Update validation to handle boolean errors (#30)

sliceline/slicefinder.py

@@ -147,7 +147,7 @@ def fit(self, X, errors):
             self.min_sup = int(self.min_sup * len(X))
 
         # Check that X and e have correct shape
-        X_array, errors = check_X_e(X, errors)
+        X_array, errors = check_X_e(X, errors, y_numeric=True)
 
         self._check_feature_names(X, reset=True)
 

sliceline/validation.py

@@ -805,8 +805,8 @@ def _check_y(y, multi_output=False, y_numeric=False, estimator=None):
         y = column_or_1d(y, warn=True)
         _assert_all_finite(y, input_name="y", estimator_name=estimator_name)
         _ensure_no_complex_data(y)
-    if y_numeric and y.dtype.kind == "O":
-        y = y.astype(np.float64)
+    if y_numeric and y.dtype.kind in ("O", "b"):
+        y = y.astype(np.float32)
 
     return y
 

tests/conftest.py

@@ -111,7 +111,7 @@ def basic_test_data():
 @pytest.fixture
 def experiments():
     """Implement several end-to-end examples with the expected outputs regarding the inputs."""
-    # Experiment 1: basic case
+    # Experiment 1: basic case with boolean errors
     np.random.seed(1)
     n_small = 10
     X_1 = np.array(
@@ -121,7 +121,7 @@ def experiments():
             np.random.randint(1, 4, size=2 * n_small),
         ]
     ).T
-    errors_1 = np.array([1] * n_small + [0] * n_small)
+    errors_1 = np.array([True] * n_small + [False] * n_small)
     expected_top_slices_1 = np.array([[1, 1, None], [None, 1, 2], [1, None, 2]])
     experiment_1 = Experiment(X_1, errors_1, expected_top_slices_1)
 
@@ -139,7 +139,7 @@ def experiments():
     ).T
     errors_2 = np.array([1] * n_small + [0] * n_small)
     expected_top_slices_2 = np.array(
-        [[None, 1.0, None, None, 1.0, None], [None, None, 4.0, None, 1.0, None]]
+        [[None, 1, None, None, 1, None], [None, None, 4, None, 1, None]]
     )
     experiment_2 = Experiment(X_2, errors_2, expected_top_slices_2)