Move functions to test modules to _utils.py

skrub/_dataframe/_namespace.py

@@ -43,19 +43,6 @@ def is_polars(dataframe):
     return isinstance(dataframe, (pl.DataFrame, pl.LazyFrame))
 
 
-def is_namespace_pandas(px):
-    return px is pd
-
-
-def is_namespace_polars(px):
-    if "polars" not in sys.modules:
-        return False
-
-    import polars as pl
-
-    return px is pl
-
-
 def get_df_namespace(*dfs):
     """Get the namespaces of dataframes.
 

skrub/_utils.py

@@ -162,3 +162,11 @@ def atleast_2d_or_none(x):
 
 def clone_if_default(estimator, default_estimator):
     return clone(estimator) if estimator is default_estimator else estimator
+
+
+def is_namespace_pandas(px):
+    return px.__name__ == "pandas"
+
+
+def is_namespace_polars(px):
+    return px.__name__ == "polars"

skrub/tests/test_datetime_encoder.py

@@ -7,14 +7,14 @@
 from numpy.testing import assert_allclose, assert_array_equal
 from pandas.testing import assert_frame_equal
 
-from skrub._dataframe._namespace import is_namespace_polars
 from skrub._dataframe._polars import POLARS_SETUP
 from skrub._datetime_encoder import (
     TIME_LEVELS,
     DatetimeEncoder,
     _is_pandas_format_mixed_available,
     to_datetime,
 )
+from skrub._utils import is_namespace_polars
 
 MODULES = [pd]
 ASSERT_TUPLES = [(pd, assert_frame_equal)]

skrub/tests/test_fuzzy_join.py

@@ -9,8 +9,8 @@
 from sklearn.feature_extraction.text import HashingVectorizer
 
 from skrub import fuzzy_join
-from skrub._dataframe._namespace import is_namespace_polars
 from skrub._dataframe._polars import POLARS_SETUP
+from skrub._utils import is_namespace_polars
 
 MODULES = [pd]
 ASSERT_TUPLES = [(pd, assert_frame_equal)]

skrub/tests/test_gap_encoder.py

@@ -5,8 +5,8 @@
 from sklearn.model_selection import train_test_split
 
 from skrub import GapEncoder
-from skrub._dataframe._namespace import is_namespace_polars
 from skrub._dataframe._polars import POLARS_SETUP
+from skrub._utils import is_namespace_polars
 from skrub.datasets import fetch_midwest_survey
 from skrub.tests.utils import generate_data
 

skrub/tests/test_interpolation_join.py

@@ -6,8 +6,8 @@
 from sklearn.neighbors import KNeighborsClassifier, KNeighborsRegressor
 
 from skrub import InterpolationJoiner
-from skrub._dataframe._namespace import is_namespace_polars
 from skrub._dataframe._polars import POLARS_SETUP
+from skrub._utils import is_namespace_polars
 
 MODULES = [pd]
 

skrub/tests/test_joiner.py

@@ -4,8 +4,8 @@
 from pandas.testing import assert_frame_equal
 
 from skrub import Joiner
-from skrub._dataframe._namespace import is_namespace_polars
 from skrub._dataframe._polars import POLARS_SETUP
+from skrub._utils import is_namespace_polars
 
 MODULES = [pd]
 ASSERT_TUPLES = [(pd, assert_frame_equal)]

skrub/tests/test_similarity_encoder.py

@@ -7,10 +7,10 @@
 from sklearn.exceptions import NotFittedError
 
 from skrub import SimilarityEncoder
-from skrub._dataframe._namespace import is_namespace_polars
 from skrub._dataframe._polars import POLARS_SETUP
 from skrub._similarity_encoder import ngram_similarity_matrix
 from skrub._string_distances import ngram_similarity
+from skrub._utils import is_namespace_polars
 
 MODULES = [pd]
 INPUT_TYPES = ["list", "numpy", "pandas"]