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

GPy/core/parameterization/lists_and_dicts.py

@@ -57,6 +57,7 @@ class ObservablesList(object):
 
     def __repr__(self):
         return self._poc.__repr__()
+    
 
     def add(self, priority, observable, callble):
         if observable is not None:
@@ -87,19 +88,17 @@ class ObservablesList(object):
     def __iter__(self):
         self.flush()
         for p, o, c in self._poc:
-            if o() is not None:
+            yield p, o(), c 
-                yield p, o(), c 
 
     def __len__(self):
         self.flush()
         return self._poc.__len__()
 
     def __deepcopy__(self, memo):
-        self.flush()
         s = ObservablesList()
-        for p,o,c in self._poc:
+        for p,o,c in self:
             import copy
-            s.add(p, copy.deepcopy(o(), memo), copy.deepcopy(c, memo))
+            s.add(p, copy.deepcopy(o, memo), copy.deepcopy(c, memo))
         s.flush()
         return s
 

GPy/core/parameterization/param.py

@@ -57,9 +57,9 @@ class Param(OptimizationHandlable, ObsAr):
 
     def build_pydot(self,G):
         import pydot
-        node = pydot.Node(id(self), shape='record', label=self.name)
+        node = pydot.Node(id(self), shape='trapezium', label=self.name)#, fontcolor='white', color='white')
         G.add_node(node)
-        for o in self.observers.keys():
+        for _, o, _ in self.observers:
             label = o.name if hasattr(o, 'name') else str(o)
             observed_node = pydot.Node(id(o), label=label)
             G.add_node(observed_node)
@@ -169,8 +169,29 @@ class Param(OptimizationHandlable, ObsAr):
     # parameterizable
     #===========================================================================
     def traverse(self, visit, *args, **kwargs):
-        visit(self, *args, **kwargs)
+        """
+        Traverse the hierarchy performing visit(self, *args, **kwargs) at every node passed by.
+        See "visitor pattern" in literature. This is implemented in pre-order fashion.
 
+        This will function will just call visit on self, as Param are leaf nodes.
+        """
+        visit(self, *args, **kwargs)
+    
+    def traverse_parents(self, visit, *args, **kwargs):
+        """
+        Traverse the hierarchy upwards, visiting all parents and their children, except self.
+        See "visitor pattern" in literature. This is implemented in pre-order fashion.
+    
+        Example:
+    
+        parents = []
+        self.traverse_parents(parents.append)
+        print parents
+        """
+        if self.has_parent():
+            self.__visited = True
+            self._parent_._traverse_parents(visit, *args, **kwargs)
+            self.__visited = False
 
     #===========================================================================
     # Convenience

GPy/core/parameterization/parameter_core.py

@@ -176,24 +176,23 @@ class Pickleable(object):
         #raise NotImplementedError, "Copy is not yet implemented, TODO: Observable hierarchy"
         import copy
         memo = {}
+        # the next part makes sure that we do not include parents in any form:
         parents = []
-        self.traverse_parents(parents.append)
+        self.traverse_parents(parents.append) # collect parents
-        # remove self, which is the first arguments
-        parents = [p for p in parents if p is not self]
         for p in parents:
-            memo[id(p)] = None
+            memo[id(p)] = None # set all parents to be None, so they will not be copied
-        memo[id(self.gradient)] = None
+        memo[id(self.gradient)] = None # reset the gradient
-        memo[id(self.param_array)] = None
+        memo[id(self.param_array)] = None # and param_array
-        memo[id(self._fixes_)] = None
+        memo[id(self._fixes_)] = None # fixes have to be reset, as this is now highest parent
-        c = copy.deepcopy(self, memo)
+        c = copy.deepcopy(self, memo) # and start the copy
         c._parent_index_ = None
         return c
 
     def __deepcopy__(self, memo):
-        s = self.__new__(self.__class__)
+        s = self.__new__(self.__class__) # fresh instance
-        memo[id(self)] = s
+        memo[id(self)] = s # be sure to break all cycles --> self is already done
         import copy
-        s.__dict__.update(copy.deepcopy(self.__dict__, memo))
+        s.__dict__.update(copy.deepcopy(self.__dict__, memo)) # standard copy
         return s
 
     def __getstate__(self):
@@ -580,12 +579,6 @@ class OptimizationHandlable(Constrainable):
     def __init__(self, name, default_constraint=None, *a, **kw):
         super(OptimizationHandlable, self).__init__(name, default_constraint=default_constraint, *a, **kw)
 
-    def transform(self):
-        [np.put(self.param_array, ind, c.finv(self.param_array.flat[ind])) for c, ind in self.constraints.iteritems() if c != __fixed__]
-
-    def untransform(self):
-        [np.put(self.param_array, ind, c.f(self.param_array.flat[ind])) for c, ind in self.constraints.iteritems() if c != __fixed__]
-
     def _get_params_transformed(self):
         # transformed parameters (apply transformation rules)
         p = self.param_array.copy()
@@ -599,15 +592,15 @@ class OptimizationHandlable(Constrainable):
         return p
 
     def _set_params_transformed(self, p):
-        if p is self.param_array:
+        if not(p is self.param_array):
-            p = p.copy()
+            if self.has_parent() and self.constraints[__fixed__].size != 0:
-        if self.has_parent() and self.constraints[__fixed__].size != 0:
+                fixes = np.ones(self.size).astype(bool)
-            fixes = np.ones(self.size).astype(bool)
+                fixes[self.constraints[__fixed__]] = FIXED
-            fixes[self.constraints[__fixed__]] = FIXED
+                self.param_array.flat[fixes] = p
-            self.param_array.flat[fixes] = p
+            elif self._has_fixes(): self.param_array.flat[self._fixes_] = p
-        elif self._has_fixes(): self.param_array.flat[self._fixes_] = p
+            else: self.param_array.flat = p
-        else: self.param_array.flat = p
+        [np.put(self.param_array, ind, c.f(self.param_array.flat[ind])) 
-        self.untransform()
+         for c, ind in self.constraints.iteritems() if c != __fixed__]
         self._trigger_params_changed()
 
     def _trigger_params_changed(self, trigger_parent=True):
@@ -621,7 +614,7 @@ class OptimizationHandlable(Constrainable):
     def num_params(self):
         """
         Return the number of parameters of this parameter_handle.
-        Param objects will allways return 0.
+        Param objects will always return 0.
         """
         raise NotImplemented, "Abstract, please implement in respective classes"
 
@@ -713,6 +706,10 @@ class Parameterizable(OptimizationHandlable):
 
     @property
     def param_array(self):
+        """
+        Array representing the parameters of this class.
+        There is only one copy of all parameters in memory, two during optimization.
+        """
         if self._param_array_ is None:
             self._param_array_ = np.empty(self.size, dtype=np.float64)
         return self._param_array_
@@ -723,7 +720,9 @@ class Parameterizable(OptimizationHandlable):
 
     def traverse(self, visit, *args, **kwargs):
         """
-        Traverse the hierarchy performing visit(self, *args, **kwargs) at every node passed by.
+        Traverse the hierarchy performing visit(self, *args, **kwargs) 
+        at every node passed by downwards. This function includes self!
+
         See "visitor pattern" in literature. This is implemented in pre-order fashion.
 
         Example:
@@ -738,25 +737,33 @@ class Parameterizable(OptimizationHandlable):
             self.__visited = True
             for c in self._parameters_:
                 c.traverse(visit, *args, **kwargs)
+            self.__visited = False
 
     def traverse_parents(self, visit, *args, **kwargs):
         """
-        Traverse the hierarchy upwards, visiting all parents and their children.
+        Traverse the hierarchy upwards, visiting all parents and their children except self.
         See "visitor pattern" in literature. This is implemented in pre-order fashion.
-
+    
         Example:
-
+    
         parents = []
         self.traverse_parents(parents.append)
         print parents
         """
-        if not self.__visited:
+        if self.has_parent():
-            visit(self, *args, **kwargs)
             self.__visited = True
+            self._parent_._traverse_parents(visit, *args, **kwargs)
+            self.__visited = False
+        
+    def _traverse_parents(self, visit, *args, **kwargs):
+        if not self.__visited:
+            self.__visited = True
+            visit(self, *args, **kwargs)
             if self.has_parent():
-                self._parent_.traverse_parents(visit, *args, **kwargs)
+                self._parent_._traverse_parents(visit, *args, **kwargs)
                 self._parent_.traverse(visit, *args, **kwargs)
             self.__visited = False
+
     #=========================================================================
     # Gradient handling
     #=========================================================================
@@ -823,11 +830,10 @@ class Parameterizable(OptimizationHandlable):
         #    raise HierarchyError, "parameter {} already in another model ({}), create new object (or copy) for adding".format(param._short(), param._highest_parent_._short())
         elif param not in self._parameters_:
             if param.has_parent():
-                parent = param._parent_
+                def visit(parent, self):
-                while parent is not None:
                     if parent is self:
                         raise HierarchyError, "You cannot add a parameter twice into the hierarchy"
-                    parent = parent._parent_
+                param.traverse_parents(visit, self)
                 param._parent_.remove_parameter(param)
             # make sure the size is set
             if index is None:
@@ -871,7 +877,7 @@ class Parameterizable(OptimizationHandlable):
         :param param: param object to remove from being a parameter of this parameterized object.
         """
         if not param in self._parameters_:
-            raise RuntimeError, "Parameter {} does not belong to this object, remove parameters directly from their respective parents".format(param._short())
+            raise RuntimeError, "Parameter {} does not belong to this object {}, remove parameters directly from their respective parents".format(param._short(), self.name)
 
         start = sum([p.size for p in self._parameters_[:param._parent_index_]])
         self._remove_parameter_name(param)

GPy/core/parameterization/parameterized.py

@@ -82,15 +82,15 @@ class Parameterized(Parameterizable):
         import pydot  # @UnresolvedImport
         iamroot = False
         if G is None:
-            G = pydot.Dot(graph_type='digraph')
+            G = pydot.Dot(graph_type='digraph', bgcolor=None)
             iamroot=True
-        node = pydot.Node(id(self), shape='record', label=self.name)
+        node = pydot.Node(id(self), shape='box', label=self.name)#, color='white')
         G.add_node(node)
         for child in self._parameters_:
             child_node = child.build_pydot(G)
-            G.add_edge(pydot.Edge(node, child_node))
+            G.add_edge(pydot.Edge(node, child_node))#, color='white'))
 
-        for o in self.observers.keys():
+        for _, o, _ in self.observers:
             label = o.name if hasattr(o, 'name') else str(o)
             observed_node = pydot.Node(id(o), label=label)
             G.add_node(observed_node)

GPy/gpy_config.cfg

@@ -6,6 +6,10 @@
 # some platforms, hence this option.
 openmp=False
 
+[datasets]
+# location for the local data cache
+dir=$HOME/tmp/GPy-datasets/
+
 [anaconda]
 # if you have an anaconda python installation please specify it here.
 installed = False

GPy/kern/_src/linear.py

@@ -12,6 +12,7 @@ from ...core.parameterization.transformations import Logexp
 from ...util.caching import Cache_this
 from ...core.parameterization import variational
 from psi_comp import linear_psi_comp
+from ...util.config import *
 
 class Linear(Kern):
     """
@@ -224,12 +225,23 @@ class Linear(Kern):
         AZZA = ZA.T[:, None, :, None] * ZA[None, :, None, :]
         AZZA = AZZA + AZZA.swapaxes(1, 2)
         AZZA_2 = AZZA/2.
+        if config.getboolean('parallel', 'openmp'):
+            pragma_string = '#pragma omp parallel for private(m,mm,q,qq,factor,tmp)'
+            header_string = '#include <omp.h>'
+            weave_options = {'headers'           : ['<omp.h>'],
+                                'extra_compile_args': ['-fopenmp -O3'],
+                                'extra_link_args'   : ['-lgomp'],
+                                'libraries': ['gomp']}
+        else:
+            pragma_string = ''
+            header_string = ''
+            weave_options = {'extra_compile_args': ['-O3']}
 
         #Using weave, we can exploit the symmetry of this problem:
         code = """
         int n, m, mm,q,qq;
         double factor,tmp;
-        #pragma omp parallel for private(m,mm,q,qq,factor,tmp)
+        %s
         for(n=0;n<N;n++){
           for(m=0;m<num_inducing;m++){
             for(mm=0;mm<=m;mm++){
@@ -253,26 +265,36 @@ class Linear(Kern):
             }
           }
         }
-        """
+        """ % pragma_string
         support_code = """
-        #include <omp.h>
+        %s
         #include <math.h>
-        """
+        """ % header_string
-        weave_options = {'headers'           : ['<omp.h>'],
-                         'extra_compile_args': ['-fopenmp -O3'],  #-march=native'],
-                         'extra_link_args'   : ['-lgomp']}
         mu = vp.mean
         N,num_inducing,input_dim,mu = mu.shape[0],Z.shape[0],mu.shape[1],param_to_array(mu)
-        weave.inline(code, support_code=support_code, libraries=['gomp'],
+        weave.inline(code, support_code=support_code,
                      arg_names=['N','num_inducing','input_dim','mu','AZZA','AZZA_2','target_mu','target_S','dL_dpsi2'],
                      type_converters=weave.converters.blitz,**weave_options)
 
 
     def _weave_dpsi2_dZ(self, dL_dpsi2, Z, vp, target):
         AZA = self.variances*self._ZAinner(vp, Z)
+
+        if config.getboolean('parallel', 'openmp'):
+            pragma_string = '#pragma omp parallel for private(n,mm,q)'
+            header_string = '#include <omp.h>'
+            weave_options =  {'headers'           : ['<omp.h>'],
+                              'extra_compile_args': ['-fopenmp -O3'],
+                              'extra_link_args'   : ['-lgomp'],
+                              'libraries': ['gomp']}
+        else:
+            pragma_string = ''
+            header_string = ''
+            weave_options = {'extra_compile_args': ['-O3']}
+
         code="""
         int n,m,mm,q;
-        #pragma omp parallel for private(n,mm,q)
+        %s
         for(m=0;m<num_inducing;m++){
           for(q=0;q<input_dim;q++){
             for(mm=0;mm<num_inducing;mm++){
@@ -282,18 +304,15 @@ class Linear(Kern):
             }
           }
         }
-        """
+        """ % pragma_string
         support_code = """
-        #include <omp.h>
+        %s
         #include <math.h>
-        """
+        """ % header_string
-        weave_options = {'headers'           : ['<omp.h>'],
-                         'extra_compile_args': ['-fopenmp -O3'],  #-march=native'],
-                         'extra_link_args'   : ['-lgomp']}
 
         N,num_inducing,input_dim = vp.mean.shape[0],Z.shape[0],vp.mean.shape[1]
         mu = param_to_array(vp.mean)
-        weave.inline(code, support_code=support_code, libraries=['gomp'],
+        weave.inline(code, support_code=support_code,
                      arg_names=['N','num_inducing','input_dim','AZA','target','dL_dpsi2'],
                      type_converters=weave.converters.blitz,**weave_options)
 

GPy/kern/_src/rbf.py

@@ -10,6 +10,7 @@ from GPy.util.caching import Cache_this
 from ...core.parameterization import variational
 from psi_comp import ssrbf_psi_comp
 from psi_comp.ssrbf_psi_gpucomp import PSICOMP_SSRBF
+from ...util.config import *
 
 class RBF(Stationary):
     """
@@ -231,6 +232,16 @@ class RBF(Stationary):
 
     @Cache_this(limit=1)
     def _psi2computations(self, Z, vp):
+
+        if config.getboolean('parallel', 'openmp'):
+            pragma_string = '#pragma omp parallel for private(tmp, exponent_tmp)'
+            header_string = '#include <omp.h>'
+            libraries = ['gomp']
+        else:
+            pragma_string = ''
+            header_string = ''
+            libraries = []
+
         mu, S = vp.mean, vp.variance
 
         N, Q = mu.shape
@@ -253,8 +264,7 @@ class RBF(Stationary):
         variance_sq = float(np.square(self.variance))
         code = """
         double tmp, exponent_tmp;
-
+        %s 
-        #pragma omp parallel for private(tmp, exponent_tmp)
         for (int n=0; n<N; n++)
         {
             for (int m=0; m<M; m++)
@@ -278,20 +288,20 @@ class RBF(Stationary):
                         tmp = -Zdist_sq(m,mm,q) - tmp - half_log_denom(n,q);
                         exponent_tmp += tmp;
                     }
-                    //compute psi2 by exponontiating
+                    //compute psi2 by exponentiating
                     psi2(n,m,mm) = variance_sq * exp(exponent_tmp);
                     psi2(n,mm,m) = psi2(n,m,mm);
                 }
             }
         }
-        """
+        """ % pragma_string
 
         support_code = """
-        #include <omp.h>
+        %s
         #include <math.h>
-        """
+        """ % header_string
         mu = param_to_array(mu)
-        weave.inline(code, support_code=support_code, libraries=['gomp'],
+        weave.inline(code, support_code=support_code, libraries=libraries,
                      arg_names=['N', 'M', 'Q', 'mu', 'Zhat', 'mudist_sq', 'mudist', 'denom_l2', 'Zdist_sq', 'half_log_denom', 'psi2', 'variance_sq'],
                      type_converters=weave.converters.blitz, **self.weave_options)
 
@@ -303,12 +313,20 @@ class RBF(Stationary):
         #return 2.*np.einsum( 'ijk,ijk,ijkl,il->l', dL_dpsi2, psi2, Zdist_sq * (2.*S[:,None,None,:]/l2 + 1.) + mudist_sq + S[:, None, None, :] / l2, 1./(2.*S + l2))*self.lengthscale
 
         result = np.zeros(self.input_dim)
+        if config.getboolean('parallel', 'openmp'):
+            pragma_string = '#pragma omp parallel for reduction(+:tmp)'
+            header_string = '#include <omp.h>'
+            libraries = ['gomp']
+        else:
+            pragma_string = ''
+            header_string = ''
+            libraries = []
         code = """
         double tmp;
         for(int q=0; q<Q; q++)
         {
             tmp = 0.0;
-            #pragma omp parallel for reduction(+:tmp)
+            %s
             for(int n=0; n<N; n++)
             {
                 for(int m=0; m<M; m++)
@@ -326,16 +344,16 @@ class RBF(Stationary):
             result(q) = tmp;
         }
 
-        """
+        """ % pragma_string
         support_code = """
-        #include <omp.h>
+        %s
         #include <math.h>
-        """
+        """ % header_string
         N,Q = S.shape
         M = psi2.shape[-1]
 
         S = param_to_array(S)
-        weave.inline(code, support_code=support_code, libraries=['gomp'],
+        weave.inline(code, support_code=support_code, libraries=libraries,
                      arg_names=['psi2', 'dL_dpsi2', 'N', 'M', 'Q', 'mudist_sq', 'l2', 'Zdist_sq', 'S', 'result'],
                      type_converters=weave.converters.blitz, **self.weave_options)
 

GPy/likelihoods/ordinal.py

@@ -0,0 +1,48 @@
+# Copyright (c) 2014 The GPy authors (see AUTHORS.txt)
+# Licensed under the BSD 3-clause license (see LICENSE.txt)
+
+
+import sympy as sym
+from GPy.util.symbolic import gammaln, normcdfln, normcdf, IndMatrix, create_matrix
+import numpy as np
+from ..util.univariate_Gaussian import std_norm_pdf, std_norm_cdf
+import link_functions
+from symbolic import Symbolic
+from scipy import stats
+
+class Ordinal(Symbolic):
+    """
+    Ordinal
+
+    .. math::
+        p(y_{i}|\pi(f_{i})) = \left(\frac{r}{r+f_i}\right)^r \frac{\Gamma(r+y_i)}{y!\Gamma(r)}\left(\frac{f_i}{r+f_i}\right)^{y_i}
+
+    .. Note::
+        Y takes non zero integer values..
+        link function should have a positive domain, e.g. log (default).
+
+    .. See also::
+        symbolic.py, for the parent class
+    """
+    def __init__(self, categories=3, gp_link=None):
+        if gp_link is None:
+            gp_link = link_functions.Identity()
+
+        dispersion = sym.Symbol('width', positive=True, real=True)
+        y_0 = sym.Symbol('y_0', nonnegative=True, integer=True)
+        f_0 = sym.Symbol('f_0', positive=True, real=True) 
+        log_pdf = create_matrix('log_pdf', 1, categories)
+        log_pdf[0] = normcdfln(-f_0)
+        if categories>2:
+            w = create_matrix('w', 1, categories)
+            log_pdf[categories-1] = normcdfln(w.sum() + f_0)
+            for i in range(1, categories-1):
+                log_pdf[i] = sym.log(normcdf(w[0, 0:i-1].sum() + f_0) - normcdf(w[0, 0:i].sum()-f_0) )
+        else:
+            log_pdf[1] = normcdfln(f_0)
+        log_pdf.index_var = y_0
+        super(Ordinal, self).__init__(log_pdf=log_pdf, gp_link=gp_link, name='Ordinal')
+
+        # TODO: Check this.
+        self.log_concave = True
+

GPy/models/bayesian_gplvm.py

@@ -42,7 +42,7 @@ class BayesianGPLVM(SparseGP):
         assert Z.shape[1] == X.shape[1]
 
         if kernel is None:
-            kernel = kern.RBF(input_dim, lengthscale=fracs, ARD=True) # + kern.white(input_dim)
+            kernel = kern.RBF(input_dim, lengthscale=1./fracs, ARD=True) # + kern.white(input_dim)
 
         if likelihood is None:
             likelihood = Gaussian()

GPy/plotting/matplot_dep/dim_reduction_plots.py

@@ -97,7 +97,7 @@ def plot_latent(model, labels=None, which_indices=None,
         elif type(ul) is np.int64:
             this_label = 'class %i' % ul
         else:
-            this_label = 'class %i' % i
+            this_label = unicode(ul)
         m = marker.next()
 
         index = np.nonzero(labels == ul)[0]

GPy/plotting/matplot_dep/models_plots.py

@@ -14,7 +14,7 @@ def plot_fit(model, plot_limits=None, which_data_rows='all',
         which_data_ycols='all', fixed_inputs=[],
         levels=20, samples=0, fignum=None, ax=None, resolution=None,
         plot_raw=False,
-        linecol=Tango.colorsHex['darkBlue'],fillcol=Tango.colorsHex['lightBlue'], Y_metadata=None):
+        linecol=Tango.colorsHex['darkBlue'],fillcol=Tango.colorsHex['lightBlue'], Y_metadata=None, data_symbol='kx'):
     """
     Plot the posterior of the GP.
       - In one dimension, the function is plotted with a shaded region identifying two standard deviations.
@@ -97,7 +97,7 @@ def plot_fit(model, plot_limits=None, which_data_rows='all',
 
         for d in which_data_ycols:
             plots['gpplot'] = gpplot(Xnew, m[:, d], lower[:, d], upper[:, d], ax=ax, edgecol=linecol, fillcol=fillcol)
-            plots['dataplot'] = ax.plot(X[which_data_rows,free_dims], Y[which_data_rows, d], 'kx', mew=1.5)
+            plots['dataplot'] = ax.plot(X[which_data_rows,free_dims], Y[which_data_rows, d], data_symbol, mew=1.5)
 
         #optionally plot some samples
         if samples: #NOTE not tested with fixed_inputs

GPy/testing/pickle_tests.py

@@ -132,6 +132,9 @@ class Test(ListDictTestCase):
         self.assertIsNot(par.full_gradient, pcopy.full_gradient)
         self.assertTrue(pcopy.checkgrad())
         self.assert_(np.any(pcopy.gradient!=0.0))
+        pcopy.optimize('bfgs')
+        par.optimize('bfgs')
+        np.testing.assert_allclose(pcopy.param_array, par.param_array, atol=.001)
         with tempfile.TemporaryFile('w+b') as f:
             par.pickle(f)
             f.seek(0)

GPy/util/data_resources.json

@@ -150,6 +150,26 @@
         ]
     },
     "fruitfly_tomancak": {
+        "citation": "",
+        "details": "",
+        "files": [
+            [
+                "tomancak_exprs.csv",
+                "tomancak_se.csv",
+                "tomancak_prctile5.csv",
+                "tomancak_prctile25.csv",
+                "tomancak_prctile50.csv",
+                "tomancak_prctile75.csv",
+                "tomancak_prctile95.csv"
+            ]
+        ],
+        "license": null,
+        "size": 59000000,
+        "urls": [
+            "http://staffwww.dcs.shef.ac.uk/people/N.Lawrence/dataset_mirror/fruitfly_tomancak/"
+        ]
+    },
+    "fruitfly_tomancak_cel_files": {
         "citation": "'Systematic determination of patterns of gene expression during Drosophila embryogenesis' Pavel Tomancak, Amy Beaton, Richard Weiszmann, Elaine Kwan, ShengQiang Shu, Suzanna E Lewis, Stephen Richards, Michael Ashburner, Volker Hartenstein, Susan E Celniker, and Gerald M Rubin",
         "details": "Gene expression results from blastoderm development in Drosophila Melanogaster.",
         "files": [ 
@@ -198,7 +218,7 @@
             ]
         ],
         "license": null,
-        "size": 1,
+        "size": 389000000,
         "urls": [
             "ftp://ftp.fruitfly.org/pub/embryo_tc_array_data/"
         ]
@@ -217,6 +237,7 @@
             "http://www.google.com/trends/"
         ]
     },
+    
     "hapmap3": {
         "citation": "Gibbs, Richard A., et al. 'The international HapMap project.' Nature 426.6968 (2003): 789-796.",
         "details": "HapMap Project: Single Nucleotide Polymorphism sequenced in all human populations. \n        The HapMap phase three SNP dataset - 1184 samples out of 11 populations.\n        See http://www.nature.com/nature/journal/v426/n6968/abs/nature02168.html for details.\n\n        SNP_matrix (A) encoding [see Paschou et all. 2007 (PCA-Correlated SNPs...)]:\n        Let (B1,B2) be the alphabetically sorted bases, which occur in the j-th SNP, then\n\n              /  1, iff SNPij==(B1,B1)\n        Aij = |  0, iff SNPij==(B1,B2)\n              \\\\ -1, iff SNPij==(B2,B2)\n\n        The SNP data and the meta information (such as iid, sex and phenotype) are\n        stored in the dataframe datadf, index is the Individual ID, \n        with following columns for metainfo:\n\n            * family_id   -> Family ID\n            * paternal_id -> Paternal ID\n            * maternal_id -> Maternal ID\n            * sex         -> Sex (1=male; 2=female; other=unknown)\n            * phenotype   -> Phenotype (-9, or 0 for unknown)\n            * population  -> Population string (e.g. 'ASW' - 'YRI')\n            * rest are SNP rs (ids)\n\n        More information is given in infodf:\n\n            * Chromosome:\n                - autosomal chromosemes                -> 1-22\n                - X    X chromosome                    -> 23\n                - Y    Y chromosome                    -> 24\n                - XY   Pseudo-autosomal region of X    -> 25\n                - MT   Mitochondrial                   -> 26\n            * Relative Positon (to Chromosome) [base pairs]\n\n        ",
@@ -434,7 +455,7 @@
     },
     "singlecell": {
         "citation": "Guoji Guo, Mikael Huss, Guo Qing Tong, Chaoyang Wang, Li Li Sun, Neil D. Clarke, Paul Robson, Resolution of Cell Fate Decisions Revealed by Single-Cell Gene Expression Analysis from Zygote to Blastocyst, Developmental Cell, Volume 18, Issue 4, 20 April 2010, Pages 675-685, ISSN 1534-5807, http://dx.doi.org/10.1016/j.devcel.2010.02.012. (http://www.sciencedirect.com/science/article/pii/S1534580710001103) Keywords: DEVBIO",
-        "details": "qPCR Singlecell experiment in Mouse, measuring 48 gene expressions in 1-64 cell states. The labels have been created as in Guo et al. [2010]",
+        "details": "qPCR TaqMan array single cell experiment in mouse. The data is taken from the early stages of development when the Blastocyst is forming. At the 32 cell stage the data is already separated into the trophectoderm (TE) which goes onto form the placenta and the inner cellular mass (ICM). The ICM further differentiates into the epiblast (EPI)---which gives rise to the endoderm, mesoderm and ectoderm---and the primitive endoderm (PE) which develops into the amniotic sack. Guo et al selected 48 genes for expression measurement. They labelled the resulting cells and their labels are included as an aide to visualization.",
         "files": [
             [
                 "singlecell.csv"
@@ -443,7 +464,22 @@
         "license": "ScienceDirect: http://www.elsevier.com/locate/termsandconditions?utm_source=sciencedirect&utm_medium=link&utm_campaign=terms",
         "size": 233.1,
         "urls": [
-            "http://staffwww.dcs.sheffield.ac.uk/people/M.Zwiessele/data/singlecell/"
+            "http://staffwww.dcs.shef.ac.uk/people/N.Lawrence/dataset_mirror/singlecell/"
+        ]
+    },
+    "sod1_mouse": {
+        "citation": "Transcriptomic indices of fast and slow disease progression in two mouse models of amyotrophic lateral sclerosis' Nardo G1, Iennaco R, Fusi N, Heath PR, Marino M, Trolese MC, Ferraiuolo L, Lawrence N, Shaw PJ, Bendotti C Brain. 2013 Nov;136(Pt 11):3305-32. doi: 10.1093/brain/awt250. Epub 2013 Sep 24.",
+        "details": "Gene expression data from two separate strains of mice: C57 and 129Sv in wild type and SOD1 mutant strains.",
+        "files": [
+            [
+                "sod1_C57_129_exprs.csv",
+                "sod1_C57_129_se.csv"
+            ]
+        ],
+        "license": null,
+        "size": 0,
+        "urls": [
+            "http://staffwww.dcs.shef.ac.uk/people/N.Lawrence/dataset_mirror/sod1_mouse/"
         ]
     },
     "swiss_roll": {

GPy/util/datasets.py

@@ -12,6 +12,8 @@ import datetime
 import json
 import re
 
+from config import *
+
 ipython_available=True
 try:
     import IPython
@@ -29,7 +31,8 @@ def reporthook(a,b,c):
     sys.stdout.flush()
 
 # Global variables
-data_path = os.path.join(os.path.dirname(__file__), 'datasets')
+data_path = os.path.expandvars(config.get('datasets', 'dir'))
+#data_path = os.path.join(os.path.dirname(__file__), 'datasets')
 default_seed = 10000
 overide_manual_authorize=False
 neil_url = 'http://staffwww.dcs.shef.ac.uk/people/N.Lawrence/dataset_mirror/'
@@ -112,7 +115,7 @@ def download_url(url, store_directory, save_name = None, messages = True, suffix
         if content_length_str:
             file_size = int(content_length_str[0])
         else:
-            file_size = 1e10
+            file_size = None
         status = ""
         file_size_dl = 0
         block_sz = 8192
@@ -124,9 +127,15 @@ def download_url(url, store_directory, save_name = None, messages = True, suffix
             file_size_dl += len(buff)
             f.write(buff)
             sys.stdout.write(" "*(len(status)) + "\r")
-            status = r"[{perc: <{ll}}] {dl:7.3f}/{full:.3f}MB".format(dl=file_size_dl/(1.*1e6), 
+            if file_size:
-                                                                       full=file_size/(1.*1e6), ll=line_length, 
+                status = r"[{perc: <{ll}}] {dl:7.3f}/{full:.3f}MB".format(dl=file_size_dl/(1048576.), 
+                                                                       full=file_size/(1048576.), ll=line_length, 
                                                                        perc="="*int(line_length*float(file_size_dl)/file_size))
+            else:
+                status = r"[{perc: <{ll}}] {dl:7.3f}MB".format(dl=file_size_dl/(1048576.), 
+                                                                       ll=line_length, 
+                                                                       perc="."*int(line_length*float(file_size_dl/(10*1048576.))))
+                
             sys.stdout.write(status)
             sys.stdout.flush()
         sys.stdout.write(" "*(len(status)) + "\r")
@@ -354,11 +363,32 @@ def football_data(season='1314', data_set='football_data'):
         Y = table[:, 4:]
     return data_details_return({'X': X, 'Y': Y}, data_set)
 
+def sod1_mouse(data_set='sod1_mouse'):
+    if not data_available(data_set):
+        download_data(data_set)
+    from pandas import read_csv
+    dirpath = os.path.join(data_path, data_set)
+    filename = os.path.join(dirpath, 'sod1_C57_129_exprs.csv')
+    Y = read_csv(filename, header=0, index_col=0)
+    num_repeats=4
+    num_time=4
+    num_cond=4
+    X = 1
+    return data_details_return({'X': X, 'Y': Y}, data_set)
+    
 def fruitfly_tomancak(data_set='fruitfly_tomancak', gene_number=None):
     if not data_available(data_set):
         download_data(data_set)
-    X = None
+    from pandas import read_csv
-    Y = None
+    dirpath = os.path.join(data_path, data_set)
+    filename = os.path.join(dirpath, 'tomancak_exprs.csv')
+    Y = read_csv(filename, header=0, index_col=0).T
+    num_repeats = 3
+    num_time = 12
+    xt = np.linspace(0, num_time-1, num_time)
+    xr = np.linspace(0, num_repeats-1, num_repeats)
+    xtime, xrepeat = np.meshgrid(xt, xr)
+    X = np.vstack((xtime.flatten(), xrepeat.flatten())).T    
     return data_details_return({'X': X, 'Y': Y, 'gene_number' : gene_number}, data_set)
 
 # This will be for downloading google trends data.
@@ -732,13 +762,16 @@ def hapmap3(data_set='hapmap3'):
 def singlecell(data_set='singlecell'):
     if not data_available(data_set):
         download_data(data_set)
+    
+    from pandas import read_csv
     dirpath = os.path.join(data_path, data_set)
-    data = np.loadtxt(os.path.join(dirpath, 'singlecell.csv'), delimiter=",", dtype=str)
+    filename = os.path.join(dirpath, 'singlecell.csv')
-    genes = data[0, 1:]
+    Y = read_csv(filename, header=0, index_col=0)
-    labels = data[1:, 0]
+    genes = Y.columns
-    Y = np.array(data[1:, 1:], dtype=float)
+    labels = Y.index
-    return data_details_return({'Y': Y, 'info' : "qPCR Singlecell experiment in Mouse, measuring 48 gene expressions in 1-64 cell states. The labels have been created as in Guo et al. [2010]",
+    # data = np.loadtxt(os.path.join(dirpath, 'singlecell.csv'), delimiter=",", dtype=str)
-                                'genes':genes, 'labels':labels,
+    return data_details_return({'Y': Y, 'info' : "qPCR singlecell experiment in Mouse, measuring 48 gene expressions in 1-64 cell states. The labels have been created as in Guo et al. [2010]",
+                                'genes': genes, 'labels':labels,
                                 }, data_set)
 
 def swiss_roll_1000():

GPy/util/initialization.py

@@ -13,11 +13,11 @@ def initialize_latent(init, input_dim, Y):
         p = pca(Y)
         PC = p.project(Y, min(input_dim, Y.shape[1]))
         Xr[:PC.shape[0], :PC.shape[1]] = PC
-        vars = p.fracs[:input_dim]
+        var = p.fracs[:input_dim]
     else:
-        vars = Xr.var(0)
+        var = Xr.var(0)
     
     Xr -= Xr.mean(0)
     Xr /= Xr.var(0)
     
-    return Xr, vars/vars.max()
+    return Xr, var/var.max()

GPy/util/misc.py

@@ -130,14 +130,14 @@ def fast_array_equal(A, B):
     """ % pragma_string
 
     if config.getboolean('parallel', 'openmp'):
-        pragma_string = '#include <omp.h>'
+        header_string = '#include <omp.h>'
     else:
-        pragma_string = ''
+        header_string = ''
 
     support_code = """
     %s
     #include <math.h>
-    """ % pragma_string
+    """ % header_string
 
 
     weave_options_openmp = {'headers'           : ['<omp.h>'],