales),-1)) + dist = np.sqrt(np.sum(np.square((X[:,None,:]-X2[None,:,:])/self.lengthscale),-1)) dvar = (1+np.sqrt(3.)*dist)*np.exp(-np.sqrt(3.)*dist) invdist = 1./np.where(dist!=0.,dist,np.inf) - dist2M = np.square(X[:,None,:]-X2[None,:,:])/self.lengthscales**3 - dl = (self.variance* 3 * dist * np.exp(-np.sqrt(3.)*dist))[:,:,np.newaxis] * dist2M*invdist[:,:,np.newaxis] + dist2M = np.square(X[:,None,:]-X2[None,:,:])/self.lengthscale**3 + #dl = (self.variance* 3 * dist * np.exp(-np.sqrt(3.)*dist))[:,:,np.newaxis] * dist2M*invdist[:,:,np.newaxis] target[0] += np.sum(dvar*partial) - target[1:] += (dl*partial[:,:,None]).sum(0).sum(0) + if self.ARD == True: + dl = (self.variance* 3 * dist * np.exp(-np.sqrt(3.)*dist))[:,:,np.newaxis] * dist2M*invdist[:,:,np.newaxis] + #dl = self.variance*dvar[:,:,None]*dist2M*invdist[:,:,None] + target[1:] += (dl*partial[:,:,None]).sum(0).sum(0) + else: + dl = (self.variance* 3 * dist * np.exp(-np.sqrt(3.)*dist)) * dist2M.sum(-1)*invdist + #dl = self.variance*dvar*dist2M.sum(-1)*invdist + target[1] += np.sum(dl*partial) def dKdiag_dtheta(self,partial,X,target): """derivative of the diagonal of the covariance matrix with respect to the parameters.""" @@ -81,8 +97,8 @@ class Matern32(kernpart): def dK_dX(self,partial,X,X2,target): """derivative of the covariance matrix with respect to X.""" if X2 is None: X2 = X - dist = np.sqrt(np.sum(np.square((X[:,None,:]-X2[None,:,:])/self.lengthscales),-1))[:,:,None] - ddist_dX = (X[:,None,:]-X2[None,:,:])/self.lengthscales**2/np.where(dist!=0.,dist,np.inf) + dist = np.sqrt(np.sum(np.square((X[:,None,:]-X2[None,:,:])/self.lengthscale),-1))[:,:,None] + ddist_dX = (X[:,None,:]-X2[None,:,:])/self.lengthscale**2/np.where(dist!=0.,dist,np.inf) dK_dX = - np.transpose(3*self.variance*dist*np.exp(-np.sqrt(3)*dist)*ddist_dX,(1,0,2)) target += np.sum(dK_dX*partial.T[:,:,None],0) @@ -104,7 +120,7 @@ class Matern32(kernpart): """ assert self.D == 1 def L(x,i): - return(3./self.lengthscales**2*F[i](x) + 2*np.sqrt(3)/self.lengthscales*F1[i](x) + F2[i](x)) + return(3./self.lengthscale**2*F[i](x) + 2*np.sqrt(3)/self.lengthscale*F1[i](x) + F2[i](x)) n = F.shape[0] G = np.zeros((n,n)) for i in range(n): @@ -114,5 +130,5 @@ class Matern32(kernpart): F1lower = np.array([f(lower) for f in F1])[:,None] #print "OLD \n", np.dot(F1lower,F1lower.T), "\n \n" #return(G) - return(self.lengthscales**3/(12.*np.sqrt(3)*self.variance) * G + 1./self.variance*np.dot(Flower,Flower.T) + self.lengthscales**2/(3.*self.variance)*np.dot(F1lower,F1lower.T)) + return(self.lengthscale**3/(12.*np.sqrt(3)*self.variance) * G + 1./self.variance*np.dot(Flower,Flower.T) + self.lengthscale**2/(3.*self.variance)*np.dot(F1lower,F1lower.T)) diff --git a/GPy/kern/exponential.py b/GPy/kern/exponential.py index 21bb8398..0ea1e922 100644 --- a/GPy/kern/exponential.py +++ b/GPy/kern/exponential.py @@ -49,13 +49,13 @@ class exponential(kernpart): def _set_params(self,x): """set the value of the parameters.""" - assert x.size==(self.D+1) + assert x.size == self.Nparam self.variance = x[0] self.lengthscale = x[1:] def _get_param_names(self): """return parameter names.""" - if self.Nparam==2: + if self.Nparam == 2: return ['variance','lengthscale'] else: return ['variance']+['lengthscale_%i'%i for i in range(self.lengthscale.size)] @@ -84,7 +84,6 @@ class exponential(kernpart): else: dl = self.variance*dvar*dist2M.sum(-1)*invdist target[1] += np.sum(dl*partial) - #foo def dKdiag_dtheta(self,partial,X,target): """derivative of the diagonal of the covariance matrix with respect to the parameters.""" From 121563cbf693f77267a77bc3f1a1274b1a5a80dc Mon Sep 17 00:00:00 2001 From: Alan Saul Date: Fri, 18 Jan 2013 16:00:10 +0000 Subject: [PATCH 17/36] Trying to get ReadTheDocs to recompile for us --- doc/conf.py | 66 ++++++++++++++++++++++++++++++++++++++++++++------- doc/index.rst | 9 +++++-- 2 files changed, 65 insertions(+), 10 deletions(-) diff --git a/doc/conf.py b/doc/conf.py index 7e1ec813..80cd5602 100644 --- a/doc/conf.py +++ b/doc/conf.py @@ -1,7 +1,7 @@ # -*- coding: utf-8 -*- # # GPy documentation build configuration file, created by -# sphinx-quickstart on Wed Jan 9 15:21:20 2013. +# sphinx-quickstart on Fri Jan 18 15:30:28 2013. # # This file is execfile()d with the current directory set to its containing dir. # @@ -25,7 +25,7 @@ import sys, os # Add any Sphinx extension module names here, as strings. They can be extensions # coming with Sphinx (named 'sphinx.ext.*') or your custom ones. -extensions = ['sphinx.ext.autodoc', 'sphinx.ext.todo', 'sphinx.ext.pngmath', 'sphinx.ext.mathjax', 'sphinx.ext.viewcode'] +extensions = ['sphinx.ext.autodoc', 'sphinx.ext.viewcode'] # Add any paths that contain templates here, relative to this directory. templates_path = ['_templates'] @@ -41,16 +41,16 @@ master_doc = 'index' # General information about the project. project = u'GPy' -copyright = u'2013, The GPy authors' +copyright = u'2013, Author' # The version info for the project you're documenting, acts as replacement for # |version| and |release|, also used in various other places throughout the # built documents. # # The short X.Y version. -version = '0.00001' +version = '' # The full version, including alpha/beta/rc tags. -release = '0.00001' +release = '' # The language for content autogenerated by Sphinx. Refer to documentation # for a list of supported languages. @@ -184,7 +184,7 @@ latex_elements = { # (source start file, target name, title, author, documentclass [howto/manual]). latex_documents = [ ('index', 'GPy.tex', u'GPy Documentation', - u'The GPy authors', 'manual'), + u'Author', 'manual'), ] # The name of an image file (relative to this directory) to place at the top of @@ -214,7 +214,7 @@ latex_documents = [ # (source start file, name, description, authors, manual section). man_pages = [ ('index', 'gpy', u'GPy Documentation', - [u'The GPy authors'], 1) + [u'Author'], 1) ] # If true, show URL addresses after external links. @@ -228,7 +228,7 @@ man_pages = [ # dir menu entry, description, category) texinfo_documents = [ ('index', 'GPy', u'GPy Documentation', - u'The GPy authors', 'GPy', 'One line description of project.', + u'Author', 'GPy', 'One line description of project.', 'Miscellaneous'), ] @@ -240,3 +240,53 @@ texinfo_documents = [ # How to display URL addresses: 'footnote', 'no', or 'inline'. #texinfo_show_urls = 'footnote' + + +# -- Options for Epub output --------------------------------------------------- + +# Bibliographic Dublin Core info. +epub_title = u'GPy' +epub_author = u'Author' +epub_publisher = u'Author' +epub_copyright = u'2013, Author' + +# The language of the text. It defaults to the language option +# or en if the language is not set. +#epub_language = '' + +# The scheme of the identifier. Typical schemes are ISBN or URL. +#epub_scheme = '' + +# The unique identifier of the text. This can be a ISBN number +# or the project homepage. +#epub_identifier = '' + +# A unique identification for the text. +#epub_uid = '' + +# A tuple containing the cover image and cover page html template filenames. +#epub_cover = () + +# HTML files that should be inserted before the pages created by sphinx. +# The format is a list of tuples containing the path and title. +#epub_pre_files = [] + +# HTML files shat should be inserted after the pages created by sphinx. +# The format is a list of tuples containing the path and title. +#epub_post_files = [] + +# A list of files that should not be packed into the epub file. +#epub_exclude_files = [] + +# The depth of the table of contents in toc.ncx. +#epub_tocdepth = 3 + +# Allow duplicate toc entries. +#epub_tocdup = True + +# ----------------------- READTHEDOCS ------------------ +on_rtd = os.environ.get('READTHEDOCS', None) == 'True' + +if on_rtd: + os.system("sphinx-apidoc -f -o . ../GPy") +#sphinx-apidoc . -o doc -H 'GPy' -f diff --git a/doc/index.rst b/doc/index.rst index 46327bb7..34d3e59c 100644 --- a/doc/index.rst +++ b/doc/index.rst @@ -1,16 +1,21 @@ .. GPy documentation master file, created by - sphinx-quickstart on Wed Jan 9 15:21:20 2013. + sphinx-quickstart on Fri Jan 18 15:30:28 2013. You can adapt this file completely to your liking, but it should at least contain the root `toctree` directive. Welcome to GPy's documentation! =============================== +Now what if I add this other hting +Just checking another thing + Contents: .. toctree:: - :maxdepth: 2 + :maxdepth: 4 + GPy + setup Indices and tables From 69743be33e207ea145d63231f96b1e6d156930cd Mon Sep 17 00:00:00 2001 From: Nicolas Date: Fri, 18 Jan 2013 16:03:20 +0000 Subject: [PATCH 18/36] last ARD flag changes to kernels --- GPy/__init__.py | 2 +- GPy/kern/Matern32.py | 6 ++-- GPy/kern/Matern52.py | 67 +++++++++++++++++++++++++--------------- GPy/kern/constructors.py | 47 +++++++++++++++++----------- GPy/kern/exponential.py | 6 ++-- GPy/kern/rbf.py | 2 +- 6 files changed, 81 insertions(+), 49 deletions(-) diff --git a/GPy/__init__.py b/GPy/__init__.py index 381d6232..0afa92e4 100644 --- a/GPy/__init__.py +++ b/GPy/__init__.py @@ -6,6 +6,6 @@ import kern import models import inference import util -import examples +#import examples #import examples TODO: discuss! from core import priors diff --git a/GPy/kern/Matern32.py b/GPy/kern/Matern32.py index 1270e3f9..cfad17c9 100644 --- a/GPy/kern/Matern32.py +++ b/GPy/kern/Matern32.py @@ -20,8 +20,10 @@ class Matern32(kernpart): :type D: int :param variance: the variance :math:`\sigma^2` :type variance: float - :param lengthscale: the lengthscale :math:`\ell_i` - :type lengthscale: np.ndarray of size (D,) + :param lengthscale: the vector of lengthscale :math:`\ell_i` + :type lengthscale: np.ndarray of size (1,) or (D,) depending on ARD + :param ARD: Auto Relevance Determination. If equal to "False", the kernel is isotropic (ie. one single lengthscale parameter \ell), otherwise there is one lengthscale parameter per dimension. + :type ARD: Boolean :rtype: kernel object """ diff --git a/GPy/kern/Matern52.py b/GPy/kern/Matern52.py index 056da32d..84c71089 100644 --- a/GPy/kern/Matern52.py +++ b/GPy/kern/Matern52.py @@ -19,42 +19,53 @@ class Matern52(kernpart): :type D: int :param variance: the variance :math:`\sigma^2` :type variance: float - :param lengthscale: the lengthscales :math:`\ell_i` - :type lengthscale: np.ndarray of size (D,) + :param lengthscale: the vector of lengthscale :math:`\ell_i` + :type lengthscale: np.ndarray of size (1,) or (D,) depending on ARD + :param ARD: Auto Relevance Determination. If equal to "False", the kernel is isotropic (ie. one single lengthscale parameter \ell), otherwise there is one lengthscale parameter per dimension. + :type ARD: Boolean :rtype: kernel object """ - def __init__(self,D,variance=1.,lengthscales=None): + def __init__(self,D,variance=1.,lengthscale=None,ARD=False): self.D = D - if lengthscales is not None: - assert lengthscales.shape==(self.D,) + self.ARD = ARD + if ARD == False: + self.Nparam = 2 + self.name = 'Mat32' + if lengthscale is not None: + assert lengthscale.shape == (1,) + else: + lengthscale = np.ones(1) else: - lengthscales = np.ones(self.D) - self.Nparam = self.D + 1 - self.name = 'Mat52' - self._set_params(np.hstack((variance,lengthscales))) + self.Nparam = self.D + 1 + self.name = 'Mat32_ARD' + if lengthscale is not None: + assert lengthscale.shape == (self.D,) + else: + lengthscale = np.ones(self.D) + self._set_params(np.hstack((variance,lengthscale))) def _get_params(self): """return the value of the parameters.""" - return np.hstack((self.variance,self.lengthscales)) + return np.hstack((self.variance,self.lengthscale)) def _set_params(self,x): """set the value of the parameters.""" - assert x.size==(self.D+1) + assert x.size == self.Nparam self.variance = x[0] - self.lengthscales = x[1:] + self.lengthscale = x[1:] def _get_param_names(self): """return parameter names.""" - if self.D==1: + if self.Nparam == 2: return ['variance','lengthscale'] else: - return ['variance']+['lengthscale_%i'%i for i in range(self.lengthscales.size)] + return ['variance']+['lengthscale_%i'%i for i in range(self.lengthscale.size)] def K(self,X,X2,target): """Compute the covariance matrix between X and X2.""" if X2 is None: X2 = X - dist = np.sqrt(np.sum(np.square((X[:,None,:]-X2[None,:,:])/self.lengthscales),-1)) + dist = np.sqrt(np.sum(np.square((X[:,None,:]-X2[None,:,:])/self.lengthscale),-1)) np.add(self.variance*(1+np.sqrt(5.)*dist+5./3*dist**2)*np.exp(-np.sqrt(5.)*dist), target,target) def Kdiag(self,X,target): @@ -64,13 +75,19 @@ class Matern52(kernpart): def dK_dtheta(self,partial,X,X2,target): """derivative of the covariance matrix with respect to the parameters.""" if X2 is None: X2 = X - dist = np.sqrt(np.sum(np.square((X[:,None,:]-X2[None,:,:])/self.lengthscales),-1)) + dist = np.sqrt(np.sum(np.square((X[:,None,:]-X2[None,:,:])/self.lengthscale),-1)) invdist = 1./np.where(dist!=0.,dist,np.inf) - dist2M = np.square(X[:,None,:]-X2[None,:,:])/self.lengthscales**3 + dist2M = np.square(X[:,None,:]-X2[None,:,:])/self.lengthscale**3 dvar = (1+np.sqrt(5.)*dist+5./3*dist**2)*np.exp(-np.sqrt(5.)*dist) - dl = (self.variance * 5./3 * dist * (1 + np.sqrt(5.)*dist ) * np.exp(-np.sqrt(5.)*dist))[:,:,np.newaxis] * dist2M*invdist[:,:,np.newaxis] target[0] += np.sum(dvar*partial) - target[1:] += (dl*partial[:,:,None]).sum(0).sum(0) + if self.ARD: + dl = (self.variance * 5./3 * dist * (1 + np.sqrt(5.)*dist ) * np.exp(-np.sqrt(5.)*dist))[:,:,np.newaxis] * dist2M*invdist[:,:,np.newaxis] + #dl = (self.variance* 3 * dist * np.exp(-np.sqrt(3.)*dist))[:,:,np.newaxis] * dist2M*invdist[:,:,np.newaxis] + target[1:] += (dl*partial[:,:,None]).sum(0).sum(0) + else: + dl = (self.variance * 5./3 * dist * (1 + np.sqrt(5.)*dist ) * np.exp(-np.sqrt(5.)*dist)) * dist2M.sum(-1)*invdist + #dl = (self.variance* 3 * dist * np.exp(-np.sqrt(3.)*dist)) * dist2M.sum(-1)*invdist + target[1] += np.sum(dl*partial) def dKdiag_dtheta(self,X,target): """derivative of the diagonal of the covariance matrix with respect to the parameters.""" @@ -79,8 +96,8 @@ class Matern52(kernpart): def dK_dX(self,partial,X,X2,target): """derivative of the covariance matrix with respect to X.""" if X2 is None: X2 = X - dist = np.sqrt(np.sum(np.square((X[:,None,:]-X2[None,:,:])/self.lengthscales),-1))[:,:,None] - ddist_dX = (X[:,None,:]-X2[None,:,:])/self.lengthscales**2/np.where(dist!=0.,dist,np.inf) + dist = np.sqrt(np.sum(np.square((X[:,None,:]-X2[None,:,:])/self.lengthscale),-1))[:,:,None] + ddist_dX = (X[:,None,:]-X2[None,:,:])/self.lengthscale**2/np.where(dist!=0.,dist,np.inf) dK_dX = - np.transpose(self.variance*5./3*dist*(1+np.sqrt(5)*dist)*np.exp(-np.sqrt(5)*dist)*ddist_dX,(1,0,2)) target += np.sum(dK_dX*partial.T[:,:,None],0) @@ -104,18 +121,18 @@ class Matern52(kernpart): """ assert self.D == 1 def L(x,i): - return(5*np.sqrt(5)/self.lengthscales**3*F[i](x) + 15./self.lengthscales**2*F1[i](x)+ 3*np.sqrt(5)/self.lengthscales*F2[i](x) + F3[i](x)) + return(5*np.sqrt(5)/self.lengthscale**3*F[i](x) + 15./self.lengthscale**2*F1[i](x)+ 3*np.sqrt(5)/self.lengthscale*F2[i](x) + F3[i](x)) n = F.shape[0] G = np.zeros((n,n)) for i in range(n): for j in range(i,n): G[i,j] = G[j,i] = integrate.quad(lambda x : L(x,i)*L(x,j),lower,upper)[0] - G_coef = 3.*self.lengthscales**5/(400*np.sqrt(5)) + G_coef = 3.*self.lengthscale**5/(400*np.sqrt(5)) Flower = np.array([f(lower) for f in F])[:,None] F1lower = np.array([f(lower) for f in F1])[:,None] F2lower = np.array([f(lower) for f in F2])[:,None] - orig = 9./8*np.dot(Flower,Flower.T) + 9.*self.lengthscales**4/200*np.dot(F2lower,F2lower.T) - orig2 = 3./5*self.lengthscales**2 * ( np.dot(F1lower,F1lower.T) + 1./8*np.dot(Flower,F2lower.T) + 1./8*np.dot(F2lower,Flower.T)) + orig = 9./8*np.dot(Flower,Flower.T) + 9.*self.lengthscale**4/200*np.dot(F2lower,F2lower.T) + orig2 = 3./5*self.lengthscale**2 * ( np.dot(F1lower,F1lower.T) + 1./8*np.dot(Flower,F2lower.T) + 1./8*np.dot(F2lower,Flower.T)) return(1./self.variance* (G_coef*G + orig + orig2)) diff --git a/GPy/kern/constructors.py b/GPy/kern/constructors.py index 5f676d9b..883357aa 100644 --- a/GPy/kern/constructors.py +++ b/GPy/kern/constructors.py @@ -32,6 +32,8 @@ def rbf(D,variance=1., lengthscale=None,ARD=False): :type variance: float :param lengthscale: the lengthscale of the kernel :type lengthscale: float + :param ARD: Auto Relevance Determination (one lengthscale per dimension) + :type ARD: Boolean """ part = rbfpart(D,variance,lengthscale,ARD) return kern(D, [part]) @@ -74,13 +76,16 @@ def white(D,variance=1.): def exponential(D,variance=1., lengthscale=None, ARD=False): """ - Construct a exponential kernel. + Construct an exponential kernel - Arguments - --------- - D (int), obligatory - variance (float) - lengthscales (np.ndarray) + :param D: dimensionality of the kernel, obligatory + :type D: int + :param variance: the variance of the kernel + :type variance: float + :param lengthscale: the lengthscale of the kernel + :type lengthscale: float + :param ARD: Auto Relevance Determination (one lengthscale per dimension) + :type ARD: Boolean """ part = exponentialpart(D,variance, lengthscale, ARD) return kern(D, [part]) @@ -89,26 +94,32 @@ def Matern32(D,variance=1., lengthscale=None, ARD=False): """ Construct a Matern 3/2 kernel. - Arguments - --------- - D (int), obligatory - variance (float) - lengthscales (np.ndarray) + :param D: dimensionality of the kernel, obligatory + :type D: int + :param variance: the variance of the kernel + :type variance: float + :param lengthscale: the lengthscale of the kernel + :type lengthscale: float + :param ARD: Auto Relevance Determination (one lengthscale per dimension) + :type ARD: Boolean """ part = Matern32part(D,variance, lengthscale, ARD) return kern(D, [part]) -def Matern52(D,variance=1., lengthscales=None): +def Matern52(D,variance=1., lengthscale=None, ARD=False): """ Construct a Matern 5/2 kernel. - Arguments - --------- - D (int), obligatory - variance (float) - lengthscales (np.ndarray) + :param D: dimensionality of the kernel, obligatory + :type D: int + :param variance: the variance of the kernel + :type variance: float + :param lengthscale: the lengthscale of the kernel + :type lengthscale: float + :param ARD: Auto Relevance Determination (one lengthscale per dimension) + :type ARD: Boolean """ - part = Matern52part(D,variance, lengthscales) + part = Matern52part(D,variance, lengthscale, ARD) return kern(D, [part]) def bias(D,variance=1.): diff --git a/GPy/kern/exponential.py b/GPy/kern/exponential.py index 0ea1e922..6c463a63 100644 --- a/GPy/kern/exponential.py +++ b/GPy/kern/exponential.py @@ -19,8 +19,10 @@ class exponential(kernpart): :type D: int :param variance: the variance :math:`\sigma^2` :type variance: float - :param lengthscale: the lengthscales :math:`\ell_i` - :type lengthscale: np.ndarray of size (D,) + :param lengthscale: the vector of lengthscale :math:`\ell_i` + :type lengthscale: np.ndarray of size (1,) or (D,) depending on ARD + :param ARD: Auto Relevance Determination. If equal to "False", the kernel is isotropic (ie. one single lengthscale parameter \ell), otherwise there is one lengthscale parameter per dimension. + :type ARD: Boolean :rtype: kernel object """ diff --git a/GPy/kern/rbf.py b/GPy/kern/rbf.py index 13ecaf86..62a9c46d 100644 --- a/GPy/kern/rbf.py +++ b/GPy/kern/rbf.py @@ -21,7 +21,7 @@ class rbf(kernpart): :param variance: the variance of the kernel :type variance: float :param lengthscale: the vector of lengthscale of the kernel - :type lengthscale: np.ndarray + :type lengthscale: np.ndarray od size (1,) or (D,) depending on ARD :param ARD: Auto Relevance Determination. If equal to "False", the kernel is isotropic (ie. one single lengthscale parameter \ell), otherwise there is one lengthscale parameter per dimension. :type ARD: Boolean From f3617eb07184f1926e0bb756ab92014582ee385e Mon Sep 17 00:00:00 2001 From: Alan Saul Date: Fri, 18 Jan 2013 16:05:50 +0000 Subject: [PATCH 19/36] Testing docs more --- doc/conf.py | 2 +- doc/index.rst | 1 + 2 files changed, 2 insertions(+), 1 deletion(-) diff --git a/doc/conf.py b/doc/conf.py index 80cd5602..268f4d0a 100644 --- a/doc/conf.py +++ b/doc/conf.py @@ -288,5 +288,5 @@ epub_copyright = u'2013, Author' on_rtd = os.environ.get('READTHEDOCS', None) == 'True' if on_rtd: - os.system("sphinx-apidoc -f -o . ../GPy") + os.system("sphinx-apidoc -f -o docs .") #sphinx-apidoc . -o doc -H 'GPy' -f diff --git a/doc/index.rst b/doc/index.rst index 34d3e59c..f7ce7e1f 100644 --- a/doc/index.rst +++ b/doc/index.rst @@ -6,6 +6,7 @@ Welcome to GPy's documentation! =============================== +And another thing Now what if I add this other hting Just checking another thing From 49530b07ee58cd095db53a93df723594172800ba Mon Sep 17 00:00:00 2001 From: Alan Saul Date: Fri, 18 Jan 2013 16:06:42 +0000 Subject: [PATCH 20/36] Adding fake kernel to test docs --- GPy/kern/rbf-testing.py | 178 ++++++++++++++++++++++++++++++++++++++++ 1 file changed, 178 insertions(+) create mode 100644 GPy/kern/rbf-testing.py diff --git a/GPy/kern/rbf-testing.py b/GPy/kern/rbf-testing.py new file mode 100644 index 00000000..785e8fb5 --- /dev/null +++ b/GPy/kern/rbf-testing.py @@ -0,0 +1,178 @@ +# Copyright (c) 2012, GPy authors (see AUTHORS.txt). +# Licensed under the BSD 3-clause license (see LICENSE.txt) + + +from kernpart import kernpart +import numpy as np +import hashlib + +class rbf_testing(kernpart): + """ + Radial Basis Function kernel, aka squared-exponential, exponentiated quadratic or Gaussian kernel. + + .. math:: + + k(r) = \sigma^2 \exp(- \frac{r^2}{2\ell}) \qquad \qquad \\text{ where } r = \sqrt{\frac{\sum_{i=1}^d (x_i-x^\prime_i)^2}{\ell^2}} + + where \ell is the lengthscale, \alpha the smoothness, \sigma^2 the variance and d the dimensionality of the input. + + :param D: the number of input dimensions + :type D: int + :param variance: the variance of the kernel + :type variance: float + :param lengthscale: the lengthscale of the kernel + :type lengthscale: float + + .. Note: for rbf_testing with different lengthscale on each dimension, see rbf_testing_ARD + """ + + def __init__(self,D,variance=1.,lengthscale=1.): + self.D = D + self.Nparam = 2 + self.name = 'rbf_testing' + self._set_params(np.hstack((variance,lengthscale))) + + #initialize cache + self._Z, self._mu, self._S = np.empty(shape=(3,1)) + self._X, self._X2, self._params = np.empty(shape=(3,1)) + + def _get_params(self): + return np.hstack((self.variance,self.lengthscale)) + + def _set_params(self,x): + self.variance, self.lengthscale = x + self.lengthscale2 = np.square(self.lengthscale) + #reset cached results + self._X, self._X2, self._params = np.empty(shape=(3,1)) + self._Z, self._mu, self._S = np.empty(shape=(3,1)) # cached versions of Z,mu,S + + def _get_param_names(self): + return ['variance','lengthscale'] + + def K(self,X,X2,target): + if X2 is None: + X2 = X + self._K_computations(X,X2) + np.add(self.variance*self._K_dvar, target,target) + + def Kdiag(self,X,target): + np.add(target,self.variance,target) + + def dK_dtheta(self,partial,X,X2,target): + self._K_computations(X,X2) + target[0] += np.sum(self._K_dvar*partial) + target[1] += np.sum(self._K_dvar*self.variance*self._K_dist2/self.lengthscale*partial) + + def dKdiag_dtheta(self,partial,X,target): + #NB: derivative of diagonal elements wrt lengthscale is 0 + target[0] += np.sum(partial) + + def dK_dX(self,partial,X,X2,target): + self._K_computations(X,X2) + _K_dist = X[:,None,:]-X2[None,:,:] + dK_dX = np.transpose(-self.variance*self._K_dvar[:,:,np.newaxis]*_K_dist/self.lengthscale2,(1,0,2)) + target += np.sum(dK_dX*partial.T[:,:,None],0) + + def dKdiag_dX(self,partial,X,target): + pass + + def _K_computations(self,X,X2): + if not (np.all(X==self._X) and np.all(X2==self._X2)): + self._X = X + self._X2 = X2 + if X2 is None: X2 = X + XXT = np.dot(X,X2.T) + if X is X2: + self._K_dist2 = (-2.*XXT + np.diag(XXT)[:,np.newaxis] + np.diag(XXT)[np.newaxis,:])/self.lengthscale2 + else: + self._K_dist2 = (-2.*XXT + np.sum(np.square(X),1)[:,None] + np.sum(np.square(X2),1)[None,:])/self.lengthscale2 + # TODO Remove comments if this is fine. + # Commented out by Neil as doesn't seem to be used elsewhere. + #self._K_exponent = -0.5*self._K_dist2 + self._K_dvar = np.exp(-0.5*self._K_dist2) + + def psi0(self,Z,mu,S,target): + target += self.variance + + def dpsi0_dtheta(self,partial,Z,mu,S,target): + target[0] += 1. + + def dpsi0_dmuS(self,Z,mu,S,target_mu,target_S): + pass + + def psi1(self,Z,mu,S,target): + self._psi_computations(Z,mu,S) + target += self._psi1 + + def dpsi1_dtheta(self,partial,Z,mu,S,target): + self._psi_computations(Z,mu,S) + denom_deriv = S[:,None,:]/(self.lengthscale**3+self.lengthscale*S[:,None,:]) + d_length = self._psi1[:,:,None]*(self.lengthscale*np.square(self._psi1_dist/(self.lengthscale2+S[:,None,:])) + denom_deriv) + target[0] += np.sum(partial*self._psi1/self.variance) + target[1] += np.sum(d_length*partial[:,:,None]) + + def dpsi1_dZ(self,partial,Z,mu,S,target): + self._psi_computations(Z,mu,S) + target += np.sum(partial[:,:,None]*-self._psi1[:,:,None]*self._psi1_dist/self.lengthscale2/self._psi1_denom,0) + + def dpsi1_dmuS(self,partial,Z,mu,S,target_mu,target_S): + self._psi_computations(Z,mu,S) + tmp = self._psi1[:,:,None]/self.lengthscale2/self._psi1_denom + target_mu += np.sum(partial*tmp*self._psi1_dist,1) + target_S += np.sum(partial*0.5*tmp*(self._psi1_dist_sq-1),1) + + def psi2(self,Z,mu,S,target): + self._psi_computations(Z,mu,S) + target += self._psi2.sum(0) #TODO: psi2 should be NxMxM (for het. noise) + + def dpsi2_dtheta(self,partial,Z,mu,S,target): + """Shape N,M,M,Ntheta""" + self._psi_computations(Z,mu,S) + d_var = np.sum(2.*self._psi2/self.variance,0) + d_length = self._psi2[:,:,:,None]*(0.5*self._psi2_Zdist_sq*self._psi2_denom + 2.*self._psi2_mudist_sq + 2.*S[:,None,None,:]/self.lengthscale2)/(self.lengthscale*self._psi2_denom) + d_length = d_length.sum(0) + target[0] += np.sum(partial*d_var) + target[1] += np.sum(d_length*partial) + + def dpsi2_dZ(self,partial,Z,mu,S,target): + """Returns shape N,M,M,Q""" + self._psi_computations(Z,mu,S) + dZ = self._psi2[:,:,:,None]/self.lengthscale2*(-0.5*self._psi2_Zdist + self._psi2_mudist/self._psi2_denom) + target += np.sum(partial[None,:,:,None]*dZ,0).sum(1) + + def dpsi2_dmuS(self,Z,mu,S,target_mu,target_S): + """Think N,M,M,Q """ + self._psi_computations(Z,mu,S) + tmp = self._psi2[:,:,:,None]/self.lengthscale2/self._psi2_denom + target_mu += (partial*-tmp*2.*self._psi2_mudist).sum(1).sum(1) + target_S += (partial*tmp*(2.*self._psi2_mudist_sq-1)).sum(1).sum(1) + + def _psi_computations(self,Z,mu,S): + #here are the "statistics" for psi1 and psi2 + if not np.all(Z==self._Z): + #Z has changed, compute Z specific stuff + self._psi2_Zhat = 0.5*(Z[:,None,:] +Z[None,:,:]) # M,M,Q + self._psi2_Zdist = Z[:,None,:]-Z[None,:,:] # M,M,Q + self._psi2_Zdist_sq = np.square(self._psi2_Zdist)/self.lengthscale2 # M,M,Q + self._Z = Z + + if not (np.all(Z==self._Z) and np.all(mu==self._mu) and np.all(S==self._S)): + #something's changed. recompute EVERYTHING + + #TODO: make more efficient for large Q (using NDL's dot product trick) + #psi1 + self._psi1_denom = S[:,None,:]/self.lengthscale2 + 1. + self._psi1_dist = Z[None,:,:]-mu[:,None,:] + self._psi1_dist_sq = np.square(self._psi1_dist)/self.lengthscale2/self._psi1_denom + self._psi1_exponent = -0.5*np.sum(self._psi1_dist_sq+np.log(self._psi1_denom),-1) + self._psi1 = self.variance*np.exp(self._psi1_exponent) + + #psi2 + self._psi2_denom = 2.*S[:,None,None,:]/self.lengthscale2+1. # N,M,M,Q + self._psi2_mudist = mu[:,None,None,:]-self._psi2_Zhat #N,M,M,Q + self._psi2_mudist_sq = np.square(self._psi2_mudist)/(self.lengthscale2*self._psi2_denom) + self._psi2_exponent = np.sum(-self._psi2_Zdist_sq/4. -self._psi2_mudist_sq -0.5*np.log(self._psi2_denom),-1) #N,M,M + self._psi2 = np.square(self.variance)*np.exp(self._psi2_exponent) # N,M,M + + self._Z, self._mu, self._S = Z, mu,S + From ad41c63d08cc250436ff9cb4530f2e178d24e7cf Mon Sep 17 00:00:00 2001 From: Alan Saul Date: Fri, 18 Jan 2013 16:10:51 +0000 Subject: [PATCH 21/36] Adding old command I read on the internet for own build --- doc/conf.py | 2 +- doc/index.rst | 1 + 2 files changed, 2 insertions(+), 1 deletion(-) diff --git a/doc/conf.py b/doc/conf.py index 268f4d0a..80cd5602 100644 --- a/doc/conf.py +++ b/doc/conf.py @@ -288,5 +288,5 @@ epub_copyright = u'2013, Author' on_rtd = os.environ.get('READTHEDOCS', None) == 'True' if on_rtd: - os.system("sphinx-apidoc -f -o docs .") + os.system("sphinx-apidoc -f -o . ../GPy") #sphinx-apidoc . -o doc -H 'GPy' -f diff --git a/doc/index.rst b/doc/index.rst index f7ce7e1f..fb721e15 100644 --- a/doc/index.rst +++ b/doc/index.rst @@ -6,6 +6,7 @@ Welcome to GPy's documentation! =============================== +Added old command again And another thing Now what if I add this other hting Just checking another thing From ffd05027cbdc4213a670c337859cd783001278ba Mon Sep 17 00:00:00 2001 From: Nicolas Date: Fri, 18 Jan 2013 16:14:13 +0000 Subject: [PATCH 22/36] More tests in unit_tests --- GPy/kern/rbf.py | 1 + GPy/testing/unit_tests.py | 41 ++++++++++++++++++++++++++------------- 2 files changed, 29 insertions(+), 13 deletions(-) diff --git a/GPy/kern/rbf.py b/GPy/kern/rbf.py index 62a9c46d..1e9b4379 100644 --- a/GPy/kern/rbf.py +++ b/GPy/kern/rbf.py @@ -24,6 +24,7 @@ class rbf(kernpart): :type lengthscale: np.ndarray od size (1,) or (D,) depending on ARD :param ARD: Auto Relevance Determination. If equal to "False", the kernel is isotropic (ie. one single lengthscale parameter \ell), otherwise there is one lengthscale parameter per dimension. :type ARD: Boolean + :rtype: kernel object """ diff --git a/GPy/testing/unit_tests.py b/GPy/testing/unit_tests.py index d2ef87f7..c5db80bd 100644 --- a/GPy/testing/unit_tests.py +++ b/GPy/testing/unit_tests.py @@ -42,51 +42,66 @@ class GradientTests(unittest.TestCase): # contrain all parameters to be positive self.assertTrue(m.checkgrad()) - def test_gp_regression_rbf_white_kern_1d(self): + def test_gp_regression_rbf_1d(self): ''' Testing the GP regression with rbf kernel with white kernel on 1d data ''' rbf = GPy.kern.rbf(1) self.check_model_with_white(rbf, model_type='GP_regression', dimension=1) - def test_GP_regression_rbf_ARD_white_kern_2D(self): - ''' Testing the GP regression with rbf and white kernel on 2d data ''' - k = GPy.kern.rbf_ARD(2) - self.check_model_with_white(k, model_type='GP_regression', dimension=2) - - def test_GP_regression_rbf_white_kern_2D(self): + def test_GP_regression_rbf_2D(self): ''' Testing the GP regression with rbf and white kernel on 2d data ''' rbf = GPy.kern.rbf(2) self.check_model_with_white(rbf, model_type='GP_regression', dimension=2) - def test_GP_regression_matern52_kern_1D(self): + def test_GP_regression_rbf_ARD_2D(self): + ''' Testing the GP regression with rbf and white kernel on 2d data ''' + k = GPy.kern.rbf(2,ARD=True) + self.check_model_with_white(k, model_type='GP_regression', dimension=2) + + def test_GP_regression_matern52_1D(self): ''' Testing the GP regression with matern52 kernel on 1d data ''' matern52 = GPy.kern.Matern52(1) self.check_model_with_white(matern52, model_type='GP_regression', dimension=1) - def test_GP_regression_matern52_kern_2D(self): + def test_GP_regression_matern52_2D(self): ''' Testing the GP regression with matern52 kernel on 2d data ''' matern52 = GPy.kern.Matern52(2) self.check_model_with_white(matern52, model_type='GP_regression', dimension=2) - def test_GP_regression_matern32_kern_1D(self): + def test_GP_regression_matern52_ARD_2D(self): + ''' Testing the GP regression with matern52 kernel on 2d data ''' + matern52 = GPy.kern.Matern52(2,ARD=True) + self.check_model_with_white(matern52, model_type='GP_regression', dimension=2) + + def test_GP_regression_matern32_1D(self): ''' Testing the GP regression with matern32 kernel on 1d data ''' matern32 = GPy.kern.Matern32(1) self.check_model_with_white(matern32, model_type='GP_regression', dimension=1) - def test_GP_regression_matern32_kern_2D(self): + def test_GP_regression_matern32_2D(self): ''' Testing the GP regression with matern32 kernel on 2d data ''' matern32 = GPy.kern.Matern32(2) self.check_model_with_white(matern32, model_type='GP_regression', dimension=2) - def test_GP_regression_exponential_kern_1D(self): + def test_GP_regression_matern32_ARD_2D(self): + ''' Testing the GP regression with matern32 kernel on 2d data ''' + matern32 = GPy.kern.Matern32(2,ARD=True) + self.check_model_with_white(matern32, model_type='GP_regression', dimension=2) + + def test_GP_regression_exponential_1D(self): ''' Testing the GP regression with exponential kernel on 1d data ''' exponential = GPy.kern.exponential(1) self.check_model_with_white(exponential, model_type='GP_regression', dimension=1) - def test_GP_regression_exponential_kern_2D(self): + def test_GP_regression_exponential_2D(self): ''' Testing the GP regression with exponential kernel on 2d data ''' exponential = GPy.kern.exponential(2) self.check_model_with_white(exponential, model_type='GP_regression', dimension=2) + def test_GP_regression_exponential_ARD_2D(self): + ''' Testing the GP regression with exponential kernel on 2d data ''' + exponential = GPy.kern.exponential(2,ARD=True) + self.check_model_with_white(exponential, model_type='GP_regression', dimension=2) + def test_GP_regression_bias_kern_1D(self): ''' Testing the GP regression with bias kernel on 1d data ''' bias = GPy.kern.bias(1) From 8696de241ac86e527dae8168068cceafd41fb247 Mon Sep 17 00:00:00 2001 From: Alan Saul Date: Fri, 18 Jan 2013 16:27:58 +0000 Subject: [PATCH 23/36] Tried changing the location of the apidoc compilation --- doc/conf.py | 4 +- doc/make.bat | 190 +++++++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 192 insertions(+), 2 deletions(-) create mode 100644 doc/make.bat diff --git a/doc/conf.py b/doc/conf.py index 80cd5602..bd8420c4 100644 --- a/doc/conf.py +++ b/doc/conf.py @@ -288,5 +288,5 @@ epub_copyright = u'2013, Author' on_rtd = os.environ.get('READTHEDOCS', None) == 'True' if on_rtd: - os.system("sphinx-apidoc -f -o . ../GPy") -#sphinx-apidoc . -o doc -H 'GPy' -f + os.system("pwd") + os.system("sphinx-apidoc -f -o . ../") diff --git a/doc/make.bat b/doc/make.bat new file mode 100644 index 00000000..40a74901 --- /dev/null +++ b/doc/make.bat @@ -0,0 +1,190 @@ +@ECHO OFF + +REM Command file for Sphinx documentation + +if "%SPHINXBUILD%" == "" ( + set SPHINXBUILD=sphinx-build +) +set BUILDDIR=_build +set ALLSPHINXOPTS=-d %BUILDDIR%/doctrees %SPHINXOPTS% . +set I18NSPHINXOPTS=%SPHINXOPTS% . +if NOT "%PAPER%" == "" ( + set ALLSPHINXOPTS=-D latex_paper_size=%PAPER% %ALLSPHINXOPTS% + set I18NSPHINXOPTS=-D latex_paper_size=%PAPER% %I18NSPHINXOPTS% +) + +if "%1" == "" goto help + +if "%1" == "help" ( + :help + echo.Please use `make ^` where ^ is one of + echo. html to make standalone HTML files + echo. dirhtml to make HTML files named index.html in directories + echo. singlehtml to make a single large HTML file + echo. pickle to make pickle files + echo. json to make JSON files + echo. htmlhelp to make HTML files and a HTML help project + echo. qthelp to make HTML files and a qthelp project + echo. devhelp to make HTML files and a Devhelp project + echo. epub to make an epub + echo. latex to make LaTeX files, you can set PAPER=a4 or PAPER=letter + echo. text to make text files + echo. man to make manual pages + echo. texinfo to make Texinfo files + echo. gettext to make PO message catalogs + echo. changes to make an overview over all changed/added/deprecated items + echo. linkcheck to check all external links for integrity + echo. doctest to run all doctests embedded in the documentation if enabled + goto end +) + +if "%1" == "clean" ( + for /d %%i in (%BUILDDIR%\*) do rmdir /q /s %%i + del /q /s %BUILDDIR%\* + goto end +) + +if "%1" == "html" ( + %SPHINXBUILD% -b html %ALLSPHINXOPTS% %BUILDDIR%/html + if errorlevel 1 exit /b 1 + echo. + echo.Build finished. The HTML pages are in %BUILDDIR%/html. + goto end +) + +if "%1" == "dirhtml" ( + %SPHINXBUILD% -b dirhtml %ALLSPHINXOPTS% %BUILDDIR%/dirhtml + if errorlevel 1 exit /b 1 + echo. + echo.Build finished. The HTML pages are in %BUILDDIR%/dirhtml. + goto end +) + +if "%1" == "singlehtml" ( + %SPHINXBUILD% -b singlehtml %ALLSPHINXOPTS% %BUILDDIR%/singlehtml + if errorlevel 1 exit /b 1 + echo. + echo.Build finished. The HTML pages are in %BUILDDIR%/singlehtml. + goto end +) + +if "%1" == "pickle" ( + %SPHINXBUILD% -b pickle %ALLSPHINXOPTS% %BUILDDIR%/pickle + if errorlevel 1 exit /b 1 + echo. + echo.Build finished; now you can process the pickle files. + goto end +) + +if "%1" == "json" ( + %SPHINXBUILD% -b json %ALLSPHINXOPTS% %BUILDDIR%/json + if errorlevel 1 exit /b 1 + echo. + echo.Build finished; now you can process the JSON files. + goto end +) + +if "%1" == "htmlhelp" ( + %SPHINXBUILD% -b htmlhelp %ALLSPHINXOPTS% %BUILDDIR%/htmlhelp + if errorlevel 1 exit /b 1 + echo. + echo.Build finished; now you can run HTML Help Workshop with the ^ +.hhp project file in %BUILDDIR%/htmlhelp. + goto end +) + +if "%1" == "qthelp" ( + %SPHINXBUILD% -b qthelp %ALLSPHINXOPTS% %BUILDDIR%/qthelp + if errorlevel 1 exit /b 1 + echo. + echo.Build finished; now you can run "qcollectiongenerator" with the ^ +.qhcp project file in %BUILDDIR%/qthelp, like this: + echo.^> qcollectiongenerator %BUILDDIR%\qthelp\GPy.qhcp + echo.To view the help file: + echo.^> assistant -collectionFile %BUILDDIR%\qthelp\GPy.ghc + goto end +) + +if "%1" == "devhelp" ( + %SPHINXBUILD% -b devhelp %ALLSPHINXOPTS% %BUILDDIR%/devhelp + if errorlevel 1 exit /b 1 + echo. + echo.Build finished. + goto end +) + +if "%1" == "epub" ( + %SPHINXBUILD% -b epub %ALLSPHINXOPTS% %BUILDDIR%/epub + if errorlevel 1 exit /b 1 + echo. + echo.Build finished. The epub file is in %BUILDDIR%/epub. + goto end +) + +if "%1" == "latex" ( + %SPHINXBUILD% -b latex %ALLSPHINXOPTS% %BUILDDIR%/latex + if errorlevel 1 exit /b 1 + echo. + echo.Build finished; the LaTeX files are in %BUILDDIR%/latex. + goto end +) + +if "%1" == "text" ( + %SPHINXBUILD% -b text %ALLSPHINXOPTS% %BUILDDIR%/text + if errorlevel 1 exit /b 1 + echo. + echo.Build finished. The text files are in %BUILDDIR%/text. + goto end +) + +if "%1" == "man" ( + %SPHINXBUILD% -b man %ALLSPHINXOPTS% %BUILDDIR%/man + if errorlevel 1 exit /b 1 + echo. + echo.Build finished. The manual pages are in %BUILDDIR%/man. + goto end +) + +if "%1" == "texinfo" ( + %SPHINXBUILD% -b texinfo %ALLSPHINXOPTS% %BUILDDIR%/texinfo + if errorlevel 1 exit /b 1 + echo. + echo.Build finished. The Texinfo files are in %BUILDDIR%/texinfo. + goto end +) + +if "%1" == "gettext" ( + %SPHINXBUILD% -b gettext %I18NSPHINXOPTS% %BUILDDIR%/locale + if errorlevel 1 exit /b 1 + echo. + echo.Build finished. The message catalogs are in %BUILDDIR%/locale. + goto end +) + +if "%1" == "changes" ( + %SPHINXBUILD% -b changes %ALLSPHINXOPTS% %BUILDDIR%/changes + if errorlevel 1 exit /b 1 + echo. + echo.The overview file is in %BUILDDIR%/changes. + goto end +) + +if "%1" == "linkcheck" ( + %SPHINXBUILD% -b linkcheck %ALLSPHINXOPTS% %BUILDDIR%/linkcheck + if errorlevel 1 exit /b 1 + echo. + echo.Link check complete; look for any errors in the above output ^ +or in %BUILDDIR%/linkcheck/output.txt. + goto end +) + +if "%1" == "doctest" ( + %SPHINXBUILD% -b doctest %ALLSPHINXOPTS% %BUILDDIR%/doctest + if errorlevel 1 exit /b 1 + echo. + echo.Testing of doctests in the sources finished, look at the ^ +results in %BUILDDIR%/doctest/output.txt. + goto end +) + +:end From e061e1a638a22f4d00b5c63e86fb3e2d71656d02 Mon Sep 17 00:00:00 2001 From: Alan Saul Date: Fri, 18 Jan 2013 16:33:59 +0000 Subject: [PATCH 24/36] Tried fixing build call --- doc/conf.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/doc/conf.py b/doc/conf.py index bd8420c4..fa6f8c34 100644 --- a/doc/conf.py +++ b/doc/conf.py @@ -289,4 +289,4 @@ on_rtd = os.environ.get('READTHEDOCS', None) == 'True' if on_rtd: os.system("pwd") - os.system("sphinx-apidoc -f -o . ../") + os.system("sphinx-apidoc -f -o . ./GPy") From b64d6e4acfdca47b23d57741a4977b7b9e1b93d6 Mon Sep 17 00:00:00 2001 From: Alan Saul Date: Fri, 18 Jan 2013 16:36:08 +0000 Subject: [PATCH 25/36] The below again --- doc/conf.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/doc/conf.py b/doc/conf.py index fa6f8c34..cf9f7ca7 100644 --- a/doc/conf.py +++ b/doc/conf.py @@ -289,4 +289,4 @@ on_rtd = os.environ.get('READTHEDOCS', None) == 'True' if on_rtd: os.system("pwd") - os.system("sphinx-apidoc -f -o . ./GPy") + os.system("sphinx-apidoc -f -o . ../GPy") From 3045ee72c8c6049e5adc90e43840a9685329e5f8 Mon Sep 17 00:00:00 2001 From: Neil Lawrence Date: Fri, 18 Jan 2013 16:38:41 +0000 Subject: [PATCH 26/36] Added unsupervised.py examples file and datasets.py@ --- GPy/examples/unsupervised.py | 25 +++++++++++++++++++++++++ 1 file changed, 25 insertions(+) create mode 100644 GPy/examples/unsupervised.py diff --git a/GPy/examples/unsupervised.py b/GPy/examples/unsupervised.py new file mode 100644 index 00000000..08d81e05 --- /dev/null +++ b/GPy/examples/unsupervised.py @@ -0,0 +1,25 @@ +""" +Usupervised learning with Gaussian Processes. +""" +import pylab as pb +import numpy as np +import GPy + + +###################################### +## Oil data subsampled to 100 points. +def oil_100(): + data = GPy.util.datasets.oil_100() + + # create simple GP model + m = GPy.models.GPLVM(data['X'], 2) + + + # optimize + m.ensure_default_constraints() + m.optimize() + + # plot + print(m) + return m + From 08ecb2f97f40272a1598f8cc46553ab92c15d6e8 Mon Sep 17 00:00:00 2001 From: Alan Saul Date: Fri, 18 Jan 2013 16:41:18 +0000 Subject: [PATCH 27/36] Added pylab to requirements --- setup.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/setup.py b/setup.py index ee44dde0..4934e072 100644 --- a/setup.py +++ b/setup.py @@ -26,7 +26,7 @@ setup(name = 'GPy', long_description=read('README.md'), #ext_modules = [Extension(name = 'GPy.kern.lfmUpsilonf2py', # sources = ['GPy/kern/src/lfmUpsilonf2py.f90'])], - install_requires=['numpy>=1.6', 'scipy','matplotlib>=1.1'], + install_requires=['pylab, numpy>=1.6', 'scipy','matplotlib>=1.1'], setup_requires=['sphinx'], cmdclass = {'build_sphinx': BuildDoc}, classifiers=[ From e4e05ea5e5082a73b8e018ee54867b612d0898a7 Mon Sep 17 00:00:00 2001 From: Alan Saul Date: Fri, 18 Jan 2013 17:05:57 +0000 Subject: [PATCH 28/36] Trying to give pylab dependency --- setup.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/setup.py b/setup.py index 4934e072..b8b8b24b 100644 --- a/setup.py +++ b/setup.py @@ -26,7 +26,7 @@ setup(name = 'GPy', long_description=read('README.md'), #ext_modules = [Extension(name = 'GPy.kern.lfmUpsilonf2py', # sources = ['GPy/kern/src/lfmUpsilonf2py.f90'])], - install_requires=['pylab, numpy>=1.6', 'scipy','matplotlib>=1.1'], + install_requires=['pylab', 'numpy>=1.6', 'scipy','matplotlib>=1.1'], setup_requires=['sphinx'], cmdclass = {'build_sphinx': BuildDoc}, classifiers=[ From 6089a34d8d96123de29cde40b2ff65aafe3d179c Mon Sep 17 00:00:00 2001 From: Alan Saul Date: Fri, 18 Jan 2013 17:32:46 +0000 Subject: [PATCH 29/36] added sympy dependency and scipy version dependency again --- setup.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/setup.py b/setup.py index b8b8b24b..432b8b13 100644 --- a/setup.py +++ b/setup.py @@ -26,7 +26,7 @@ setup(name = 'GPy', long_description=read('README.md'), #ext_modules = [Extension(name = 'GPy.kern.lfmUpsilonf2py', # sources = ['GPy/kern/src/lfmUpsilonf2py.f90'])], - install_requires=['pylab', 'numpy>=1.6', 'scipy','matplotlib>=1.1'], + install_requires=['sympy', 'numpy>=1.6', 'scipy>=0.9','matplotlib>=1.1'], setup_requires=['sphinx'], cmdclass = {'build_sphinx': BuildDoc}, classifiers=[ From a76c53e6996712295b4f98b5ce21564f7165a769 Mon Sep 17 00:00:00 2001 From: Nicolas Date: Fri, 18 Jan 2013 17:43:32 +0000 Subject: [PATCH 30/36] bug fixed in example (in regression.py) --- GPy/kern/rbf.py | 14 ++++++-------- GPy/util/datasets.py | 2 +- 2 files changed, 7 insertions(+), 9 deletions(-) diff --git a/GPy/kern/rbf.py b/GPy/kern/rbf.py index 1e9b4379..b1134e25 100644 --- a/GPy/kern/rbf.py +++ b/GPy/kern/rbf.py @@ -8,13 +8,13 @@ import hashlib class rbf(kernpart): """ - Radial Basis Function kernel, aka squared-exponential, exponentiated quadratic or Gaussian kernel. + Radial Basis Function kernel, aka squared-exponential, exponentiated quadratic or Gaussian kernel: .. math:: - k(r) = \sigma^2 \exp(- \frac{r^2}{2\ell}) \qquad \qquad \\text{ where } r = \sqrt{\frac{\sum_{i=1}^d (x_i-x^\prime_i)^2}{\ell^2}} + k(r) = \sigma^2 \exp(- \frac{1}{2}r^2) \qquad \qquad \\text{ where } r^2 = \sum_{i=1}^d \frac{ (x_i-x^\prime_i)^2}{\ell_i^2}} - where \ell is the lengthscale, \alpha the smoothness, \sigma^2 the variance and d the dimensionality of the input. + where \ell_i is the lengthscale, \sigma^2 the variance and d the dimensionality of the input. :param D: the number of input dimensions :type D: int @@ -37,8 +37,7 @@ class rbf(kernpart): if lengthscale is not None: assert lengthscale.shape == (1,) else: - lengthscale = np.ones(1) - + lengthscale = np.ones(1) else: self.Nparam = self.D + 1 self.name = 'rbf_ARD' @@ -109,11 +108,10 @@ class rbf(kernpart): self._X2 = X2 if X2 is None: X2 = X self._K_dist = X[:,None,:]-X2[None,:,:] # this can be computationally heavy - self._params = np.empty(shape=(1,0))#ensure the next section gets called + self._params = np.empty(shape=(1,0)) #ensure the next section gets called if not np.all(self._params == self._get_params()): self._params == self._get_params() - self._K_dist2 = np.square(self._K_dist/self.lengthscale) - #self._K_exponent = -0.5*self._K_dist2.sum(-1) #ND: commented out because seems not to be used + self._K_dist2 = np.square(self._K_dist/self.lengthscale) self._K_dvar = np.exp(-0.5*self._K_dist2.sum(-1)) def psi0(self,Z,mu,S,target): diff --git a/GPy/util/datasets.py b/GPy/util/datasets.py index bc7bf546..071db5d6 100644 --- a/GPy/util/datasets.py +++ b/GPy/util/datasets.py @@ -90,7 +90,7 @@ def toy_rbf_1d(seed=default_seed): N = 500 X = np.random.uniform(low=-1.0, high=1.0, size=(N, numIn)) X.sort(axis=0) - rbf = GPy.kern.rbf(numIn, variance=1., lengthscale=0.25) + rbf = GPy.kern.rbf(numIn, variance=1., lengthscale=np.array((0.25,))) white = GPy.kern.white(numIn, variance=1e-2) kernel = rbf + white K = kernel.K(X) From 3cac7c5094a6fae14f6ad7968905e7a453acff7f Mon Sep 17 00:00:00 2001 From: Alan Saul Date: Fri, 18 Jan 2013 17:44:25 +0000 Subject: [PATCH 31/36] Updated index.rst --- GPy/kern/rbf-testing.py | 178 ---------------------------------------- doc/index.rst | 8 +- 2 files changed, 1 insertion(+), 185 deletions(-) delete mode 100644 GPy/kern/rbf-testing.py diff --git a/GPy/kern/rbf-testing.py b/GPy/kern/rbf-testing.py deleted file mode 100644 index 785e8fb5..00000000 --- a/GPy/kern/rbf-testing.py +++ /dev/null @@ -1,178 +0,0 @@ -# Copyright (c) 2012, GPy authors (see AUTHORS.txt). -# Licensed under the BSD 3-clause license (see LICENSE.txt) - - -from kernpart import kernpart -import numpy as np -import hashlib - -class rbf_testing(kernpart): - """ - Radial Basis Function kernel, aka squared-exponential, exponentiated quadratic or Gaussian kernel. - - .. math:: - - k(r) = \sigma^2 \exp(- \frac{r^2}{2\ell}) \qquad \qquad \\text{ where } r = \sqrt{\frac{\sum_{i=1}^d (x_i-x^\prime_i)^2}{\ell^2}} - - where \ell is the lengthscale, \alpha the smoothness, \sigma^2 the variance and d the dimensionality of the input. - - :param D: the number of input dimensions - :type D: int - :param variance: the variance of the kernel - :type variance: float - :param lengthscale: the lengthscale of the kernel - :type lengthscale: float - - .. Note: for rbf_testing with different lengthscale on each dimension, see rbf_testing_ARD - """ - - def __init__(self,D,variance=1.,lengthscale=1.): - self.D = D - self.Nparam = 2 - self.name = 'rbf_testing' - self._set_params(np.hstack((variance,lengthscale))) - - #initialize cache - self._Z, self._mu, self._S = np.empty(shape=(3,1)) - self._X, self._X2, self._params = np.empty(shape=(3,1)) - - def _get_params(self): - return np.hstack((self.variance,self.lengthscale)) - - def _set_params(self,x): - self.variance, self.lengthscale = x - self.lengthscale2 = np.square(self.lengthscale) - #reset cached results - self._X, self._X2, self._params = np.empty(shape=(3,1)) - self._Z, self._mu, self._S = np.empty(shape=(3,1)) # cached versions of Z,mu,S - - def _get_param_names(self): - return ['variance','lengthscale'] - - def K(self,X,X2,target): - if X2 is None: - X2 = X - self._K_computations(X,X2) - np.add(self.variance*self._K_dvar, target,target) - - def Kdiag(self,X,target): - np.add(target,self.variance,target) - - def dK_dtheta(self,partial,X,X2,target): - self._K_computations(X,X2) - target[0] += np.sum(self._K_dvar*partial) - target[1] += np.sum(self._K_dvar*self.variance*self._K_dist2/self.lengthscale*partial) - - def dKdiag_dtheta(self,partial,X,target): - #NB: derivative of diagonal elements wrt lengthscale is 0 - target[0] += np.sum(partial) - - def dK_dX(self,partial,X,X2,target): - self._K_computations(X,X2) - _K_dist = X[:,None,:]-X2[None,:,:] - dK_dX = np.transpose(-self.variance*self._K_dvar[:,:,np.newaxis]*_K_dist/self.lengthscale2,(1,0,2)) - target += np.sum(dK_dX*partial.T[:,:,None],0) - - def dKdiag_dX(self,partial,X,target): - pass - - def _K_computations(self,X,X2): - if not (np.all(X==self._X) and np.all(X2==self._X2)): - self._X = X - self._X2 = X2 - if X2 is None: X2 = X - XXT = np.dot(X,X2.T) - if X is X2: - self._K_dist2 = (-2.*XXT + np.diag(XXT)[:,np.newaxis] + np.diag(XXT)[np.newaxis,:])/self.lengthscale2 - else: - self._K_dist2 = (-2.*XXT + np.sum(np.square(X),1)[:,None] + np.sum(np.square(X2),1)[None,:])/self.lengthscale2 - # TODO Remove comments if this is fine. - # Commented out by Neil as doesn't seem to be used elsewhere. - #self._K_exponent = -0.5*self._K_dist2 - self._K_dvar = np.exp(-0.5*self._K_dist2) - - def psi0(self,Z,mu,S,target): - target += self.variance - - def dpsi0_dtheta(self,partial,Z,mu,S,target): - target[0] += 1. - - def dpsi0_dmuS(self,Z,mu,S,target_mu,target_S): - pass - - def psi1(self,Z,mu,S,target): - self._psi_computations(Z,mu,S) - target += self._psi1 - - def dpsi1_dtheta(self,partial,Z,mu,S,target): - self._psi_computations(Z,mu,S) - denom_deriv = S[:,None,:]/(self.lengthscale**3+self.lengthscale*S[:,None,:]) - d_length = self._psi1[:,:,None]*(self.lengthscale*np.square(self._psi1_dist/(self.lengthscale2+S[:,None,:])) + denom_deriv) - target[0] += np.sum(partial*self._psi1/self.variance) - target[1] += np.sum(d_length*partial[:,:,None]) - - def dpsi1_dZ(self,partial,Z,mu,S,target): - self._psi_computations(Z,mu,S) - target += np.sum(partial[:,:,None]*-self._psi1[:,:,None]*self._psi1_dist/self.lengthscale2/self._psi1_denom,0) - - def dpsi1_dmuS(self,partial,Z,mu,S,target_mu,target_S): - self._psi_computations(Z,mu,S) - tmp = self._psi1[:,:,None]/self.lengthscale2/self._psi1_denom - target_mu += np.sum(partial*tmp*self._psi1_dist,1) - target_S += np.sum(partial*0.5*tmp*(self._psi1_dist_sq-1),1) - - def psi2(self,Z,mu,S,target): - self._psi_computations(Z,mu,S) - target += self._psi2.sum(0) #TODO: psi2 should be NxMxM (for het. noise) - - def dpsi2_dtheta(self,partial,Z,mu,S,target): - """Shape N,M,M,Ntheta""" - self._psi_computations(Z,mu,S) - d_var = np.sum(2.*self._psi2/self.variance,0) - d_length = self._psi2[:,:,:,None]*(0.5*self._psi2_Zdist_sq*self._psi2_denom + 2.*self._psi2_mudist_sq + 2.*S[:,None,None,:]/self.lengthscale2)/(self.lengthscale*self._psi2_denom) - d_length = d_length.sum(0) - target[0] += np.sum(partial*d_var) - target[1] += np.sum(d_length*partial) - - def dpsi2_dZ(self,partial,Z,mu,S,target): - """Returns shape N,M,M,Q""" - self._psi_computations(Z,mu,S) - dZ = self._psi2[:,:,:,None]/self.lengthscale2*(-0.5*self._psi2_Zdist + self._psi2_mudist/self._psi2_denom) - target += np.sum(partial[None,:,:,None]*dZ,0).sum(1) - - def dpsi2_dmuS(self,Z,mu,S,target_mu,target_S): - """Think N,M,M,Q """ - self._psi_computations(Z,mu,S) - tmp = self._psi2[:,:,:,None]/self.lengthscale2/self._psi2_denom - target_mu += (partial*-tmp*2.*self._psi2_mudist).sum(1).sum(1) - target_S += (partial*tmp*(2.*self._psi2_mudist_sq-1)).sum(1).sum(1) - - def _psi_computations(self,Z,mu,S): - #here are the "statistics" for psi1 and psi2 - if not np.all(Z==self._Z): - #Z has changed, compute Z specific stuff - self._psi2_Zhat = 0.5*(Z[:,None,:] +Z[None,:,:]) # M,M,Q - self._psi2_Zdist = Z[:,None,:]-Z[None,:,:] # M,M,Q - self._psi2_Zdist_sq = np.square(self._psi2_Zdist)/self.lengthscale2 # M,M,Q - self._Z = Z - - if not (np.all(Z==self._Z) and np.all(mu==self._mu) and np.all(S==self._S)): - #something's changed. recompute EVERYTHING - - #TODO: make more efficient for large Q (using NDL's dot product trick) - #psi1 - self._psi1_denom = S[:,None,:]/self.lengthscale2 + 1. - self._psi1_dist = Z[None,:,:]-mu[:,None,:] - self._psi1_dist_sq = np.square(self._psi1_dist)/self.lengthscale2/self._psi1_denom - self._psi1_exponent = -0.5*np.sum(self._psi1_dist_sq+np.log(self._psi1_denom),-1) - self._psi1 = self.variance*np.exp(self._psi1_exponent) - - #psi2 - self._psi2_denom = 2.*S[:,None,None,:]/self.lengthscale2+1. # N,M,M,Q - self._psi2_mudist = mu[:,None,None,:]-self._psi2_Zhat #N,M,M,Q - self._psi2_mudist_sq = np.square(self._psi2_mudist)/(self.lengthscale2*self._psi2_denom) - self._psi2_exponent = np.sum(-self._psi2_Zdist_sq/4. -self._psi2_mudist_sq -0.5*np.log(self._psi2_denom),-1) #N,M,M - self._psi2 = np.square(self.variance)*np.exp(self._psi2_exponent) # N,M,M - - self._Z, self._mu, self._S = Z, mu,S - diff --git a/doc/index.rst b/doc/index.rst index fb721e15..818a71db 100644 --- a/doc/index.rst +++ b/doc/index.rst @@ -1,23 +1,17 @@ .. GPy documentation master file, created by - sphinx-quickstart on Fri Jan 18 15:30:28 2013. + sphinx-quickstart on Fri Jan 18 17:36:01 2013. You can adapt this file completely to your liking, but it should at least contain the root `toctree` directive. Welcome to GPy's documentation! =============================== -Added old command again -And another thing -Now what if I add this other hting -Just checking another thing - Contents: .. toctree:: :maxdepth: 4 GPy - setup Indices and tables From 6d98a230771a210a1144d1b7f6732655a8e1bb6a Mon Sep 17 00:00:00 2001 From: James Hensman Date: Fri, 18 Jan 2013 17:38:50 +0000 Subject: [PATCH 32/36] fixed optimize_restarts --- GPy/core/model.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/GPy/core/model.py b/GPy/core/model.py index 46cf6ac9..f21db4be 100644 --- a/GPy/core/model.py +++ b/GPy/core/model.py @@ -181,7 +181,7 @@ class model(parameterised): else: raise e if len(self.optimization_runs): - i = np.argmax([o.f_opt for o in self.optimization_runs]) + i = np.argmin([o.f_opt for o in self.optimization_runs]) self.expand_param(self.optimization_runs[i].x_opt) else: self.expand_param(initial_parameters) From fddbaeac77c55068f62593ea18ee0ef7ced87d18 Mon Sep 17 00:00:00 2001 From: Nicolas Date: Fri, 18 Jan 2013 17:55:39 +0000 Subject: [PATCH 33/36] delete unnecessary rbf_ARD.py --- GPy/kern/rbf_ARD.py | 251 -------------------------------------------- 1 file changed, 251 deletions(-) delete mode 100644 GPy/kern/rbf_ARD.py diff --git a/GPy/kern/rbf_ARD.py b/GPy/kern/rbf_ARD.py deleted file mode 100644 index 12924b3f..00000000 --- a/GPy/kern/rbf_ARD.py +++ /dev/null @@ -1,251 +0,0 @@ -# Copyright (c) 2012, GPy authors (see AUTHORS.txt). -# Licensed under the BSD 3-clause license (see LICENSE.txt) - - -from kernpart import kernpart -import numpy as np -import hashlib - -class rbf_ARD(kernpart): - def __init__(self,D,variance=1.,lengthscales=None): - """ - Arguments - ---------- - D: int - the number of input dimensions - variance: float - lengthscales : np.ndarray of shape (D,) - """ - self.D = D - if lengthscales is not None: - assert lengthscales.shape==(self.D,) - else: - lengthscales = np.ones(self.D) - self.Nparam = self.D + 1 - self.name = 'rbf_ARD' - self._set_params(np.hstack((variance,lengthscales))) - - #initialize cache - self._Z, self._mu, self._S = np.empty(shape=(3,1)) - self._X, self._X2, self._params = np.empty(shape=(3,1)) - - def _get_params(self): - return np.hstack((self.variance,self.lengthscales)) - - def _set_params(self,x): - assert x.size==(self.D+1) - self.variance = x[0] - self.lengthscales = x[1:] - self.lengthscales2 = np.square(self.lengthscales) - #reset cached results - self._Z, self._mu, self._S = np.empty(shape=(3,1)) # cached versions of Z,mu,S - - def _get_param_names(self): - if self.D==1: - return ['variance','lengthscale'] - else: - return ['variance']+['lengthscale_%i'%i for i in range(self.lengthscales.size)] - - def K(self,X,X2,target): - self._K_computations(X,X2) - np.add(self.variance*self._K_dvar, target,target) - - def Kdiag(self,X,target): - np.add(target,self.variance,target) - - def dK_dtheta(self,partial,X,X2,target): - self._K_computations(X,X2) - dl = self._K_dvar[:,:,None]*self.variance*self._K_dist2/self.lengthscales - target[0] += np.sum(self._K_dvar*partial) - target[1:] += (dl*partial[:,:,None]).sum(0).sum(0) - - def dKdiag_dtheta(self,X,target): - target[0] += np.sum(partial) - - def dK_dX(self,partial,X,X2,target): - self._K_computations(X,X2) - dZ = self.variance*self._K_dvar[:,:,None]*self._K_dist/self.lengthscales2 - dK_dX = -dZ.transpose(1,0,2) - target += np.sum(dK_dX*partial.T[:,:,None],0) - - def dKdiag_dX(self,partial,X,target): - pass - - def psi0(self,Z,mu,S,target): - target += self.variance - - def dpsi0_dtheta(self,partial,Z,mu,S,target): - target[0] += 1. - - def dpsi0_dmuS(self,Z,mu,S,target_mu,target_S): - pass - - def psi1(self,Z,mu,S,target): - self._psi_computations(Z,mu,S) - np.add(target, self._psi1,target) - - def dpsi1_dtheta(self,partial,Z,mu,S,target): - self._psi_computations(Z,mu,S) - denom_deriv = S[:,None,:]/(self.lengthscales**3+self.lengthscales*S[:,None,:]) - d_length = self._psi1[:,:,None]*(self.lengthscales*np.square(self._psi1_dist/(self.lengthscales2+S[:,None,:])) + denom_deriv) - target[0] += np.sum(partial*self._psi1/self.variance) - target[1:] += (d_length*partial[:,:,None]).sum(0).sum(0) - - def dpsi1_dZ(self,partial,Z,mu,S,target): - self._psi_computations(Z,mu,S) - np.add(target,-self._psi1[:,:,None]*self._psi1_dist/self.lengthscales2/self._psi1_denom,target) - target += np.sum(partial[:,:,None]*-self._psi1[:,:,None]*self._psi1_dist/self.lengthscales2/self._psi1_denom,0) - - def dpsi1_dmuS(self,partial,Z,mu,S,target_mu,target_S): - """return shapes are N,M,Q""" - self._psi_computations(Z,mu,S) - tmp = self._psi1[:,:,None]/self.lengthscales2/self._psi1_denom - target_mu += np.sum(partial*tmp*self._psi1_dist,1) - target_S += np.sum(partial*0.5*tmp*(self._psi1_dist_sq-1),1) - - def psi2(self,Z,mu,S,target): - self._psi_computations(Z,mu,S) - target += self._psi2.sum(0) #TODO: psi2 should be NxMxM (for het. noise) - - def dpsi2_dtheta(self,Z,mu,S,target): - """Shape N,M,M,Ntheta""" - self._psi_computations(Z,mu,S) - d_var = np.sum(2.*self._psi2/self.variance,0) - d_length = self._psi2[:,:,:,None]*(0.5*self._psi2_Zdist_sq*self._psi2_denom + 2.*self._psi2_mudist_sq + 2.*S[:,None,None,:]/self.lengthscales2)/(self.lengthscales*self._psi2_denom) - d_length = d_length.sum(0) - target[0] += np.sum(partial*d_var) - target[1:] += (d_length*partial[:,:,None]).sum(0).sum(0) - - def dpsi2_dZ(self,Z,mu,S,target): - """Returns shape N,M,M,Q""" - self._psi_computations(Z,mu,S) - dZ = self._psi2[:,:,:,None]/self.lengthscales2*(-0.5*self._psi2_Zdist + self._psi2_mudist/self._psi2_denom) - target += np.sum(partial[None,:,:,None]*dZ,0).sum(1) - - def dpsi2_dmuS(self,Z,mu,S,target_mu,target_S): - """Think N,M,M,Q """ - self._psi_computations(Z,mu,S) - tmp = self._psi2[:,:,:,None]/self.lengthscales2/self._psi2_denom - target_mu += (partial*-tmp*2.*self._psi2_mudist).sum(1).sum(1) - target_S += (partial*tmp*(2.*self._psi2_mudist_sq-1)).sum(1).sum(1) - - def _K_computations(self,X,X2): - if not (np.all(X==self._X) and np.all(X2==self._X2)): - self._X = X - self._X2 = X2 - if X2 is None: X2 = X - self._K_dist = X[:,None,:]-X2[None,:,:] # this can be computationally heavy - self._params = np.empty(shape=(1,0))#ensure the next section gets called - if not np.all(self._params == self._get_params()): - self._params == self._get_params() - self._K_dist2 = np.square(self._K_dist/self.lengthscales) - self._K_exponent = -0.5*self._K_dist2.sum(-1) - self._K_dvar = np.exp(-0.5*self._K_dist2.sum(-1)) - - def _psi_computations(self,Z,mu,S): - #here are the "statistics" for psi1 and psi2 - if not np.all(Z==self._Z): - #Z has changed, compute Z specific stuff - self._psi2_Zhat = 0.5*(Z[:,None,:] +Z[None,:,:]) # M,M,Q - self._psi2_Zdist = Z[:,None,:]-Z[None,:,:] # M,M,Q - self._psi2_Zdist_sq = np.square(self._psi2_Zdist)/self.lengthscales2 # M,M,Q - self._Z = Z - - if not (np.all(Z==self._Z) and np.all(mu==self._mu) and np.all(S==self._S)): - #something's changed. recompute EVERYTHING - - #psi1 - self._psi1_denom = S[:,None,:]/self.lengthscales2 + 1. - self._psi1_dist = Z[None,:,:]-mu[:,None,:] - self._psi1_dist_sq = np.square(self._psi1_dist)/self.lengthscales2/self._psi1_denom - self._psi1_exponent = -0.5*np.sum(self._psi1_dist_sq+np.log(self._psi1_denom),-1) - self._psi1 = self.variance*np.exp(self._psi1_exponent) - - #psi2 - self._psi2_denom = 2.*S[:,None,None,:]/self.lengthscales2+1. # N,M,M,Q - self._psi2_mudist = mu[:,None,None,:]-self._psi2_Zhat #N,M,M,Q - self._psi2_mudist_sq = np.square(self._psi2_mudist)/(self.lengthscales2*self._psi2_denom) - self._psi2_exponent = np.sum(-self._psi2_Zdist_sq/4. -self._psi2_mudist_sq -0.5*np.log(self._psi2_denom),-1) #N,M,M - self._psi2 = np.square(self.variance)*np.exp(self._psi2_exponent) # N,M,M - - self._Z, self._mu, self._S = Z, mu,S - - -if __name__=='__main__': - #run some simple tests on the kernel (TODO:move these to unititest) - #TODO: these are broken in this new structure! - N = 10 - M = 5 - Q = 3 - - Z = np.random.randn(M,Q) - mu = np.random.randn(N,Q) - S = np.random.rand(N,Q) - - var = 2.5 - lengthscales = np.ones(Q)*0.7 - - k = rbf(Q,var,lengthscales) - - from checkgrad import checkgrad - - def k_theta_test(param,k): - k._set_params(param) - K = k.K(Z) - dK_dtheta = k.dK_dtheta(Z) - f = np.sum(K) - df = dK_dtheta.sum(0).sum(0) - return f,np.array(df) - print "dk_dtheta_test" - checkgrad(k_theta_test,np.random.randn(1+Q),args=(k,)) - - - def psi1_mu_test(mu,k): - mu = mu.reshape(N,Q) - f = np.sum(k.psi1(Z,mu,S)) - df = k.dpsi1_dmuS(Z,mu,S)[0].sum(1) - return f,df.flatten() - print "psi1_mu_test" - checkgrad(psi1_mu_test,np.random.randn(N*Q),args=(k,)) - - def psi1_S_test(S,k): - S = S.reshape(N,Q) - f = np.sum(k.psi1(Z,mu,S)) - df = k.dpsi1_dmuS(Z,mu,S)[1].sum(1) - return f,df.flatten() - print "psi1_S_test" - checkgrad(psi1_S_test,np.random.rand(N*Q),args=(k,)) - - def psi1_theta_test(theta,k): - k._set_params(theta) - f = np.sum(k.psi1(Z,mu,S)) - df = np.array([np.sum(grad) for grad in k.dpsi1_dtheta(Z,mu,S)]) - return f,df - print "psi1_theta_test" - checkgrad(psi1_theta_test,np.random.rand(1+Q),args=(k,)) - - - def psi2_mu_test(mu,k): - mu = mu.reshape(N,Q) - f = np.sum(k.psi2(Z,mu,S)) - df = k.dpsi2_dmuS(Z,mu,S)[0].sum(1).sum(1) - return f,df.flatten() - print "psi2_mu_test" - checkgrad(psi2_mu_test,np.random.randn(N*Q),args=(k,)) - - def psi2_S_test(S,k): - S = S.reshape(N,Q) - f = np.sum(k.psi2(Z,mu,S)) - df = k.dpsi2_dmuS(Z,mu,S)[1].sum(1).sum(1) - return f,df.flatten() - print "psi2_S_test" - checkgrad(psi2_S_test,np.random.rand(N*Q),args=(k,)) - - def psi2_theta_test(theta,k): - k._set_params(theta) - f = np.sum(k.psi2(Z,mu,S)) - df = np.array([np.sum(grad) for grad in k.dpsi2_dtheta(Z,mu,S)]) - return f,df - print "psi2_theta_test" - checkgrad(psi2_theta_test,np.random.rand(1+Q),args=(k,)) - - From a908664996713ff17b6f7014bc6043553f24ea73 Mon Sep 17 00:00:00 2001 From: Alan Saul Date: Fri, 18 Jan 2013 18:10:03 +0000 Subject: [PATCH 34/36] Added path for RTD --- doc/conf.py | 14 ++++++++------ 1 file changed, 8 insertions(+), 6 deletions(-) diff --git a/doc/conf.py b/doc/conf.py index cf9f7ca7..474836a2 100644 --- a/doc/conf.py +++ b/doc/conf.py @@ -27,6 +27,14 @@ import sys, os # coming with Sphinx (named 'sphinx.ext.*') or your custom ones. extensions = ['sphinx.ext.autodoc', 'sphinx.ext.viewcode'] +# ----------------------- READTHEDOCS ------------------ +on_rtd = os.environ.get('READTHEDOCS', None) == 'True' + +if on_rtd: + sys.path.append("../GPy") + os.system("pwd") + os.system("sphinx-apidoc -f -o . ../GPy") + # Add any paths that contain templates here, relative to this directory. templates_path = ['_templates'] @@ -284,9 +292,3 @@ epub_copyright = u'2013, Author' # Allow duplicate toc entries. #epub_tocdup = True -# ----------------------- READTHEDOCS ------------------ -on_rtd = os.environ.get('READTHEDOCS', None) == 'True' - -if on_rtd: - os.system("pwd") - os.system("sphinx-apidoc -f -o . ../GPy") From e8383c2d71ddfa91f52f805d266e251868a57f23 Mon Sep 17 00:00:00 2001 From: Nicolas Date: Fri, 18 Jan 2013 18:15:36 +0000 Subject: [PATCH 35/36] restored examples folder --- GPy/__init__.py | 2 +- GPy/kern/constructors.py | 1 - 2 files changed, 1 insertion(+), 2 deletions(-) diff --git a/GPy/__init__.py b/GPy/__init__.py index 0afa92e4..381d6232 100644 --- a/GPy/__init__.py +++ b/GPy/__init__.py @@ -6,6 +6,6 @@ import kern import models import inference import util -#import examples +import examples #import examples TODO: discuss! from core import priors diff --git a/GPy/kern/constructors.py b/GPy/kern/constructors.py index 883357aa..ab39fcb2 100644 --- a/GPy/kern/constructors.py +++ b/GPy/kern/constructors.py @@ -6,7 +6,6 @@ import numpy as np from kern import kern from rbf import rbf as rbfpart -from rbf_ARD import rbf_ARD as rbf_ARD_part from white import white as whitepart from linear import linear as linearpart from linear_ARD import linear_ARD as linear_ARD_part From eedf1f99d1f94967a82efb559992ff0d265aeb4c Mon Sep 17 00:00:00 2001 From: Nicolas Date: Mon, 21 Jan 2013 12:16:55 +0000 Subject: [PATCH 36/36] added some documentation and example files --- GPy/examples/__init__.py | 0 doc/GPy.core.rst | 35 ++++++++++ doc/GPy.inference.rst | 35 ++++++++++ doc/GPy.kern.rst | 139 +++++++++++++++++++++++++++++++++++++++ doc/GPy.models.rst | 75 +++++++++++++++++++++ doc/GPy.rst | 22 +++++++ doc/GPy.util.rst | 67 +++++++++++++++++++ doc/modules.rst | 7 ++ 8 files changed, 380 insertions(+) create mode 100644 GPy/examples/__init__.py create mode 100644 doc/GPy.core.rst create mode 100644 doc/GPy.inference.rst create mode 100644 doc/GPy.kern.rst create mode 100644 doc/GPy.models.rst create mode 100644 doc/GPy.rst create mode 100644 doc/GPy.util.rst create mode 100644 doc/modules.rst diff --git a/GPy/examples/__init__.py b/GPy/examples/__init__.py new file mode 100644 index 00000000..e69de29b diff --git a/doc/GPy.core.rst b/doc/GPy.core.rst new file mode 100644 index 00000000..e02aaa2a --- /dev/null +++ b/doc/GPy.core.rst @@ -0,0 +1,35 @@ +core Package +============ + +:mod:`core` Package +------------------- + +.. automodule:: GPy.core + :members: + :undoc-members: + :show-inheritance: + +:mod:`model` Module +------------------- + +.. automodule:: GPy.core.model + :members: + :undoc-members: + :show-inheritance: + +:mod:`parameterised` Module +--------------------------- + +.. automodule:: GPy.core.parameterised + :members: + :undoc-members: + :show-inheritance: + +:mod:`priors` Module +-------------------- + +.. automodule:: GPy.core.priors + :members: + :undoc-members: + :show-inheritance: + diff --git a/doc/GPy.inference.rst b/doc/GPy.inference.rst new file mode 100644 index 00000000..6f4ab691 --- /dev/null +++ b/doc/GPy.inference.rst @@ -0,0 +1,35 @@ +inference Package +================= + +:mod:`Expectation_Propagation` Module +------------------------------------- + +.. automodule:: GPy.inference.Expectation_Propagation + :members: + :undoc-members: + :show-inheritance: + +:mod:`likelihoods` Module +------------------------- + +.. automodule:: GPy.inference.likelihoods + :members: + :undoc-members: + :show-inheritance: + +:mod:`optimization` Module +-------------------------- + +.. automodule:: GPy.inference.optimization + :members: + :undoc-members: + :show-inheritance: + +:mod:`samplers` Module +---------------------- + +.. automodule:: GPy.inference.samplers + :members: + :undoc-members: + :show-inheritance: + diff --git a/doc/GPy.kern.rst b/doc/GPy.kern.rst new file mode 100644 index 00000000..95943178 --- /dev/null +++ b/doc/GPy.kern.rst @@ -0,0 +1,139 @@ +kern Package +============ + +:mod:`kern` Package +------------------- + +.. automodule:: GPy.kern + :members: + :undoc-members: + :show-inheritance: + +:mod:`Brownian` Module +---------------------- + +.. automodule:: GPy.kern.Brownian + :members: + :undoc-members: + :show-inheritance: + +:mod:`Matern32` Module +---------------------- + +.. automodule:: GPy.kern.Matern32 + :members: + :undoc-members: + :show-inheritance: + +:mod:`Matern52` Module +---------------------- + +.. automodule:: GPy.kern.Matern52 + :members: + :undoc-members: + :show-inheritance: + +:mod:`bias` Module +------------------ + +.. automodule:: GPy.kern.bias + :members: + :undoc-members: + :show-inheritance: + +:mod:`constructors` Module +-------------------------- + +.. automodule:: GPy.kern.constructors + :members: + :undoc-members: + :show-inheritance: + +:mod:`exponential` Module +------------------------- + +.. automodule:: GPy.kern.exponential + :members: + :undoc-members: + :show-inheritance: + +:mod:`finite_dimensional` Module +-------------------------------- + +.. automodule:: GPy.kern.finite_dimensional + :members: + :undoc-members: + :show-inheritance: + +:mod:`kern` Module +------------------ + +.. automodule:: GPy.kern.kern + :members: + :undoc-members: + :show-inheritance: + +:mod:`kernpart` Module +---------------------- + +.. automodule:: GPy.kern.kernpart + :members: + :undoc-members: + :show-inheritance: + +:mod:`linear` Module +-------------------- + +.. automodule:: GPy.kern.linear + :members: + :undoc-members: + :show-inheritance: + +:mod:`linear_ARD` Module +------------------------ + +.. automodule:: GPy.kern.linear_ARD + :members: + :undoc-members: + :show-inheritance: + +:mod:`rbf-testing` Module +------------------------- + +.. automodule:: GPy.kern.rbf-testing + :members: + :undoc-members: + :show-inheritance: + +:mod:`rbf` Module +----------------- + +.. automodule:: GPy.kern.rbf + :members: + :undoc-members: + :show-inheritance: + +:mod:`spline` Module +-------------------- + +.. automodule:: GPy.kern.spline + :members: + :undoc-members: + :show-inheritance: + +:mod:`sympykern` Module +----------------------- + +.. automodule:: GPy.kern.sympykern + :members: + :undoc-members: + :show-inheritance: + +:mod:`white` Module +------------------- + +.. automodule:: GPy.kern.white + :members: + :undoc-members: + :show-inheritance: + diff --git a/doc/GPy.models.rst b/doc/GPy.models.rst new file mode 100644 index 00000000..47af78ab --- /dev/null +++ b/doc/GPy.models.rst @@ -0,0 +1,75 @@ +models Package +============== + +:mod:`models` Package +--------------------- + +.. automodule:: GPy.models + :members: + :undoc-members: + :show-inheritance: + +:mod:`GPLVM` Module +------------------- + +.. automodule:: GPy.models.GPLVM + :members: + :undoc-members: + :show-inheritance: + +:mod:`GP_EP` Module +------------------- + +.. automodule:: GPy.models.GP_EP + :members: + :undoc-members: + :show-inheritance: + +:mod:`GP_regression` Module +--------------------------- + +.. automodule:: GPy.models.GP_regression + :members: + :undoc-members: + :show-inheritance: + +:mod:`generalized_FITC` Module +------------------------------ + +.. automodule:: GPy.models.generalized_FITC + :members: + :undoc-members: + :show-inheritance: + +:mod:`sparse_GPLVM` Module +-------------------------- + +.. automodule:: GPy.models.sparse_GPLVM + :members: + :undoc-members: + :show-inheritance: + +:mod:`sparse_GP_regression` Module +---------------------------------- + +.. automodule:: GPy.models.sparse_GP_regression + :members: + :undoc-members: + :show-inheritance: + +:mod:`uncollapsed_sparse_GP` Module +----------------------------------- + +.. automodule:: GPy.models.uncollapsed_sparse_GP + :members: + :undoc-members: + :show-inheritance: + +:mod:`warped_GP` Module +----------------------- + +.. automodule:: GPy.models.warped_GP + :members: + :undoc-members: + :show-inheritance: + diff --git a/doc/GPy.rst b/doc/GPy.rst new file mode 100644 index 00000000..61a4242c --- /dev/null +++ b/doc/GPy.rst @@ -0,0 +1,22 @@ +GPy Package +=========== + +:mod:`GPy` Package +------------------ + +.. automodule:: GPy.__init__ + :members: + :undoc-members: + :show-inheritance: + +Subpackages +----------- + +.. toctree:: + + GPy.core + GPy.inference + GPy.kern + GPy.models + GPy.util + diff --git a/doc/GPy.util.rst b/doc/GPy.util.rst new file mode 100644 index 00000000..5bec990b --- /dev/null +++ b/doc/GPy.util.rst @@ -0,0 +1,67 @@ +util Package +============ + +:mod:`util` Package +------------------- + +.. automodule:: GPy.util + :members: + :undoc-members: + :show-inheritance: + +:mod:`Tango` Module +------------------- + +.. automodule:: GPy.util.Tango + :members: + :undoc-members: + :show-inheritance: + +:mod:`datasets` Module +---------------------- + +.. automodule:: GPy.util.datasets + :members: + :undoc-members: + :show-inheritance: + +:mod:`linalg` Module +-------------------- + +.. automodule:: GPy.util.linalg + :members: + :undoc-members: + :show-inheritance: + +:mod:`misc` Module +------------------ + +.. automodule:: GPy.util.misc + :members: + :undoc-members: + :show-inheritance: + +:mod:`plot` Module +------------------ + +.. automodule:: GPy.util.plot + :members: + :undoc-members: + :show-inheritance: + +:mod:`squashers` Module +----------------------- + +.. automodule:: GPy.util.squashers + :members: + :undoc-members: + :show-inheritance: + +:mod:`warping_functions` Module +------------------------------- + +.. automodule:: GPy.util.warping_functions + :members: + :undoc-members: + :show-inheritance: + diff --git a/doc/modules.rst b/doc/modules.rst new file mode 100644 index 00000000..9c698ca2 --- /dev/null +++ b/doc/modules.rst @@ -0,0 +1,7 @@ +GPy +=== + +.. toctree:: + :maxdepth: 4 + + GPy