minor refactoring and version change

Author: jsalvatier

pymc/NumpyDeterministics.py

@@ -15,7 +15,7 @@
 
 
 #transformations (broadcasted)
-_generic = ['abs', 'exp', 'log', 'sqrt','expm1', 'log1p]
+_generic = ['abs', 'exp', 'log', 'sqrt','expm1', 'log1p']
 _trig = ['sin', 'cos', 'tan', 'arcsin', 'arccos', 'arctan']
 _hyp_trig = ['sinh', 'cosh', 'tanh', 'arcsinh', 'arccosh', 'arctanh']
 _transformation_deterministics = _generic + _trig + _hyp_trig

pymc/PyMCObjects.py

@@ -437,6 +437,21 @@ def set_value(self,value):
 
     value = property(fget = get_value, fset=set_value, doc="Self's value computed from current values of parents.")
 
+    def apply_jacobian(self, parameter, variable, gradient):
+        try :
+            jacobian_func = self._jacobians[parameter]
+        except KeyError:
+            raise NotImplementedError(repr(self) + " has no jacobian function for parameter " + parameter)
+        
+        jacobian = jacobian_func.get()
+        
+        
+        mapping = self._jacobian_formats.get(parameter, 'full')
+
+            
+        return self._format_mapping[mapping](self, variable, jacobian, gradient)
+    
+        
     def logp_partial_gradient(self, variable, calculation_set = None):
         """
         gets the logp gradient of this deterministic with respect to variable
@@ -446,28 +461,15 @@ def logp_partial_gradient(self, variable, calculation_set = None):
 
         if not (variable.dtype in float_dtypes and self.dtype in float_dtypes):
                 return zeros(shape(variable.value))
-            
-        #gradient = 0
 
         # loop through all the parameters and add up all the gradients of log p with respect to the approrpiate variable
         gradient = __builtin__.sum([child.logp_partial_gradient(self, calculation_set) for child in self.children ])
 
         totalGradient = 0
         for parameter, value in self.parents.iteritems():
             if value is variable:
-
-                try :
-                    jacobian_func = self._jacobians[parameter]
-                except KeyError:
-                    raise NotImplementedError(repr(self) + " has no jacobian function for parameter " + parameter)
-                
-                jacobian = jacobian_func.get()
-                
-                
-                mapping = self._jacobian_formats.get(parameter, 'full')
-
 
-                totalGradient += self._format_mapping[mapping](self, variable, jacobian, gradient)
+                totalGradient += self.apply_gradient(parameter, variable, gradient )
 
         return np.reshape(totalGradient, shape(variable.value))
 
@@ -478,20 +480,20 @@ def full_jacobian(self, variable, jacobian, gradient):
     def transformation_operation_jacobian(self, variable, jacobian, gradient):
         return jacobian * gradient
 
-    _bog_history = {}
+    _BO_history = {}
     def broadcast_operation_jacobian(self, variable, jacobian, gradient):
 
         tgradient = jacobian * gradient
         try :
-            axes, lx = self._bog_history[id(variable)]
+            axes, lx = self._BO_history[id(variable)]
         except KeyError: 
             sshape = array(shape(self.value))
             vshape = zeros(sshape.size)
             vshape[0:ndim(variable.value)] += array(shape(variable.value))
             axes = np.where(sshape != vshape)
             lx = size(axes)
 
-            self._bog_history[id(variable)] = (axes, lx )
+            self._BO_history[id(variable)] = (axes, lx )
 
         if lx > 0:
             return np.apply_over_axes(np.sum, tgradient, axes)   
@@ -506,7 +508,6 @@ def accumulation_operation_jacobian(self, variable, jacobian, gradient):
         return gradient * jacobian
 
     def index_operation_jacobian(self, variable, jacobian, gradient):
-        #index = jacobian
         derivative = zeros(variable.shape)
         derivative[jacobian] = gradient
         return derivative 

pymc/__init__.py

@@ -10,7 +10,7 @@
 
 """
 
-__version__ = '2.1alpha'
+__version__ = '2.2grad'
 
 try:
     import numpy