Merge branch 'devel' of github.com:SheffieldML/GPy into devel

GPy/core/gp.py

@@ -96,7 +96,8 @@ def log_likelihood(self):
         model for a new variable Y* = v_tilde/tau_tilde, with a covariance
         matrix K* = K + diag(1./tau_tilde) plus a normalization term.
         """
-        return - 0.5 * self.num_data * self.output_dim * np.log(2.*np.pi) - 0.5 * self.output_dim * self.K_logdet + self._model_fit_term() + self.likelihood.Z
+        return (-0.5 * self.num_data * self.output_dim * np.log(2.*np.pi) -
+            0.5 * self.output_dim * self.K_logdet + self._model_fit_term() + self.likelihood.Z)
 
 
     def _log_likelihood_gradients(self):

GPy/core/sparse_gp.py

@@ -108,7 +108,7 @@ def _computations(self):
         self.B = np.eye(self.num_inducing) + self.A
         self.LB = jitchol(self.B)
 
-        #VVT_factor is a matrix such that tdot(VVT_factor) = VVT...this is for efficiency!
+        # VVT_factor is a matrix such that tdot(VVT_factor) = VVT...this is for efficiency!
         self.psi1Vf = np.dot(self.psi1.T, self.likelihood.VVT_factor)
 
         # back substutue C into psi1Vf
@@ -163,7 +163,7 @@ def _computations(self):
     def log_likelihood(self):
         """ Compute the (lower bound on the) log marginal likelihood """
         if self.likelihood.is_heteroscedastic:
-            A = -0.5 * self.num_data * self.output_dim * np.log(2.*np.pi) + 0.5 * np.sum(np.log(self.likelihood.precision)) - 0.5 * np.sum(self.likelihood.V*self.likelihood.Y)
+            A = -0.5 * self.num_data * self.output_dim * np.log(2.*np.pi) + 0.5 * np.sum(np.log(self.likelihood.precision)) - 0.5 * np.sum(self.likelihood.V * self.likelihood.Y)
             B = -0.5 * self.output_dim * (np.sum(self.likelihood.precision.flatten() * self.psi0) - np.trace(self.A))
         else:
             A = -0.5 * self.num_data * self.output_dim * (np.log(2.*np.pi) - np.log(self.likelihood.precision)) - 0.5 * self.likelihood.precision * self.likelihood.trYYT
@@ -246,7 +246,7 @@ def _raw_predict(self, Xnew, X_variance_new=None, which_parts='all', full_cov=Fa
         Kmmi_LmiBLmi = backsub_both_sides(self.Lm, np.eye(self.num_inducing) - Bi)
 
         if self.Cpsi1V is None:
-            psi1V = np.dot(self.psi1.T,self.likelihood.V)
+            psi1V = np.dot(self.psi1.T, self.likelihood.V)
             tmp, _ = dtrtrs(self.Lm, np.asfortranarray(psi1V), lower=1, trans=0)
             tmp, _ = dpotrs(self.LB, tmp, lower=1)
             self.Cpsi1V, _ = dtrtrs(self.Lm, tmp, lower=1, trans=1)

GPy/examples/dimensionality_reduction.py

@@ -7,26 +7,29 @@
 import GPy
 from GPy.core.transformations import logexp
 from GPy.models.bayesian_gplvm import BayesianGPLVM
+from GPy.likelihoods.gaussian import Gaussian
 
 default_seed = np.random.seed(123344)
 
 def BGPLVM(seed=default_seed):
     N = 10
     num_inducing = 3
-    Q = 2
-    D = 4
+    Q = 5
+    D = 10
     # generate GPLVM-like data
     X = np.random.rand(N, Q)
-    k = GPy.kern.rbf(Q) + GPy.kern.white(Q, 0.00001)
+    lengthscales = np.random.rand(Q)
+    k = GPy.kern.rbf(Q, .5, lengthscales, ARD=True) + GPy.kern.white(Q, 0.01)
     K = k.K(X)
     Y = np.random.multivariate_normal(np.zeros(N), K, Q).T
+    lik = Gaussian(Y, normalize=True)
 
-    k = GPy.kern.rbf(Q, ARD=True) + GPy.kern.linear(Q, ARD=True) + GPy.kern.rbf(Q, ARD=True) + GPy.kern.white(Q)
-    # k = GPy.kern.rbf(Q) + GPy.kern.rbf(Q) + GPy.kern.white(Q)
+    k = GPy.kern.rbf_inv(Q, ARD=True) + GPy.kern.bias(Q) + GPy.kern.white(Q)
     # k = GPy.kern.rbf(Q) + GPy.kern.bias(Q) + GPy.kern.white(Q, 0.00001)
     # k = GPy.kern.rbf(Q, ARD = False)  + GPy.kern.white(Q, 0.00001)
 
-    m = GPy.models.BayesianGPLVM(Y, Q, kernel=k, num_inducing=num_inducing)
+    m = GPy.models.BayesianGPLVM(lik, Q, kernel=k, num_inducing=num_inducing)
+    m.lengthscales = lengthscales
     # m.constrain_positive('(rbf|bias|noise|white|S)')
     # m.constrain_fixed('S', 1)
 
@@ -37,8 +40,8 @@ def BGPLVM(seed=default_seed):
     # m.optimize(messages = 1)
     # m.plot()
     # pb.title('After optimisation')
-    m.randomize()
-    m.checkgrad(verbose=1)
+    # m.randomize()
+    # m.checkgrad(verbose=1)
 
     return m
 
@@ -70,16 +73,16 @@ def sparseGPLVM_oil(optimize=True, N=100, Q=6, num_inducing=15, max_iters=50):
 
     # create simple GP model
     kernel = GPy.kern.rbf(Q, ARD=True) + GPy.kern.bias(Q)
-    m = GPy.models.SparseGPLVM(Y, Q, kernel=kernel, num_inducing = num_inducing)
+    m = GPy.models.SparseGPLVM(Y, Q, kernel=kernel, num_inducing=num_inducing)
     m.data_labels = data['Y'].argmax(axis=1)
 
     # optimize
     if optimize:
-        m.optimize('scg', messages=1, max_iters = max_iters)
+        m.optimize('scg', messages=1, max_iters=max_iters)
 
     # plot
     print(m)
-    #m.plot_latent(labels=m.data_labels)
+    # m.plot_latent(labels=m.data_labels)
     return m
 
 def swiss_roll(optimize=True, N=1000, num_inducing=15, Q=4, sigma=.2, plot=False):
@@ -136,12 +139,13 @@ def swiss_roll(optimize=True, N=1000, num_inducing=15, Q=4, sigma=.2, plot=False
         m.optimize('scg', messages=1)
     return m
 
-def BGPLVM_oil(optimize=True, N=200, Q=10, num_inducing=15, max_iters=50, plot=False, **k):
+def BGPLVM_oil(optimize=True, N=200, Q=10, num_inducing=15, max_iters=150, plot=False, **k):
     np.random.seed(0)
     data = GPy.util.datasets.oil()
 
     # create simple GP model
-    kernel = GPy.kern.rbf(Q, ARD=True) + GPy.kern.bias(Q, np.exp(-2)) + GPy.kern.white(Q, np.exp(-2))
+    kernel = GPy.kern.rbf_inv(Q, ARD=True) + GPy.kern.bias(Q, np.exp(-2)) + GPy.kern.white(Q, np.exp(-2))
+    kernel += GPy.kern.rbf_inv(Q, ARD=True) + GPy.kern.bias(Q, np.exp(-2)) + GPy.kern.white(Q, np.exp(-2))
     Y = data['X'][:N]
     Yn = Y - Y.mean(0)
     Yn /= Yn.std(0)
@@ -150,14 +154,14 @@ def BGPLVM_oil(optimize=True, N=200, Q=10, num_inducing=15, max_iters=50, plot=F
     m.data_labels = data['Y'][:N].argmax(axis=1)
 
     # m.constrain('variance|leng', logexp_clipped())
-    m['.*lengt'] = 1. # m.X.var(0).max() / m.X.var(0)
+    # m['.*lengt'] = m.X.var(0).max() / m.X.var(0)
     m['noise'] = Yn.var() / 100.
 
 
     # optimize
     if optimize:
         m.constrain_fixed('noise')
-        m.optimize('scg', messages=1, max_iters=100, gtol=.05)
+        m.optimize('scg', messages=1, max_iters=200, gtol=.05)
         m.constrain_positive('noise')
         m.optimize('scg', messages=1, max_iters=max_iters, gtol=.05)
 
@@ -208,7 +212,7 @@ def _simulate_sincos(D1, D2, D3, N, num_inducing, Q, plot_sim=False):
 
     Y1 += .3 * np.random.randn(*Y1.shape)
     Y2 += .2 * np.random.randn(*Y2.shape)
-    Y3 += .1 * np.random.randn(*Y3.shape)
+    Y3 += .25 * np.random.randn(*Y3.shape)
 
     Y1 -= Y1.mean(0)
     Y2 -= Y2.mean(0)
@@ -263,24 +267,23 @@ def bgplvm_simulation_matlab_compare():
 
 def bgplvm_simulation(optimize='scg',
                       plot=True,
-                      max_iters=2e4):
+                      max_iters=2e4,
+                      plot_sim=False):
 #     from GPy.core.transformations import logexp_clipped
-    D1, D2, D3, N, num_inducing, Q = 15, 8, 8, 100, 3, 5
-    slist, Slist, Ylist = _simulate_sincos(D1, D2, D3, N, num_inducing, Q, plot)
+    D1, D2, D3, N, num_inducing, Q = 15, 5, 8, 300, 30, 6
+    slist, Slist, Ylist = _simulate_sincos(D1, D2, D3, N, num_inducing, Q, plot_sim)
 
     from GPy.models import mrd
     from GPy import kern
     reload(mrd); reload(kern)
 
-
     Y = Ylist[0]
 
     k = kern.linear(Q, ARD=True) + kern.bias(Q, np.exp(-2)) + kern.white(Q, np.exp(-2)) # + kern.bias(Q)
     m = BayesianGPLVM(Y, Q, init="PCA", num_inducing=num_inducing, kernel=k)
 
     # m.constrain('variance|noise', logexp_clipped())
     m['noise'] = Y.var() / 100.
-    m['linear_variance'] = .01
 
     if optimize:
         print "Optimizing model:"
@@ -346,7 +349,7 @@ def brendan_faces():
 def stick_play(range=None, frame_rate=15):
     data = GPy.util.datasets.stick()
     # optimize
-    if range==None:
+    if range == None:
         Y = data['Y'].copy()
     else:
         Y = data['Y'][range[0]:range[1], :].copy()
@@ -374,10 +377,10 @@ def stick_bgplvm(model=None):
     data = GPy.util.datasets.stick()
     Q = 6
     kernel = GPy.kern.rbf(Q, ARD=True) + GPy.kern.bias(Q, np.exp(-2)) + GPy.kern.white(Q, np.exp(-2))
-    m = BayesianGPLVM(data['Y'], Q, init="PCA", num_inducing=20,kernel=kernel)
+    m = BayesianGPLVM(data['Y'], Q, init="PCA", num_inducing=20, kernel=kernel)
     # optimize
     m.ensure_default_constraints()
-    m.optimize(messages=1, max_f_eval=3000,xtol=1e-300,ftol=1e-300)
+    m.optimize(messages=1, max_f_eval=3000, xtol=1e-300, ftol=1e-300)
     m._set_params(m._get_params())
     plt.clf, (latent_axes, sense_axes) = plt.subplots(1, 2)
     plt.sca(latent_axes)

GPy/kern/kern.py

@@ -66,12 +66,26 @@ def setstate(self, state):
         Parameterized.setstate(self, state)
 
 
-    def plot_ARD(self, fignum=None, ax=None, title=None):
-        """If an ARD kernel is present, it bar-plots the ARD parameters"""
+    def plot_ARD(self, fignum=None, ax=None, title='', legend=False):
+        """If an ARD kernel is present, it bar-plots the ARD parameters,
+        :param fignum: figure number of the plot
+        :param ax: matplotlib axis to plot on
+        :param title: 
+            title of the plot, 
+            pass '' to not print a title
+            pass None for a generic title
+        """
         if ax is None:
             fig = pb.figure(fignum)
             ax = fig.add_subplot(111)
+        from GPy.util import Tango
+        from matplotlib.textpath import TextPath
+        Tango.reset()
+        xticklabels = []
+        bars = []
+        x0 = 0
         for p in self.parts:
+            c = Tango.nextMedium()
             if hasattr(p, 'ARD') and p.ARD:
                 if title is None:
                     ax.set_title('ARD parameters, %s kernel' % p.name)
@@ -82,10 +96,32 @@ def plot_ARD(self, fignum=None, ax=None, title=None):
                 else:
                     ard_params = 1. / p.lengthscale
 
-                x = np.arange(len(ard_params))
-                ax.bar(x - 0.4, ard_params)
-                ax.set_xticks(x)
-                ax.set_xticklabels([r"${}$".format(i) for i in x])
+                x = np.arange(x0, x0 + len(ard_params))
+                bars.append(ax.bar(x, ard_params, align='center', color=c, edgecolor='k', linewidth=1.2, label=p.name))
+                xticklabels.extend([r"$\mathrm{{{name}}}\ {x}$".format(name=p.name, x=i) for i in np.arange(len(ard_params))])
+                x0 += len(ard_params)
+        x = np.arange(x0)
+        for bar in bars:
+            for patch, num in zip(bar.patches, np.arange(len(bar.patches))):
+                height = patch.get_height()
+                xi = patch.get_x() + patch.get_width() / 2.
+                va = 'top'
+                c = 'w'
+                t = TextPath((0, 0), "${xi}$".format(xi=xi), rotation=0, usetex=True, ha='center')
+                if patch.get_extents().height <= t.get_extents().height + 2:
+                    va = 'bottom'
+                    c = 'k'
+                ax.text(xi, height, "${xi}$".format(xi=int(num)), color=c, rotation=0, ha='center', va=va)
+        # for xi, t in zip(x, xticklabels):
+        #    ax.text(xi, maxi / 2, t, rotation=90, ha='center', va='center')
+        # ax.set_xticklabels(xticklabels, rotation=17)
+        ax.set_xticks([])
+        ax.set_xlim(-.5, x0 - .5)
+        if title is '':
+            ax.legend(bbox_to_anchor=(0., 1.02, 1., .102), loc=3,
+                      ncol=len(bars), mode="expand", borderaxespad=0.)
+        else:
+            ax.legend()
         return ax
 
     def _transform_gradients(self, g):
@@ -361,6 +397,7 @@ def dpsi2_dtheta(self, dL_dpsi2, Z, mu, S):
 
         # compute the "cross" terms
         # TODO: better looping, input_slices
+
         for i1, i2 in itertools.permutations(range(len(self.parts)), 2):
             p1, p2 = self.parts[i1], self.parts[i2]
 #             ipsl1, ipsl2 = self.input_slices[i1], self.input_slices[i2]

GPy/models/bayesian_gplvm.py

@@ -115,8 +115,8 @@ def _log_likelihood_gradients(self):
         self.dbound_dZtheta = SparseGP._log_likelihood_gradients(self)
         return np.hstack((self.dbound_dmuS.flatten(), self.dbound_dZtheta))
 
-    def plot_latent(self, *args, **kwargs):
-        return plot_latent.plot_latent(self, *args, **kwargs)
+    def plot_latent(self, plot_inducing=True, *args, **kwargs):
+        return plot_latent.plot_latent(self, plot_inducing=plot_inducing, *args, **kwargs)
 
     def do_test_latents(self, Y):
         """

GPy/models/gplvm.py

@@ -36,10 +36,10 @@ def __init__(self, Y, input_dim, init='PCA', X=None, kernel=None, normalize_Y=Fa
         self.ensure_default_constraints()
 
     def initialise_latent(self, init, input_dim, Y):
+        Xr = np.random.randn(Y.shape[0], input_dim)
         if init == 'PCA':
-            return PCA(Y, input_dim)[0]
-        else:
-            return np.random.randn(Y.shape[0], input_dim)
+            Xr[:, :Y.shape[1]] = PCA(Y, input_dim)[0]
+        return Xr
 
     def getstate(self):
         return GP.getstate(self)

GPy/testing/unit_tests.py

@@ -23,7 +23,7 @@ def setUp(self):
         self.X2D = np.random.uniform(-3., 3., (40, 2))
         self.Y2D = np.sin(self.X2D[:, 0:1]) * np.sin(self.X2D[:, 1:2]) + np.random.randn(40, 1) * 0.05
 
-    def check_model_with_white(self, kern, model_type='GPRegression', dimension=1):
+    def check_model_with_white(self, kern, model_type='GPRegression', dimension=1, uncertain_inputs=False):
         # Get the correct gradients
         if dimension == 1:
             X = self.X1D
@@ -36,7 +36,10 @@ def check_model_with_white(self, kern, model_type='GPRegression', dimension=1):
 
         noise = GPy.kern.white(dimension)
         kern = kern + noise
-        m = model_fit(X, Y, kernel=kern)
+        if uncertain_inputs:
+            m = model_fit(X, Y, kernel=kern, X_variance=np.random.rand(X.shape[0], X.shape[1]))
+        else:
+            m = model_fit(X, Y, kernel=kern)
         m.randomize()
         # contrain all parameters to be positive
         self.assertTrue(m.checkgrad())
@@ -141,6 +144,26 @@ def test_SparseGPRegression_rbf_white_kern_2D(self):
         rbf = GPy.kern.rbf(2)
         self.check_model_with_white(rbf, model_type='SparseGPRegression', dimension=2)
 
+    def test_SparseGPRegression_rbf_linear_white_kern_1D(self):
+        ''' Testing the sparse GP regression with rbf and white kernel on 2d data '''
+        rbflin = GPy.kern.rbf(1) + GPy.kern.linear(1)
+        self.check_model_with_white(rbflin, model_type='SparseGPRegression', dimension=1)
+
+    def test_SparseGPRegression_rbf_linear_white_kern_2D(self):
+        ''' Testing the sparse GP regression with rbf and white kernel on 2d data '''
+        rbflin = GPy.kern.rbf(2) + GPy.kern.linear(2)
+        self.check_model_with_white(rbflin, model_type='SparseGPRegression', dimension=2)
+
+    def test_SparseGPRegression_rbf_linear_white_kern_2D_uncertain_inputs(self):
+        ''' Testing the sparse GP regression with rbf, linear and white kernel on 2d data with uncertain inputs'''
+        rbflin = GPy.kern.rbf(2) + GPy.kern.linear(2)
+        self.check_model_with_white(rbflin, model_type='SparseGPRegression', dimension=2, uncertain_inputs=1)
+
+    def test_SparseGPRegression_rbf_linear_white_kern_1D_uncertain_inputs(self):
+        ''' Testing the sparse GP regression with rbf, linear and white kernel on 1d data with uncertain inputs'''
+        rbflin = GPy.kern.rbf(1) + GPy.kern.linear(1)
+        self.check_model_with_white(rbflin, model_type='SparseGPRegression', dimension=1, uncertain_inputs=1)
+
     def test_GPLVM_rbf_bias_white_kern_2D(self):
         """ Testing GPLVM with rbf + bias and white kernel """
         N, input_dim, D = 50, 1, 2