Merge pull request #3460 from rpgoldman/grapher-dfs

Fix for timeout in graph_model

RELEASE-NOTES.md

@@ -12,6 +12,7 @@
 - Used `numpy.vectorize` in `distributions.distribution._compile_theano_function`. This enables `sample_prior_predictive` and `sample_posterior_predictive` to ask for tuples of samples instead of just integers. This fixes issue #3422.
 
 ### Maintenance
+- Fixed an issue in `model_graph` that caused construction of the graph of the model for rendering to hang: replaced a search over the powerset of the nodes with a breadth-first search over the nodes. Fix for #3458.
 - All occurances of `sd` as a parameter name have been renamed to `sigma`. `sd` will continue to function for backwards compatibility.
 - Made `BrokenPipeError` for parallel sampling more verbose on Windows.
 - Added the `broadcast_distribution_samples` function that helps broadcasting arrays of drawn samples, taking into account the requested `size` and the inferred distribution shape. This sometimes is needed by distributions that call several `rvs` separately within their `random` method, such as the `ZeroInflatedPoisson` (Fix issue #3310).

pymc3/model_graph.py

@@ -1,11 +1,18 @@
 import itertools
+from collections import deque
+from typing import Iterator, Optional, MutableSet
 
-from theano.gof.graph import ancestors
+from theano.gof.graph import stack_search
 from theano.compile import SharedVariable
+from theano.tensor import Tensor
 
 from .util import get_default_varnames
 import pymc3 as pm
 
+# this is a placeholder for a better characterization of the type
+# of variables in a model.
+RV = Tensor
+
 
 def powerset(iterable):
     """All *nonempty* subsets of an iterable.
@@ -27,37 +34,41 @@ def __init__(self, model):
         self._deterministics = None
 
     def get_deterministics(self, var):
-        """Compute the deterministic nodes of the graph"""
+        """Compute the deterministic nodes of the graph, **not** including var itself."""
         deterministics = []
         attrs = ('transformed', 'logpt')
         for v in self.var_list:
             if v != var and all(not hasattr(v, attr) for attr in attrs):
                 deterministics.append(v)
         return deterministics
 
-    def _ancestors(self, var, func, blockers=None):
-        """Get ancestors of a function that are also named PyMC3 variables"""
-        return set([j for j in ancestors([func], blockers=blockers) if j in self.var_list and j != var])
+    def _get_ancestors(self, var, func) -> MutableSet[RV]:
+        """Get all ancestors of a function, doing some accounting for deterministics.
+        """
 
-    def _get_ancestors(self, var, func):
-        """Get all ancestors of a function, doing some accounting for deterministics
+        # this contains all of the variables in the model EXCEPT var...
+        vars: MutableSet[RV] = set(self.var_list)
+        vars.remove(var)
+
+        blockers: MutableSet[RV] = set()
+        retval = set()
+        def _expand(node) -> Optional[Iterator[Tensor]]:
+            if node in blockers:
+                return None
+            elif node in vars:
+                blockers.add(node)
+                retval.add(node)
+                return None
+            elif node.owner:
+                blockers.add(node)
+                return reversed(node.owner.inputs)
+            else:
+                return None
 
-        Specifically, if a deterministic is an input, theano.gof.graph.ancestors will
-        return only the inputs *to the deterministic*.  However, if we pass in the
-        deterministic as a blocker, it will skip those nodes.
-        """
-        deterministics = self.get_deterministics(var)
-        upstream = self._ancestors(var, func)
-
-        # Usual case
-        if upstream == self._ancestors(var, func, blockers=upstream):
-            return upstream
-        else: # deterministic accounting
-            for d in powerset(upstream):
-                blocked = self._ancestors(var, func, blockers=d)
-                if set(d) == blocked:
-                    return d
-        raise RuntimeError('Could not traverse graph. Consider raising an issue with developers.')
+        stack_search(start = deque([func]),
+                     expand=_expand,
+                     mode='bfs')
+        return retval
 
     def _filter_parents(self, var, parents):
         """Get direct parents of a var, as strings"""