Upped the code coverage

jmafoster1 · jmafoster1 · commit 73f360c3d0fb · 2023-03-14T09:54:13.000Z
diff --git a/causal_testing/testing/estimators.py b/causal_testing/testing/estimators.py
@@ -48,7 +48,7 @@ def __init__(
         adjustment_set: set,
         outcome: str,
         df: pd.DataFrame = None,
-        effect_modifiers: dict[Variable:Any] = None,
+        effect_modifiers: dict[str:Any] = None,
     ):
         self.treatment = treatment
         self.treatment_value = treatment_value
@@ -59,9 +59,9 @@ def __init__(
         if effect_modifiers is None:
             self.effect_modifiers = {}
         elif isinstance(effect_modifiers, (list, set)):
-            self.effect_modifiers = {k.name for k in effect_modifiers}
+            self.effect_modifiers = {k for k in effect_modifiers}
         elif isinstance(effect_modifiers, dict):
-            self.effect_modifiers = {k.name: v for k, v in effect_modifiers.items()}
+            self.effect_modifiers = {k: v for k, v in effect_modifiers.items()}
         else:
             raise ValueError(f"Unsupported type for effect_modifiers {effect_modifiers}. Expected iterable")
         self.modelling_assumptions = []
@@ -106,19 +106,17 @@ def __init__(
         adjustment_set: set,
         outcome: str,
         df: pd.DataFrame = None,
-        effect_modifiers: dict[Variable:Any] = None,
+        effect_modifiers: dict[str:Any] = None,
         formula: str = None,
     ):
         super().__init__(treatment, treatment_value, control_value, adjustment_set, outcome, df, effect_modifiers)
 
         self.model = None
-        if effect_modifiers is None:
-            effect_modifiers = []
 
         if formula is not None:
             self.formula = formula
         else:
-            terms = [treatment] + sorted(list(adjustment_set)) + sorted(list(effect_modifiers))
+            terms = [treatment] + sorted(list(adjustment_set)) + sorted(list(self.effect_modifiers))
             self.formula = f"{outcome} ~ {'+'.join(((terms)))}"
 
         for term in self.effect_modifiers:
@@ -170,7 +168,6 @@ def estimate(self, data: pd.DataFrame, adjustment_config=None) -> RegressionResu
         :param data: A pandas dataframe containing execution data from the system-under-test.
 
         """
-        print(data)
         if adjustment_config is None:
             adjustment_config = {}
 
@@ -211,6 +208,11 @@ def estimate_control_treatment(self, bootstrap_size=100) -> tuple[pd.Series, pd.
                 "a small dataset."
             )
             return (y.iloc[1], None), (y.iloc[0], None)
+        except np.linalg.LinAlgError:
+            logger.warning(
+                "Singular matrix detected. Confidence intervals not available. Try with a larger data set"
+            )
+            return (y.iloc[1], None), (y.iloc[0], None)
 
         # Delta method confidence intervals from
         # https://stackoverflow.com/questions/47414842/confidence-interval-of-probability-prediction-from-logistic-regression-statsmode
@@ -272,8 +274,6 @@ def estimate_risk_ratio(self, bootstrap_size=100) -> float:
 
         bootstraps = sorted(list(treatment_bootstraps / control_bootstraps))
         bound = ceil((bootstrap_size * 0.05) / 2)
-        print("bootstraps", bootstraps)
-        print("bound", bound)
         ci_low = bootstraps[bound]
         ci_high = bootstraps[bootstrap_size - bound]
 
diff --git a/tests/testing_tests/test_causal_test_engine.py b/tests/testing_tests/test_causal_test_engine.py
@@ -240,7 +240,7 @@ def test_execute_observational_causal_forest_estimator_cates(self):
             self.minimal_adjustment_set,
             "C",
             self.causal_test_engine.scenario_execution_data_df,
-            effect_modifiers={Input("M", int): None},
+            effect_modifiers={"M": None},
         )
         causal_test_result = self.causal_test_engine.execute_test(
             estimation_model, self.causal_test_case, estimate_type="cate"
diff --git a/tests/testing_tests/test_causal_test_outcome.py b/tests/testing_tests/test_causal_test_outcome.py
@@ -1,5 +1,5 @@
 import unittest
-from causal_testing.testing.causal_test_outcome import ExactValue, SomeEffect
+from causal_testing.testing.causal_test_outcome import ExactValue, SomeEffect, Positive, Negative
 from causal_testing.testing.causal_test_result import CausalTestResult, TestValue
 from causal_testing.testing.estimators import LinearRegressionEstimator
 
@@ -63,6 +63,50 @@ def test_empty_adjustment_set(self):
             ),
         )
 
+    def test_Positive_pass(self):
+        test_value = TestValue(type="ate", value=5.05)
+        ctr = CausalTestResult(
+            estimator=self.estimator,
+            test_value=test_value,
+            confidence_intervals=None,
+            effect_modifier_configuration=None,
+        )
+        ev = Positive()
+        self.assertTrue(ev.apply(ctr))
+
+    def test_Positive_fail(self):
+        test_value = TestValue(type="ate", value=0)
+        ctr = CausalTestResult(
+            estimator=self.estimator,
+            test_value=test_value,
+            confidence_intervals=None,
+            effect_modifier_configuration=None,
+        )
+        ev = Positive()
+        self.assertFalse(ev.apply(ctr))
+
+    def test_Negative_pass(self):
+        test_value = TestValue(type="ate", value=-5.05)
+        ctr = CausalTestResult(
+            estimator=self.estimator,
+            test_value=test_value,
+            confidence_intervals=None,
+            effect_modifier_configuration=None,
+        )
+        ev = Negative()
+        self.assertTrue(ev.apply(ctr))
+
+    def test_Negative_fail(self):
+        test_value = TestValue(type="ate", value=0)
+        ctr = CausalTestResult(
+            estimator=self.estimator,
+            test_value=test_value,
+            confidence_intervals=None,
+            effect_modifier_configuration=None,
+        )
+        ev = Negative()
+        self.assertFalse(ev.apply(ctr))
+
     def test_exactValue_pass(self):
         test_value = TestValue(type="ate", value=5.05)
         ctr = CausalTestResult(
diff --git a/tests/testing_tests/test_estimators.py b/tests/testing_tests/test_estimators.py
@@ -76,40 +76,44 @@ class TestLogisticRegressionEstimator(unittest.TestCase):
 
     @classmethod
     def setUpClass(cls) -> None:
-        cls.scarf_df = pd.DataFrame(
-            [
-                {"length_in": 55, "completed": 1},
-                {"length_in": 55, "completed": 1},
-                {"length_in": 55, "completed": 1},
-                {"length_in": 60, "completed": 1},
-                {"length_in": 60, "completed": 0},
-                {"length_in": 70, "completed": 1},
-                {"length_in": 70, "completed": 0},
-                {"length_in": 82, "completed": 1},
-                {"length_in": 82, "completed": 0},
-                {"length_in": 82, "completed": 0},
-                {"length_in": 82, "completed": 0},
-            ]
-        )
+        cls.scarf_df = pd.DataFrame([
+            { 'length_in': 55, 'large_gauge': 1, 'color': 'orange', 'completed': 1 },
+            { 'length_in': 55, 'large_gauge': 0, 'color': 'orange', 'completed': 1 },
+            { 'length_in': 55, 'large_gauge': 0, 'color': 'brown', 'completed': 1 },
+            { 'length_in': 60, 'large_gauge': 0, 'color': 'brown', 'completed': 1 },
+            { 'length_in': 60, 'large_gauge': 0, 'color': 'grey', 'completed': 0 },
+            { 'length_in': 70, 'large_gauge': 0, 'color': 'grey', 'completed': 1 },
+            { 'length_in': 70, 'large_gauge': 0, 'color': 'orange', 'completed': 0 },
+            { 'length_in': 82, 'large_gauge': 1, 'color': 'grey', 'completed': 1 },
+            { 'length_in': 82, 'large_gauge': 0, 'color': 'brown', 'completed': 0 },
+            { 'length_in': 82, 'large_gauge': 0, 'color': 'orange', 'completed': 0 },
+            { 'length_in': 82, 'large_gauge': 1, 'color': 'brown', 'completed': 0 },
+        ])
 
     def test_ate(self):
-        df = self.scarf_df
+        df = self.scarf_df.copy()
         logistic_regression_estimator = LogisticRegressionEstimator("length_in", 65, 55, set(), "completed", df)
         ate, _ = logistic_regression_estimator.estimate_ate()
         self.assertEqual(round(ate, 4), -0.1987)
 
     def test_risk_ratio(self):
-        df = self.scarf_df
+        df = self.scarf_df.copy()
         logistic_regression_estimator = LogisticRegressionEstimator("length_in", 65, 55, set(), "completed", df)
         rr, _ = logistic_regression_estimator.estimate_risk_ratio()
         self.assertEqual(round(rr, 4), 0.7664)
 
     def test_odds_ratio(self):
-        df = self.scarf_df
+        df = self.scarf_df.copy()
         logistic_regression_estimator = LogisticRegressionEstimator("length_in", 65, 55, set(), "completed", df)
         odds = logistic_regression_estimator.estimate_unit_odds_ratio()
         self.assertEqual(round(odds, 4), 0.8948)
 
+    def test_ate_effect_modifiers(self):
+        df = self.scarf_df.copy()
+        logistic_regression_estimator = LogisticRegressionEstimator("length_in", 65, 55, set(), "completed", df, effect_modifiers={"large_gauge": 0})
+        ate, _ = logistic_regression_estimator.estimate_ate()
+        self.assertEqual(round(ate, 4), -0.3388)
+
 
 class TestInstrumentalVariableEstimator(unittest.TestCase):
     """
@@ -368,7 +372,7 @@ def test_program_15_ate(self):
             "smokeyrs",
         }
         causal_forest = CausalForestEstimator(
-            "qsmk", 1, 0, covariates, "wt82_71", df, {Input("smokeintensity", int): 40}
+            "qsmk", 1, 0, covariates, "wt82_71", df, {"smokeintensity": 40}
         )
         ate, _ = causal_forest.estimate_ate()
         self.assertGreater(round(ate, 1), 2.5)
@@ -395,7 +399,7 @@ def test_program_15_cate(self):
             "smokeyrs",
         }
         causal_forest = CausalForestEstimator(
-            "qsmk", 1, 0, covariates, "wt82_71", smoking_intensity_5_and_40_df, {Input("smokeintensity", int): 40}
+            "qsmk", 1, 0, covariates, "wt82_71", smoking_intensity_5_and_40_df, {"smokeintensity": 40}
         )
         cates_df, _ = causal_forest.estimate_cates()
         self.assertGreater(cates_df["cate"].mean(), 0)
@@ -415,7 +419,7 @@ def test_X1_effect(self):
         """When we fix the value of X2 to 0, the effect of X1 on Y should become ~2 (because X2 terms are cancelled)."""
         x2 = Input("X2", float)
         lr_model = LinearRegressionEstimator(
-            "X1", 1, 0, {"X2"}, "Y", effect_modifiers={x2: 0}, formula="Y ~ X1 + X2 + (X1 * X2)", df=self.df
+            "X1", 1, 0, {"X2"}, "Y", effect_modifiers={x2.name: 0}, formula="Y ~ X1 + X2 + (X1 * X2)", df=self.df
         )
         test_results = lr_model.estimate_ate()
         ate = test_results[0]

Original file line number	Diff line number	Diff line change
`@@ -240,7 +240,7 @@ def test_execute_observational_causal_forest_estimator_cates(self):`
`240`	`240`	`self.minimal_adjustment_set,`
`241`	`241`	`"C",`
`242`	`242`	`self.causal_test_engine.scenario_execution_data_df,`
`243`		`- effect_modifiers={Input("M", int): None},`
	`243`	`+ effect_modifiers={"M": None},`
`244`	`244`	`)`
`245`	`245`	`causal_test_result = self.causal_test_engine.execute_test(`
`246`	`246`	`estimation_model, self.causal_test_case, estimate_type="cate"`