Rename MeasurableVariable to MeasurableOp

Also introduce MeasurableOpMixin for string representation

Author: ricardoV94

pymc/logprob/basic.py

@@ -485,19 +485,19 @@ def conditional_logp(
 
     # This is the updated random-to-value-vars map with the lifted/rewritten
     # variables.  The rewrites are supposed to produce new
-    # `MeasurableVariable`s that are amenable to `_logprob`.
+    # `MeasurableOp`s whose variables are amenable to `_logprob`.
     updated_rv_values = rv_remapper.rv_values
 
     # Some rewrites also transform the original value variables. This is the
     # updated map from the new value variables to the original ones, which
     # we want to use as the keys in the final dictionary output
     original_values = rv_remapper.original_values
 
-    # When a `_logprob` has been produced for a `MeasurableVariable` node, all
+    # When a `_logprob` has been produced for a `MeasurableOp` node, all
     # other references to it need to be replaced with its value-variable all
     # throughout the `_logprob`-produced graphs.  The following `dict`
     # cumulatively maintains remappings for all the variables/nodes that needed
-    # to be recreated after replacing `MeasurableVariable`s with their
+    # to be recreated after replacing `MeasurableOp` variables with their
     # value-variables.  Since these replacements work in topological order, all
     # the necessary value-variable replacements should be present for each
     # node.

pymc/logprob/rewriting.py

@@ -152,7 +152,7 @@ def apply(self, fgraph):
                         rewriter_name = getattr(node_rewriter, "name", None) or getattr(
                             node_rewriter, "__name__", ""
                         )
-                    # If we converted to a MeasurableVariable we're done here!
+                    # If we converted to a MeasurableOp we're done here!
                     if node not in fgraph.apply_nodes or isinstance(node.op, MeasurableOp):
                         # go to next node
                         break
@@ -364,11 +364,11 @@ def construct_ir_fgraph(
 
     Our measurable IR takes the form of an PyTensor graph that is more-or-less
     equivalent to a given PyTensor graph (i.e. the keys of `rv_values`) but
-    contains `Op`s that are subclasses of the `MeasurableVariable` type in
-    place of ones that do not inherit from `MeasurableVariable` in the original
+    contains `Op`s that are subclasses of the `MeasurableOp` type in
+    place of ones that do not inherit from `MeasurableOp` in the original
     graph but are nevertheless measurable.
 
-    `MeasurableVariable`\s are mapped to log-probabilities, so this IR is how
+    `MeasurableOp` variables are mapped to log-probabilities, so this IR is how
     non-trivial log-probabilities are constructed, especially when the
     "measurability" of a term depends on the measurability of its inputs
     (e.g. a mixture).
@@ -381,11 +381,6 @@ def construct_ir_fgraph(
     measurable IR includes manipulations that are not applicable to outside of
     the context of measurability/log-probabilities.
 
-    For instance, some `Op`s will be lifted through `MeasurableVariable`\s in
-    this IR, and the resulting graphs will not be computationally sound,
-    because they wouldn't produce independent samples when the original graph
-    would.  See https://github.com/aesara-devs/aeppl/pull/78.
-
     Returns
     -------
     A `FunctionGraph` of the measurable IR, a copy of `rv_values` containing

pymc/logprob/scan.py

@@ -269,13 +269,13 @@ def remove(x, i):
 def get_random_outer_outputs(
     scan_args: ScanArgs,
 ) -> list[tuple[int, TensorVariable, TensorVariable]]:
-    """Get the `MeasurableVariable` outputs of a `Scan` (well, its `ScanArgs`).
+    """Get the measurable outputs of a `Scan` (well, its `ScanArgs`).
 
     Returns
     -------
     A tuple of tuples containing the index of each outer-output variable, the
     outer-output variable itself, and the inner-output variable that
-    is an instance of `MeasurableVariable`.
+    is an instance of `MeasurableOp` variable.
     """
     rv_vars = []
     for n, oo_var in enumerate(
@@ -376,7 +376,7 @@ def find_measurable_scans(fgraph, node):
 
     curr_scanargs = ScanArgs.from_node(node)
 
-    # Find the un-output `MeasurableVariable`s created in the inner-graph
+    # Find the un-output `MeasurablOp` variables created in the inner-graph
     if not any(out in rv_map_feature.rv_values for out in node.outputs):
         # TODO: T
         # We need to remap user inputs that have been specified in terms of

pymc/logprob/utils.py

@@ -132,7 +132,7 @@ def populate_replacements(var):
 
 
 def rvs_in_graph(vars: Variable | Sequence[Variable]) -> set[Variable]:
-    """Assert that there are no `MeasurableVariable` nodes in a graph."""
+    """Assert that there are no `MeasurableOp` nodes in a graph."""
 
     def expand(r):
         owner = r.owner
@@ -178,7 +178,7 @@ def check_potential_measurability(
     valued_rvs = set(valued_rvs)
 
     def expand_fn(var):
-        # expand_fn does not go beyond valued_rvs or any MeasurableVariable
+        # expand_fn does not go beyond valued_rvs or any MeasurableOp variables
         if var.owner and not isinstance(var.owner.op, MeasurableOp) and var not in valued_rvs:
             return reversed(var.owner.inputs)
         else:

pymc/testing.py

@@ -978,7 +978,7 @@ def seeded_numpy_distribution_builder(dist_name: str) -> Callable:
 
 
 def assert_no_rvs(vars: Sequence[Variable]) -> None:
-    """Assert that there are no `MeasurableVariable` nodes in a graph."""
+    """Assert that there are no `MeasurableOp` nodes in a graph."""
 
     rvs = rvs_in_graph(vars)
     if rvs:

tests/logprob/test_transform_value.py

@@ -470,7 +470,7 @@ def test_transformed_rv_and_value():
 
 @pytest.mark.filterwarnings("error")
 def test_mixture_transform():
-    """Make sure that non-`RandomVariable` `MeasurableVariable`s can be transformed.
+    """Make sure that non-`RandomVariable` `MeasurableOp` variables can be transformed.
 
     This test is specific to `MixtureRV`, which is derived from an `OpFromGraph`.
     """