Remove .log_density field from Delta funsor

funsor/delta.py

@@ -13,12 +13,11 @@
 
 class DeltaMeta(FunsorMeta):
     """
-    Wrapper to fill in defaults.
+    Wrapper to convert point to a funsor.
     """
-    def __call__(cls, name, point, log_density=0):
+    def __call__(cls, name, point):
         point = to_funsor(point)
-        log_density = to_funsor(log_density)
-        return super(DeltaMeta, cls).__call__(name, point, log_density)
+        return super(DeltaMeta, cls).__call__(name, point)
 
 
 @add_metaclass(DeltaMeta)
@@ -28,23 +27,16 @@ class Delta(Funsor):
 
     :param str name: Name of the bound variable.
     :param Funsor point: Value of the bound variable.
-    :param Funsor log_density: Optional log density to be added when evaluating
-        at a point. This is needed to make :class:`Delta` closed under
-        differentiable substitution.
     """
-    def __init__(self, name, point, log_density=0):
+    def __init__(self, name, point):
         assert isinstance(name, str)
         assert isinstance(point, Funsor)
-        assert isinstance(log_density, Funsor)
-        assert log_density.output == reals()
         inputs = OrderedDict([(name, point.output)])
         inputs.update(point.inputs)
-        inputs.update(log_density.inputs)
         output = reals()
         super(Delta, self).__init__(inputs, output)
         self.name = name
         self.point = point
-        self.log_density = log_density
 
     def eager_subs(self, subs):
         value = None
@@ -62,17 +54,16 @@ def eager_subs(self, subs):
 
         name = self.name
         point = self.point.eager_subs(index_part)
-        log_density = self.log_density.eager_subs(index_part)
         if value is not None:
             if isinstance(value, Variable):
                 name = value.name
             elif isinstance(value, (Number, Tensor)) and isinstance(point, (Number, Tensor)):
-                return (value == point).all().log() + log_density
+                return (value == point).all().log()
             else:
                 # TODO Compute a jacobian, update log_prob, and emit another Delta.
                 raise ValueError('Cannot substitute a {} into a Delta'
                                  .format(type(value).__name__))
-        return Delta(name, point, log_density)
+        return Delta(name, point)
 
     def eager_reduce(self, op, reduced_vars):
         if op is ops.logaddexp:

funsor/distributions.py

@@ -151,17 +151,16 @@ def eager_delta(v, log_density, value):
     return Tensor(data, inputs)
 
 
-@eager.register(Delta, Funsor, Funsor, Variable)
-@eager.register(Delta, Variable, Funsor, Variable)
+@eager.register(Delta, (Funsor, Variable), Funsor, Variable)
 def eager_delta(v, log_density, value):
     assert v.output == value.output
-    return funsor.delta.Delta(value.name, v, log_density)
+    return funsor.delta.Delta(value.name, v) + log_density
 
 
 @eager.register(Delta, Variable, Funsor, Funsor)
 def eager_delta(v, log_density, value):
     assert v.output == value.output
-    return funsor.delta.Delta(v.name, value, log_density)
+    return funsor.delta.Delta(v.name, value) + log_density
 
 
 class Normal(Distribution):

test/test_distributions.py

@@ -86,7 +86,7 @@ def test_delta_delta():
     point = Tensor(torch.randn(2))
     log_density = Tensor(torch.tensor(0.5))
     d = dist.Delta(point, log_density, v)
-    assert d is Delta('v', point, log_density)
+    assert d is Delta('v', point) + log_density
 
 
 def test_normal_defaults():