Add compatibility between GCM and CausalModel class

Before, the causal graphs between the GCM and CausalModel part required some additional work to be converted.
Now, a GCM can be used to initiate a CausalModel or CausalGraph and vice versa.

Signed-off-by: Patrick Bloebaum <bloebp@amazon.com>

dowhy/causal_graph.py

@@ -4,6 +4,7 @@
 
 import networkx as nx
 
+from dowhy.gcm.causal_models import ProbabilisticCausalModel
 from dowhy.utils.api import parse_state
 from dowhy.utils.graph_operations import daggity_to_dot
 from dowhy.utils.plotting import plot
@@ -54,6 +55,10 @@ def __init__(
         if graph is None:
             self._graph = nx.DiGraph()
             self._graph = self.build_graph(common_cause_names, instrument_names, effect_modifier_names, mediator_names)
+        elif isinstance(graph, nx.DiGraph):
+            self._graph = nx.DiGraph(graph)
+        elif isinstance(graph, ProbabilisticCausalModel):
+            self._graph = nx.DiGraph(graph.graph)
         elif re.match(r".*\.dot", graph):
             # load dot file
             try:
@@ -90,9 +95,18 @@ def __init__(
         elif re.match(".*graph\s*\[.*\]\s*", graph):
             self._graph = nx.DiGraph(nx.parse_gml(graph))
         else:
-            self.logger.error("Error: Please provide graph (as string or text file) in dot or gml format.")
+            self.logger.error(
+                "Error: Please provide graph (as string or text file) in dot, gml format, networkx graph "
+                "or GCM model."
+            )
             self.logger.error("Error: Incorrect graph format")
             raise ValueError
+
+        if observed_node_names is None and (
+            isinstance(graph, nx.DiGraph) or isinstance(graph, ProbabilisticCausalModel)
+        ):
+            observed_node_names = list(self._graph.nodes)
+
         if missing_nodes_as_confounders:
             self._graph = self.add_missing_nodes_as_common_causes(observed_node_names)
         # Adding node attributes

dowhy/causal_model.py

@@ -11,7 +11,6 @@
 import dowhy.causal_estimators as causal_estimators
 import dowhy.causal_refuters as causal_refuters
 import dowhy.graph_learners as graph_learners
-import dowhy.utils.cli_helpers as cli
 from dowhy.causal_estimator import CausalEstimate, estimate_effect
 from dowhy.causal_graph import CausalGraph
 from dowhy.causal_identifier import AutoIdentifier, BackdoorAdjustment, IDIdentifier
@@ -23,7 +22,6 @@
 
 
 class CausalModel:
-
     """Main class for storing the causal model state."""
 
     def __init__(

dowhy/gcm/causal_models.py

@@ -39,8 +39,17 @@ def __init__(
         :param graph_copier: Optional function that can copy a causal graph. Defaults to a networkx.DiGraph
                              constructor.
         """
+        # Todo: Remove after https://github.com/py-why/dowhy/pull/943.
+        from dowhy.causal_graph import CausalGraph
+        from dowhy.causal_model import CausalModel
+
         if graph is None:
             graph = nx.DiGraph()
+        elif isinstance(graph, CausalModel):
+            graph = graph_copier(graph._graph._graph)
+        elif isinstance(graph, CausalGraph):
+            graph = graph_copier(graph._graph)
+
         self.graph = graph
         self.graph_copier = graph_copier
 

tests/test_causal_model.py

@@ -1,3 +1,4 @@
+import networkx as nx
 import pandas as pd
 import pytest
 from flaky import flaky
@@ -7,6 +8,8 @@
 import dowhy
 import dowhy.datasets
 from dowhy import CausalModel
+from dowhy.causal_graph import CausalGraph
+from dowhy.gcm import ProbabilisticCausalModel, StructuralCausalModel
 
 
 class TestCausalModel(object):
@@ -498,6 +501,30 @@ def test_unobserved_graph_variables_log_warning(self, caplog):
             f"Only the following log records were emitted instead: '{caplog.records}'."
         )
 
+    def test_compability_with_gcm(self):
+        data = pd.DataFrame({"X": [0], "Y": [0], "Z": [0]})
+        model = CausalModel(
+            data=data,
+            treatment="Y",
+            outcome="Z",
+            graph=StructuralCausalModel(nx.DiGraph([("X", "Y"), ("Y", "Z")])),
+        )
+
+        assert set(model._graph._graph.nodes) == {"X", "Y", "Z"}
+        assert set(model._graph._graph.edges) == {("X", "Y"), ("Y", "Z")}
+
+        causal_graph = CausalGraph("Y", "Z", graph=StructuralCausalModel(nx.DiGraph([("X", "Y"), ("Y", "Z")])))
+        assert set(causal_graph._graph.nodes) == {"X", "Y", "Z"}
+        assert set(causal_graph._graph.edges) == {("X", "Y"), ("Y", "Z")}
+
+        pcm = ProbabilisticCausalModel(model)
+        assert set(pcm.graph.nodes) == {"X", "Y", "Z"}
+        assert set(pcm.graph.edges) == {("X", "Y"), ("Y", "Z")}
+
+        pcm = ProbabilisticCausalModel(model._graph)
+        assert set(pcm.graph.nodes) == {"X", "Y", "Z"}
+        assert set(pcm.graph.edges) == {("X", "Y"), ("Y", "Z")}
+
 
 if __name__ == "__main__":
     pytest.main([__file__])