Backport PR pandas-dev#57174 on branch 2.2.x (BUG: Interchange protocol implementation allows non-string column names) (pandas-dev#57203)

Backport PR pandas-dev#57174: BUG: Interchange protocol implementation allows non-string column names

Co-authored-by: Marco Edward Gorelli <[email protected]>

Author: meeseeksmachine

doc/source/whatsnew/v2.2.1.rst

@@ -32,6 +32,7 @@ Bug fixes
 ~~~~~~~~~
 - Fixed bug in :func:`pandas.api.interchange.from_dataframe` which was raising for Nullable integers (:issue:`55069`)
 - Fixed bug in :func:`pandas.api.interchange.from_dataframe` which was raising for empty inputs (:issue:`56700`)
+- Fixed bug in :func:`pandas.api.interchange.from_dataframe` which wasn't converting columns names to strings (:issue:`55069`)
 - Fixed bug in :meth:`DataFrame.__getitem__` for empty :class:`DataFrame` with Copy-on-Write enabled (:issue:`57130`)
 
 .. ---------------------------------------------------------------------------

pandas/core/interchange/column.py

@@ -77,6 +77,14 @@ def __init__(self, column: pd.Series, allow_copy: bool = True) -> None:
         Note: doesn't deal with extension arrays yet, just assume a regular
         Series/ndarray for now.
         """
+        if isinstance(column, pd.DataFrame):
+            raise TypeError(
+                "Expected a Series, got a DataFrame. This likely happened "
+                "because you called __dataframe__ on a DataFrame which, "
+                "after converting column names to string, resulted in duplicated "
+                f"names: {column.columns}. Please rename these columns before "
+                "using the interchange protocol."
+            )
         if not isinstance(column, pd.Series):
             raise NotImplementedError(f"Columns of type {type(column)} not handled yet")
 

pandas/core/interchange/dataframe.py

@@ -32,7 +32,7 @@ def __init__(self, df: DataFrame, allow_copy: bool = True) -> None:
         Constructor - an instance of this (private) class is returned from
         `pd.DataFrame.__dataframe__`.
         """
-        self._df = df
+        self._df = df.rename(columns=str, copy=False)
         self._allow_copy = allow_copy
 
     def __dataframe__(

pandas/tests/interchange/test_impl.py

@@ -180,8 +180,6 @@ def test_missing_from_masked():
         }
     )
 
-    df2 = df.__dataframe__()
-
     rng = np.random.default_rng(2)
     dict_null = {col: rng.integers(low=0, high=len(df)) for col in df.columns}
     for col, num_nulls in dict_null.items():
@@ -382,6 +380,30 @@ def test_large_string():
     tm.assert_frame_equal(result, expected)
 
 
+def test_non_str_names():
+    # https://github.com/pandas-dev/pandas/issues/56701
+    df = pd.Series([1, 2, 3], name=0).to_frame()
+    names = df.__dataframe__().column_names()
+    assert names == ["0"]
+
+
+def test_non_str_names_w_duplicates():
+    # https://github.com/pandas-dev/pandas/issues/56701
+    df = pd.DataFrame({"0": [1, 2, 3], 0: [4, 5, 6]})
+    dfi = df.__dataframe__()
+    with pytest.raises(
+        TypeError,
+        match=(
+            "Expected a Series, got a DataFrame. This likely happened because you "
+            "called __dataframe__ on a DataFrame which, after converting column "
+            r"names to string, resulted in duplicated names: Index\(\['0', '0'\], "
+            r"dtype='object'\). Please rename these columns before using the "
+            "interchange protocol."
+        ),
+    ):
+        pd.api.interchange.from_dataframe(dfi, allow_copy=False)
+
+
 @pytest.mark.parametrize(
     "dtype", ["Int8", pytest.param("Int8[pyarrow]", marks=td.skip_if_no("pyarrow"))]
 )