FIX-modin-project#4060: Fix _to_pandas() dtypes for empty frames.

Signed-off-by: mvashishtha <[email protected]>
mvashishtha · Jan 28, 2022 · 8312829 · 8312829
1 parent cafd213
commit 8312829
Show file tree

Hide file tree

Showing 3 changed files with 17 additions and 2 deletions.
diff --git a/docs/release_notes/release_notes-0.14.0.rst b/docs/release_notes/release_notes-0.14.0.rst
@@ -4,7 +4,7 @@ Key Features and Updates
 ------------------------
 
 * Stability and Bugfixes
-  *
+  * Fix to_pandas() dtypes for empty dataframes and series (0d7c273)
 * Performance enhancements
   *
 * Benchmarking enhancements

diff --git a/modin/core/dataframe/pandas/dataframe/dataframe.py b/modin/core/dataframe/pandas/dataframe/dataframe.py
@@ -2698,7 +2698,9 @@ def to_pandas(self):
         """
         df = self._partition_mgr_cls.to_pandas(self._partitions)
         if df.empty:
-            df = pandas.DataFrame(columns=self.columns, index=self.index)
+            df = pandas.DataFrame(columns=self.columns, index=self.index).astype(
+                self.dtypes
+            )
         else:
             for axis in [0, 1]:
                 ErrorMessage.catch_bugs_and_request_email(

diff --git a/modin/pandas/test/test_general.py b/modin/pandas/test/test_general.py
@@ -19,6 +19,7 @@
 from numpy.testing import assert_array_equal
 from modin.utils import get_current_execution, to_pandas
 from modin.test.test_utils import warns_that_defaulting_to_pandas
+from pandas.testing import assert_frame_equal, assert_series_equal
 
 from .utils import (
     test_data_values,
@@ -765,3 +766,15 @@ def test_empty_dataframe():
 def test_empty_series():
     s = pd.Series([])
     pd.to_numeric(s)
+
+
+def test_to_pandas_empty_series():
+    pandas_series = pandas.Series([], dtype="bool")
+    modin_series = pd.Series(pandas_series)
+    assert_series_equal(pandas_series, modin_series._to_pandas(), check_dtype=True)
+
+
+def test_to_pandas_empty_dataframe():
+    pandas_df = pandas.DataFrame([[]], dtype="bool")
+    modin_df = pd.DataFrame(pandas_df)
+    assert_frame_equal(pandas_df, modin_df._to_pandas(), check_dtype=True)