causal-testing-framework 14.3.0__tar.gz → 14.3.1__tar.gz

{causal_testing_framework-14.3.0 → causal_testing_framework-14.3.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: causal_testing_framework
-Version: 14.3.0
+Version: 14.3.1
 Summary: A framework for causal testing using causal directed acyclic graphs.
 Author: The CITCOM team
 License: MIT

{causal_testing_framework-14.3.0 → causal_testing_framework-14.3.1}/causal_testing/_version.py

@@ -18,7 +18,7 @@ version_tuple: tuple[int | str, ...]
 commit_id: str | None
 __commit_id__: str | None
 
-__version__ = version = '14.3.0'
-__version_tuple__ = version_tuple = (14, 3, 0)
+__version__ = version = '14.3.1'
+__version_tuple__ = version_tuple = (14, 3, 1)
 
-__commit_id__ = commit_id = 'ge6a19c77b'
+__commit_id__ = commit_id = 'g3b5b62ad2'

{causal_testing_framework-14.3.0 → causal_testing_framework-14.3.1}/causal_testing/discovery/abstract_discovery.py

@@ -10,6 +10,7 @@ from enum import Enum
 from itertools import permutations
 
 import networkx as nx
+import numpy as np
 import pandas as pd
 import rustworkx as rx
 
@@ -207,28 +208,27 @@ class Discovery(ABC):
 
         results = []
 
-        # We use "silent=False" here to allow for inestimable edges, but it'd be good to have a more stringent
-        # error catching strategy to catch "genuine" problems (e.g. to do with the structure of the data)
-        for test_case, result in zip(ctf.test_cases, ctf.run_tests(silent=False)):
-            if result.effect_estimate is None:
+        for test_case, result in zip(ctf.test_cases, ctf.test_cases):
+            try:
+                result = test_case.execute_test()
                 results.append(
                     {
-                        "result": TestResult.INESTIMABLE,
+                        "result": (
+                            TestResult.PASS if test_case.expected_causal_effect.apply(result) else TestResult.FAIL
+                        ),
                         "expected_effect": test_case.expected_causal_effect.__class__.__name__,
                         "treatment": test_case.base_test_case.treatment_variable.name,
                         "outcome": test_case.base_test_case.outcome_variable.name,
+                        "effect": effect_direction(result),
                     }
                 )
-            else:
+            except np.linalg.LinAlgError:
                 results.append(
                     {
-                        "result": (
-                            TestResult.PASS if test_case.expected_causal_effect.apply(result) else TestResult.FAIL
-                        ),
+                        "result": TestResult.INESTIMABLE,
                         "expected_effect": test_case.expected_causal_effect.__class__.__name__,
                         "treatment": test_case.base_test_case.treatment_variable.name,
                         "outcome": test_case.base_test_case.outcome_variable.name,
-                        "effect": effect_direction(result),
                     }
                 )
 

{causal_testing_framework-14.3.0 → causal_testing_framework-14.3.1}/causal_testing_framework.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: causal_testing_framework
-Version: 14.3.0
+Version: 14.3.1
 Summary: A framework for causal testing using causal directed acyclic graphs.
 Author: The CITCOM team
 License: MIT

causal_testing_framework-14.3.1/causal_testing_framework.egg-info/scm_version.json

@@ -0,0 +1,8 @@
+{
+  "tag": "14.3.1",
+  "distance": 0,
+  "node": "g3b5b62ad20e1ae57075b5e5bf16ac1e2cc154317",
+  "dirty": false,
+  "branch": "HEAD",
+  "node_date": "2026-07-13"
+}

{causal_testing_framework-14.3.0 → causal_testing_framework-14.3.1}/tests/discovery_tests/test_abstract_discovery.py

@@ -176,6 +176,105 @@ class TestAbstractHillClimber(unittest.TestCase):
         with self.assertRaises(ValueError):
             abstract_discovery.write_dot(dag, "dag.dot")
 
+    def test_evaluate_tests_invalid_datatype(self):
+        scarf_df = pd.read_csv("tests/resources/data/scarf_data.csv")
+        scarf_df["completed"] = pd.to_datetime(["2026-01-01" for _ in range(len(scarf_df))], format="%Y-%m-%d")
+
+        dag = CausalDAG()
+        dag.add_nodes_from(scarf_df.columns)
+        dag.add_edges_from([("length_in", "completed"), ("large_gauge", "completed")])
+
+        abstract_discovery = AbstractDiscovery(scarf_df)
+        with self.assertRaises(ValueError):
+            abstract_discovery.evaluate_tests(dag)
+
+    def test_evaluate_tests_inestimable(self):
+        scarf_df = pd.read_csv("tests/resources/data/scarf_data.csv")
+        scarf_df["completed"] = scarf_df["completed"].astype(bool)
+        scarf_df = scarf_df.loc[(scarf_df["completed"]) & (scarf_df["color"] != "grey")]
+
+        dag = CausalDAG()
+        dag.add_nodes_from(scarf_df.columns)
+        dag.add_edges_from([("length_in", "completed"), ("large_gauge", "completed")])
+
+        abstract_discovery = AbstractDiscovery(scarf_df)
+        test_results = abstract_discovery.evaluate_tests(dag)
+        expected_results = pd.DataFrame(
+            [
+                {
+                    "result": TestResult.PASS,
+                    "expected_effect": "NoEffect",
+                    "treatment": "length_in",
+                    "outcome": "large_gauge",
+                    "effect": "negative",
+                },
+                {
+                    "result": TestResult.PASS,
+                    "expected_effect": "NoEffect",
+                    "treatment": "large_gauge",
+                    "outcome": "length_in",
+                    "effect": "negative",
+                },
+                {
+                    "result": TestResult.PASS,
+                    "expected_effect": "NoEffect",
+                    "treatment": "length_in",
+                    "outcome": "color",
+                    "effect": None,
+                },
+                {
+                    "result": TestResult.PASS,
+                    "expected_effect": "NoEffect",
+                    "treatment": "color",
+                    "outcome": "length_in",
+                    "effect": None,
+                },
+                {
+                    "result": TestResult.FAIL,
+                    "expected_effect": "SomeEffect",
+                    "treatment": "length_in",
+                    "outcome": "completed",
+                    "effect": "positive",
+                },
+                {
+                    "result": TestResult.PASS,
+                    "expected_effect": "NoEffect",
+                    "treatment": "large_gauge",
+                    "outcome": "color",
+                    "effect": None,
+                },
+                {
+                    "result": TestResult.PASS,
+                    "expected_effect": "NoEffect",
+                    "treatment": "color",
+                    "outcome": "large_gauge",
+                    "effect": None,
+                },
+                {
+                    "result": TestResult.FAIL,
+                    "expected_effect": "SomeEffect",
+                    "treatment": "large_gauge",
+                    "outcome": "completed",
+                    "effect": "negative",
+                },
+                {
+                    "result": TestResult.INESTIMABLE,
+                    "expected_effect": "NoEffect",
+                    "treatment": "color",
+                    "outcome": "completed",
+                    "effect": None,
+                },
+                {
+                    "result": TestResult.INESTIMABLE,
+                    "expected_effect": "NoEffect",
+                    "treatment": "completed",
+                    "outcome": "color",
+                    "effect": None,
+                },
+            ]
+        )
+        pd.testing.assert_frame_equal(test_results, expected_results)
+
     def test_evaluate_tests(self):
         scarf_df = pd.read_csv("tests/resources/data/scarf_data.csv")
 

causal_testing_framework-14.3.0/causal_testing_framework.egg-info/scm_version.json

@@ -1,8 +0,0 @@
-{
-  "tag": "14.3.0",
-  "distance": 0,
-  "node": "ge6a19c77bff74b8bd610c83493609841e058d0d6",
-  "dirty": false,
-  "branch": "HEAD",
-  "node_date": "2026-07-10"
-}