Remove causal specification from execute_test

Author: christopher-wild

causal_testing/json_front/json_class.py

@@ -265,8 +265,7 @@ def _execute_test_case(
 
         estimation_model = self._setup_test(causal_test_case=causal_test_case, test=test)
         causal_test_result = causal_test_case.execute_test(estimator=estimation_model,
-                                                           data_collector=self.data_collector,
-                                                           causal_specification=self.causal_specification)
+                                                           data_collector=self.data_collector)
 
         test_passes = causal_test_case.expected_causal_effect.apply(causal_test_result)
 
@@ -310,7 +309,7 @@ def _setup_test(
             "control_value": causal_test_case.control_value,
             "adjustment_set": minimal_adjustment_set,
             "outcome": causal_test_case.outcome_variable.name,
-            "df": self.data_collector.data,
+            "df": self.data_collector.collect_data(),
             "effect_modifiers": causal_test_case.effect_modifier_configuration,
             "alpha": test["alpha"] if "alpha" in test else 0.05,
         }

causal_testing/testing/causal_test_suite.py

@@ -77,7 +77,7 @@ def execute_test_suite(self, data_collector: ObservationalDataCollector,
                         test.outcome_variable.name,
                     )
                     if estimator.df is None:
-                        estimator.df = data_collector.data
+                        estimator.df = data_collector.collect_data()
                     causal_test_result = test._return_causal_test_results(estimator)
                     causal_test_results.append(causal_test_result)
 

examples/covasim_/doubling_beta/example_beta.py

@@ -65,8 +65,7 @@ def doubling_beta_CATE_on_csv(
 
     # Add squared terms for beta, since it has a quadratic relationship with cumulative infections
     causal_test_result = causal_test_case.execute_test(estimator=linear_regression_estimator,
-                                                       data_collector=data_collector,
-                                                       causal_specification=causal_specification)
+                                                       data_collector=data_collector)
 
     # Repeat for association estimate (no adjustment)
     no_adjustment_linear_regression_estimator = LinearRegressionEstimator(
@@ -79,8 +78,7 @@ def doubling_beta_CATE_on_csv(
         formula="cum_infections ~ beta + np.power(beta, 2)",
     )
     association_test_result = causal_test_case.execute_test(estimator=no_adjustment_linear_regression_estimator,
-                                                            data_collector=data_collector,
-                                                            causal_specification=causal_specification)
+                                                            data_collector=data_collector)
 
     # Store results for plotting
     results_dict["association"] = {
@@ -111,9 +109,8 @@ def doubling_beta_CATE_on_csv(
             formula="cum_infections ~ beta + np.power(beta, 2) + avg_age + contacts",
         )
         counterfactual_causal_test_result = causal_test_case.execute_test(
-            estimator=linear_regression_estimator, data_collector=data_collector,
-            causal_specification=causal_specification
-        )
+            estimator=linear_regression_estimator, data_collector=data_collector)
+
         results_dict["counterfactual"] = {
             "ate": counterfactual_causal_test_result.test_value.value,
             "cis": counterfactual_causal_test_result.confidence_intervals,

examples/covasim_/vaccinating_elderly/example_vaccine.py

@@ -95,8 +95,7 @@ def test_experimental_vaccinate_elderly(runs_per_test_per_config: int = 30, verb
         )
 
         # 9. Execute test and save results in dict
-        causal_test_result = causal_test_case.execute_test(linear_regression_estimator, data_collector,
-                                                           causal_specification)
+        causal_test_result = causal_test_case.execute_test(linear_regression_estimator, data_collector)
         if verbose:
             logging.info("Causation:\n%s", causal_test_result)
         results_dict[outcome_variable.name]["ate"] = causal_test_result.test_value.value

examples/lr91/example_max_conductances.py

@@ -144,7 +144,7 @@ def effects_on_APD90(observational_data_path, treatment_var, control_val, treatm
     )
 
     # 9. Run the causal test and print results
-    causal_test_result = causal_test_case.execute_test(linear_regression_estimator, data_collector, causal_specification)
+    causal_test_result = causal_test_case.execute_test(linear_regression_estimator, data_collector)
     logger.info("%s", causal_test_result)
     return causal_test_result.test_value.value, causal_test_result.confidence_intervals
 

tests/testing_tests/test_causal_test_case.py

@@ -108,7 +108,7 @@ def setUp(self) -> None:
         # Obsolete?
         self.data_collector = ObservationalDataCollector(self.scenario, df)
         self.data_collector.collect_data()
-        self.df = self.data_collector.data
+        self.df = self.data_collector.collect_data()
         self.minimal_adjustment_set = self.causal_dag.identification(self.base_test_case)
         # 6. Easier to access treatment and outcome values
         self.treatment_value = 1
@@ -130,7 +130,7 @@ def test_execute_test_observational_causal_forest_estimator(self):
             "C",
             self.df,
         )
-        causal_test_result = self.causal_test_case.execute_test(estimation_model, self.data_collector, self.causal_specification)
+        causal_test_result = self.causal_test_case.execute_test(estimation_model, self.data_collector)
         self.assertAlmostEqual(causal_test_result.test_value.value, 4, delta=1)
 
     def test_invalid_causal_effect(self):
@@ -152,7 +152,7 @@ def test_execute_test_observational_linear_regression_estimator(self):
             "C",
             self.df,
         )
-        causal_test_result = self.causal_test_case.execute_test(estimation_model, self.data_collector, self.causal_specification)
+        causal_test_result = self.causal_test_case.execute_test(estimation_model, self.data_collector)
         self.assertAlmostEqual(causal_test_result.test_value.value, 4, delta=1e-10)
 
     def test_execute_test_observational_linear_regression_estimator_direct_effect(self):
@@ -179,7 +179,7 @@ def test_execute_test_observational_linear_regression_estimator_direct_effect(se
             "C",
             self.df,
         )
-        causal_test_result = causal_test_case.execute_test(estimation_model, self.data_collector, self.causal_specification)
+        causal_test_result = causal_test_case.execute_test(estimation_model, self.data_collector)
         self.assertAlmostEqual(causal_test_result.test_value.value, 4, delta=1e-10)
 
     def test_execute_test_observational_linear_regression_estimator_coefficient(self):
@@ -194,7 +194,7 @@ def test_execute_test_observational_linear_regression_estimator_coefficient(self
             self.df,
         )
         self.causal_test_case.estimate_type = "coefficient"
-        causal_test_result = self.causal_test_case.execute_test(estimation_model, self.data_collector, self.causal_specification)
+        causal_test_result = self.causal_test_case.execute_test(estimation_model, self.data_collector)
         self.assertEqual(int(causal_test_result.test_value.value), 0)
 
     def test_execute_test_observational_linear_regression_estimator_risk_ratio(self):
@@ -209,7 +209,7 @@ def test_execute_test_observational_linear_regression_estimator_risk_ratio(self)
             self.df,
         )
         self.causal_test_case.estimate_type = "risk_ratio"
-        causal_test_result = self.causal_test_case.execute_test(estimation_model, self.data_collector, self.causal_specification)
+        causal_test_result = self.causal_test_case.execute_test(estimation_model, self.data_collector)
         self.assertEqual(int(causal_test_result.test_value.value), 0)
 
     def test_invalid_estimate_type(self):
@@ -225,7 +225,7 @@ def test_invalid_estimate_type(self):
         )
         self.causal_test_case.estimate_type = "invalid"
         with self.assertRaises(AttributeError):
-            self.causal_test_case.execute_test(estimation_model, self.data_collector, self.causal_specification)
+            self.causal_test_case.execute_test(estimation_model, self.data_collector)
 
     def test_execute_test_observational_linear_regression_estimator_squared_term(self):
         """Check that executing the causal test case returns the correct results for dummy data with a squared term
@@ -239,7 +239,7 @@ def test_execute_test_observational_linear_regression_estimator_squared_term(sel
             self.df,
             formula=f"C ~ A + {'+'.join(self.minimal_adjustment_set)} + (D ** 2)",
         )
-        causal_test_result = self.causal_test_case.execute_test(estimation_model, self.data_collector, self.causal_specification)
+        causal_test_result = self.causal_test_case.execute_test(estimation_model, self.data_collector)
         self.assertAlmostEqual(round(causal_test_result.test_value.value, 1), 4, delta=1)
 
     def test_execute_observational_causal_forest_estimator_cates(self):
@@ -262,7 +262,7 @@ def test_execute_observational_causal_forest_estimator_cates(self):
             effect_modifiers={"M": None},
         )
         self.causal_test_case.estimate_type = "cates"
-        causal_test_result = self.causal_test_case.execute_test(estimation_model, self.data_collector, self.causal_specification)
+        causal_test_result = self.causal_test_case.execute_test(estimation_model, self.data_collector)
         causal_test_result = causal_test_result.test_value.value
         # Check that each effect modifier's strata has a greater ATE than the last (ascending order)
         causal_test_result_m1 = causal_test_result.loc[causal_test_result["M"] == 1]