Add na_tuple kwarg

Author: jschendel

doc/source/whatsnew/v0.22.0.txt

@@ -138,6 +138,7 @@ Other Enhancements
 - :func:`Series` / :func:`DataFrame` tab completion also returns identifiers in the first level of a :func:`MultiIndex`. (:issue:`16326`)
 - :func:`read_excel()` has gained the ``nrows`` parameter (:issue:`16645`)
 - :func:``DataFrame.to_json`` and ``Series.to_json`` now accept an ``index`` argument which allows the user to exclude the index from the JSON output (:issue:`17394`)
+- ``IntervalIndex.to_tuples()`` has gained the ``na_tuple`` parameter to control whether NA is returned as a tuple of NA, or NA itself (:issue:`18756`)
 
 .. _whatsnew_0220.api_breaking:
 
@@ -264,7 +265,6 @@ Conversion
 - Fixed a bug where ``FY5253`` date offsets could incorrectly raise an ``AssertionError`` in arithmetic operatons (:issue:`14774`)
 - Bug in :meth:`Index.astype` with a categorical dtype where the resultant index is not converted to a :class:`CategoricalIndex` for all types of index (:issue:`18630`)
 - Bug in :meth:`Series.astype` and ``Categorical.astype()`` where an existing categorical data does not get updated (:issue:`10696`, :issue:`18593`)
-- Bug in ``IntervalIndex.to_tuples()`` where NA values are returned as a tuple of NA values instead of the NA value itself (:issue:`18756`)
 
 
 Indexing

pandas/core/indexes/interval.py

@@ -544,10 +544,28 @@ def from_tuples(cls, data, closed='right', name=None, copy=False):
 
         return cls.from_arrays(left, right, closed, name=name, copy=False)
 
-    def to_tuples(self):
-        """Return an Index of tuples of the form (left, right)"""
+    def to_tuples(self, na_tuple=True):
+        """
+        Return an Index of tuples of the form (left, right)
+
+        Parameters
+        ----------
+        na_tuple : boolean, default True
+            Returns NA as a tuple if True, ``(nan, nan)``, or just as the NA
+            value itself if False, ``nan``.
+
+            ..versionadded:: 0.22.0
+
+        Examples
+        --------
+        >>>  idx = pd.IntervalIndex.from_arrays([0, np.nan, 2], [1, np.nan, 3])
+        >>>  idx.to_tuples()
+        Index([(0.0, 1.0), (nan, nan), (2.0, 3.0)], dtype='object')
+        >>>  idx.to_tuples(na_tuple=False)
+        Index([(0.0, 1.0), nan, (2.0, 3.0)], dtype='object')
+        """
         tuples = _asarray_tuplesafe(zip(self.left, self.right))
-        if self.hasnans:
+        if not na_tuple:
             # GH 18756
             tuples = np.where(~self._isnan, tuples, np.nan)
         return Index(tuples)

pandas/tests/indexes/test_interval.py

@@ -1075,17 +1075,43 @@ def test_is_non_overlapping_monotonic(self, closed):
 
     @pytest.mark.parametrize('tuples', [
         lzip(range(10), range(1, 11)),
-        lzip(range(10), range(1, 11)) + [np.nan],
         lzip(date_range('20170101', periods=10),
              date_range('20170101', periods=10)),
-        [np.nan] + lzip(date_range('20170101', periods=10),
-                        date_range('20170101', periods=10))])
+        lzip(timedelta_range('0 days', periods=10),
+             timedelta_range('1 day', periods=10))])
     def test_to_tuples(self, tuples):
         # GH 18756
-        result = IntervalIndex.from_tuples(tuples).to_tuples()
+        idx = IntervalIndex.from_tuples(tuples)
+        result = idx.to_tuples()
         expected = Index(_asarray_tuplesafe(tuples))
         tm.assert_index_equal(result, expected)
 
+    @pytest.mark.parametrize('tuples', [
+        lzip(range(10), range(1, 11)) + [np.nan],
+        lzip(date_range('20170101', periods=10),
+             date_range('20170101', periods=10)) + [np.nan],
+        lzip(timedelta_range('0 days', periods=10),
+             timedelta_range('1 day', periods=10)) + [np.nan]])
+    @pytest.mark.parametrize('na_tuple', [True, False])
+    def test_to_tuples_na(self, tuples, na_tuple):
+        # GH 18756
+        idx = IntervalIndex.from_tuples(tuples)
+        result = idx.to_tuples(na_tuple=na_tuple)
+
+        # check the non-NA portion
+        expected_notna = Index(_asarray_tuplesafe(tuples[:-1]))
+        result_notna = result[:-1]
+        tm.assert_index_equal(result_notna, expected_notna)
+
+        # check the NA portion
+        result_na = result[-1]
+        if na_tuple:
+            assert isinstance(result_na, tuple)
+            assert len(result_na) == 2
+            assert all(isna(x) for x in result_na)
+        else:
+            assert isna(result_na)
+
 
 class TestIntervalRange(object):