Set check_dtype=False for tests

lumen/tests/sources/test_base.py

@@ -17,6 +17,9 @@
 except ImportError:
     fastparquet = None
 
+from ..utils import assert_frame_equal_no_dtype_check
+
+
 @pytest.fixture
 def source(make_filesource):
     root = os.path.dirname(__file__)
@@ -95,7 +98,7 @@
     column, value, _ = column_value_type
     kwargs = {column: value, '__dask': dask}
     filtered = source.get('test', **kwargs)
-    pd.testing.assert_frame_equal(filtered, expected_df)
+    assert_frame_equal_no_dtype_check(filtered, expected_df)
 
 
 @pytest.mark.parametrize(
@@ -122,8 +125,8 @@
     assert cache_key in source._cache
     cached_df = source._cache[cache_key]
     if dask:
         cached_df = cached_df.compute()
-    pd.testing.assert_frame_equal(cached_df, expected_df)
+    assert_frame_equal_no_dtype_check(cached_df, expected_df)
     cache_key = source._get_key('test', **kwargs)
     assert cache_key in source._cache
 
@@ -160,7 +163,7 @@
 
     # Patch index names due to https://github.com/dask/fastparquet/issues/732
     df.index.names = [None]
-    pd.testing.assert_frame_equal(
+    assert_frame_equal_no_dtype_check(
         df,
         makeMixedDataFrame().iloc[1:3]
     )
@@ -177,7 +180,7 @@
     state.variables.tables = {'test': 'test2.csv'}
     df = source.get('test')
     expected = mixed_df.iloc[::-1].reset_index(drop=True)
-    pd.testing.assert_frame_equal(df, expected)
+    assert_frame_equal_no_dtype_check(df, expected)
 
 
 def test_extension_of_comlicated_url(source):

lumen/tests/sources/test_derived.py

@@ -1,10 +1,11 @@
 import os
 
-import pandas as pd
 import pytest
 
 from lumen.sources.base import DerivedSource, Source
 
+from ..utils import assert_frame_equal_no_dtype_check
+
 
 @pytest.fixture
 def original(make_filesource):
@@ -60,7 +61,7 @@ def test_derived_mirror_source(original, mirror_mode_spec):
     assert derived.get_tables() == original.get_tables()
     assert derived.get_schema() == original.get_schema()
     for table in original.get_tables():
-        pd.testing.assert_frame_equal(derived.get(table), original.get(table))
+        assert_frame_equal_no_dtype_check(derived.get(table), original.get(table))
 
 
 @pytest.mark.parametrize("additional_spec", [
@@ -75,13 +76,13 @@ def test_derived_mirror_source_apply(
     derived = Source.from_spec(mirror_mode_spec)
     assert derived.get_tables() == original.get_tables()
     assert derived.get_schema('test') == expected_schema
-    pd.testing.assert_frame_equal(derived.get('test'), expected_table)
+    assert_frame_equal_no_dtype_check(derived.get('test'), expected_table)
 
 
 def test_derived_tables_source(original, tables_mode_spec, mixed_df):
     derived = Source.from_spec(tables_mode_spec)
     assert derived.get_tables() == ['derived']
-    pd.testing.assert_frame_equal(derived.get('derived'), mixed_df)
+    assert_frame_equal_no_dtype_check(derived.get('derived'), mixed_df)
     assert original.get_schema('test') == derived.get_schema('derived')
 
 
@@ -96,7 +97,7 @@ def test_derived_tables_source_apply(
     tables_mode_spec['tables']['derived'][spec_key] = spec_value
     derived = Source.from_spec(tables_mode_spec)
     assert derived.get_tables() == ['derived']
-    pd.testing.assert_frame_equal(derived.get('derived'), expected_table)
+    assert_frame_equal_no_dtype_check(derived.get('derived'), expected_table)
     assert derived.get_schema('derived') == expected_schema
 
 
@@ -106,7 +107,7 @@ def test_derived_get_query_cache(original, mirror_mode_spec):
     cache_key = derived._get_key('test')
     assert cache_key in derived._cache
     cached_df = derived._cache[cache_key]
-    pd.testing.assert_frame_equal(cached_df, df)
+    assert_frame_equal_no_dtype_check(cached_df, df)
 
 
 def test_derived_clear_cache(original, mirror_mode_spec):

lumen/tests/sources/test_duckdb.py

@@ -1,7 +1,6 @@
 import datetime as dt
 import os
 
-import pandas as pd
 import pytest
 
 from lumen.transforms.sql import SQLGroupBy
@@ -12,6 +11,7 @@
 except ImportError:
     pytestmark = pytest.mark.skip(reason="Duckdb is not installed")
 
+from ..utils import assert_frame_equal_no_dtype_check
 
 
 @pytest.fixture
@@ -49,7 +49,7 @@ def test_duckdb_get_tables(duckdb_source, source_tables):
     tables = duckdb_source.get_tables()
     assert not len(set(tables) - set(source_tables.keys()))
     for table in tables:
-        pd.testing.assert_frame_equal(
+        assert_frame_equal_no_dtype_check(
             duckdb_source.get(table),
             source_tables[table],
         )
@@ -119,7 +119,7 @@ def test_duckdb_filter(duckdb_source, table_column_value_type, dask, expected_fi
     table, column, value, _ = table_column_value_type
     kwargs = {column: value}
     filtered = duckdb_source.get(table, __dask=dask, **kwargs)
-    pd.testing.assert_frame_equal(filtered, expected_filtered_df.reset_index(drop=True))
+    assert_frame_equal_no_dtype_check(filtered, expected_filtered_df.reset_index(drop=True))
 
 
 @pytest.mark.flaky(reruns=3)
@@ -128,7 +128,7 @@ def test_duckdb_transforms(duckdb_source, source_tables):
     transforms = [SQLGroupBy(by=['B'], aggregates={'SUM': 'A'})]
     transformed = duckdb_source.get('test_sql', sql_transforms=transforms)
     expected = df_test_sql.groupby('B')['A'].sum().reset_index()
-    pd.testing.assert_frame_equal(transformed, expected)
+    assert_frame_equal_no_dtype_check(transformed, expected)
 
 
 @pytest.mark.flaky(reruns=3)
@@ -141,7 +141,7 @@ def test_duckdb_transforms_cache(duckdb_source, source_tables):
     assert cache_key in duckdb_source._cache
 
     expected = df_test_sql.groupby('B')['A'].sum().reset_index()
-    pd.testing.assert_frame_equal(duckdb_source._cache[cache_key], expected)
+    assert_frame_equal_no_dtype_check(duckdb_source._cache[cache_key], expected)
 
     cache_key = duckdb_source._get_key('test_sql', sql_transforms=transforms)
     assert cache_key in duckdb_source._cache
@@ -166,4 +166,4 @@ def test_duckdb_source_ephemeral_roundtrips(duckdb_memory_source, mixed_df):
 
 def test_duckdb_source_mirrors_source(duckdb_source):
     mirrored = DuckDBSource(uri=':memory:', mirrors={'mixed': (duckdb_source, 'test_sql')})
-    pd.testing.assert_frame_equal(duckdb_source.get('test_sql'), mirrored.get('mixed'))
+    assert_frame_equal_no_dtype_check(duckdb_source.get('test_sql'), mirrored.get('mixed'))

lumen/tests/sources/test_intake.py

@@ -1,12 +1,13 @@
 import datetime as dt
 import os
 
-import pandas as pd
 import pytest
 import yaml
 
 from lumen.sources.intake import IntakeSource
 
+from ..utils import assert_frame_equal_no_dtype_check
+
 
 @pytest.fixture
 def source():
@@ -36,15 +37,15 @@ def test_intake_resolve_module_type():
 
 
 def test_intake_source_from_file(source, mixed_df):
-    pd.testing.assert_frame_equal(source.get('test'), mixed_df)
+    assert_frame_equal_no_dtype_check(source.get('test'), mixed_df)
 
 
 def test_intake_source_from_dict(mixed_df):
     root = os.path.dirname(__file__)
     with open(os.path.join(root, 'catalog.yml')) as f:
         catalog = yaml.load(f, Loader=yaml.Loader)
     source = IntakeSource(catalog=catalog, root=root)
-    pd.testing.assert_frame_equal(source.get('test'), mixed_df)
+    assert_frame_equal_no_dtype_check(source.get('test'), mixed_df)
 
 
 @pytest.mark.parametrize(
@@ -65,4 +66,4 @@ def test_intake_filter(source, table_column_value_type, dask, expected_filtered_
     table, column, value, _ = table_column_value_type
     kwargs = {column: value}
     filtered = source.get(table, __dask=dask, **kwargs)
-    pd.testing.assert_frame_equal(filtered, expected_filtered_df)
+    assert_frame_equal_no_dtype_check(filtered, expected_filtered_df)

lumen/tests/sources/test_intake_sql.py

@@ -1,7 +1,6 @@
 import datetime as dt
 import os
 
-import pandas as pd
 import pytest
 
 try:
@@ -12,6 +11,8 @@
 from lumen.sources.intake_sql import IntakeSQLSource
 from lumen.transforms.sql import SQLGroupBy
 
+from ..utils import assert_frame_equal_no_dtype_check
+
 
 @pytest.fixture
 def source():
@@ -31,7 +32,7 @@ def test_intake_sql_get_tables(source, source_tables):
     tables = source.get_tables()
     assert tables == list(source_tables.keys())
     for table in tables:
-        pd.testing.assert_frame_equal(source.get(table), source_tables[table])
+        assert_frame_equal_no_dtype_check(source.get(table), source_tables[table])
 
 
 def test_intake_sql_get_schema(source):
@@ -99,15 +100,15 @@ def test_intake_sql_filter(source, table_column_value_type, dask, expected_filte
     table, column, value, _ = table_column_value_type
     kwargs = {column: value}
     filtered = source.get(table, __dask=dask, **kwargs)
-    pd.testing.assert_frame_equal(filtered, expected_filtered_df.reset_index(drop=True))
+    assert_frame_equal_no_dtype_check(filtered, expected_filtered_df.reset_index(drop=True))
 
 
 def test_intake_sql_transforms(source, source_tables):
     df_test_sql = source_tables['test_sql']
     transforms = [SQLGroupBy(by=['B'], aggregates={'SUM': 'A'})]
     transformed = source.get('test_sql', sql_transforms=transforms)
     expected = df_test_sql.groupby('B')['A'].sum().reset_index()
-    pd.testing.assert_frame_equal(transformed, expected)
+    assert_frame_equal_no_dtype_check(transformed, expected)
 
 
 def test_intake_sql_transforms_cache(source, source_tables):
@@ -119,7 +120,7 @@ def test_intake_sql_transforms_cache(source, source_tables):
     assert cache_key in source._cache
 
     expected = df_test_sql.groupby('B')['A'].sum().reset_index()
-    pd.testing.assert_frame_equal(source._cache[cache_key], expected)
+    assert_frame_equal_no_dtype_check(source._cache[cache_key], expected)
 
     cache_key = source._get_key('test_sql', sql_transforms=transforms)
     assert cache_key in source._cache

lumen/tests/test_dashboard.py

@@ -1,6 +1,5 @@
 import pathlib
 
-import pandas as pd
 import panel as pn
 import pytest
 
@@ -10,6 +9,7 @@
 from lumen.validation import ValidationError
 
 from .test_pipeline import sql_available
+from .utils import assert_frame_equal_no_dtype_check
 
 
 def test_dashboard_with_local_view(set_root):
@@ -131,11 +131,11 @@ def test_dashboard_with_sql_source_and_transforms(set_root, document, mixed_df_o
     layout.update()
 
     table = layout._cards[0]._card[0][0]
-    pd.testing.assert_frame_equal(table.value, mixed_df_object_type)
+    assert_frame_equal_no_dtype_check(table.value, mixed_df_object_type)
 
     dashboard._sidebar[0][0][0]._widgets['limit'].value = 2
 
-    pd.testing.assert_frame_equal(table.value, mixed_df_object_type.iloc[:2])
+    assert_frame_equal_no_dtype_check(table.value, mixed_df_object_type.iloc[:2])
 
 def test_dashboard_with_transform_variable(set_root, document, mixed_df):
     root = pathlib.Path(__file__).parent / 'sample_dashboard'
@@ -146,11 +146,11 @@ def test_dashboard_with_transform_variable(set_root, document, mixed_df):
     layout.update()
 
     table = layout._cards[0]._card[0][0]
-    pd.testing.assert_frame_equal(table.value, mixed_df)
+    assert_frame_equal_no_dtype_check(table.value, mixed_df)
 
     state.variables.length = 2
 
-    pd.testing.assert_frame_equal(table.value, mixed_df.iloc[:2])
+    assert_frame_equal_no_dtype_check(table.value, mixed_df.iloc[:2])
 
 def test_dashboard_with_source_variable(set_root, document, mixed_df):
     root = pathlib.Path(__file__).parent / 'sample_dashboard'
@@ -161,11 +161,11 @@ def test_dashboard_with_source_variable(set_root, document, mixed_df):
     layout.update()
 
     table = layout._cards[0]._card[0][0]
-    pd.testing.assert_frame_equal(table.value, mixed_df)
+    assert_frame_equal_no_dtype_check(table.value, mixed_df)
 
     state.variables.tables = {'test': '../sources/test2.csv'}
 
-    pd.testing.assert_frame_equal(table.value, mixed_df.iloc[::-1].reset_index(drop=True))
+    assert_frame_equal_no_dtype_check(table.value, mixed_df.iloc[::-1].reset_index(drop=True))
 
 def test_dashboard_with_nested_source_variable(set_root, document, mixed_df):
     root = pathlib.Path(__file__).parent / 'sample_dashboard'
@@ -176,11 +176,11 @@ def test_dashboard_with_nested_source_variable(set_root, document, mixed_df):
     layout.update()
 
     table = layout._cards[0]._card[0][0]
-    pd.testing.assert_frame_equal(table.value, mixed_df)
+    assert_frame_equal_no_dtype_check(table.value, mixed_df)
 
     state.variables.ticker = '../sources/test2.csv'
 
-    pd.testing.assert_frame_equal(table.value, mixed_df.iloc[::-1].reset_index(drop=True))
+    assert_frame_equal_no_dtype_check(table.value, mixed_df.iloc[::-1].reset_index(drop=True))
 
 def test_dashboard_with_view_variable(set_root, document):
     root = pathlib.Path(__file__).parent / 'sample_dashboard'