Commit bab4b9a

committed

BUG: Fix Series.str.zfill for ArrowDtype string arrays #61485

1 parent cfe54bd commit bab4b9aCopy full SHA for bab4b9a

File tree

+30

-2

lines changed

+30

-2

lines changed

Lines changed: 1 addition & 1 deletion

Original file line number	Diff line number	Diff line change
`@@ -725,8 +725,8 @@ Conversion`
`725`	`725`
`726`	`726`	`Strings`
`727`	`727`	`^^^^^^^`
	`728`	+- Bug in :meth:`Series.str.zfill` raising ``AttributeError`` for ``ArrowDtype(pa.string())``. Now supported via ``_str_zfill`` implementation in ``ArrowExtensionArray`` (:issue:`61485`)
`728`	`729`	- Bug in :meth:`Series.value_counts` would not respect ``sort=False`` for series having ``string`` dtype (:issue:`55224`)
`729`		`--`
`730`	`730`
`731`	`731`	`Interval`
`732`	`732`	`^^^^^^^^`

Lines changed: 6 additions & 0 deletions

Original file line number	Diff line number	Diff line change
`@@ -2601,6 +2601,12 @@ def _str_wrap(self, width: int, **kwargs) -> Self:`
`2601`	`2601`	`result = self._apply_elementwise(predicate)`
`2602`	`2602`	`return type(self)(pa.chunked_array(result))`
`2603`	`2603`
	`2604`	`+ def _str_zfill(self, width: int) -> Self:`
	`2605`	`+ # TODO: Replace with pc.utf8_zfill when supported by arrow`
	`2606`	`+ predicate = lambda val: val.zfill(width)`
	`2607`	`+ result = self._apply_elementwise(predicate)`
	`2608`	`+ return type(self)(pa.chunked_array(result))`
	`2609`	`+`
`2604`	`2610`	`@property`
`2605`	`2611`	`def _dt_days(self) -> Self:`
`2606`	`2612`	`return type(self)(`

Lines changed: 7 additions & 1 deletion

Original file line number	Diff line number	Diff line change
`@@ -1939,7 +1939,13 @@ def zfill(self, width: int):`
`1939`	`1939`	`msg = f"width must be of integer type, not {type(width).__name__}"`
`1940`	`1940`	`raise TypeError(msg)`
`1941`	`1941`	`f = lambda x: x.zfill(width)`
`1942`		`- result = self._data.array._str_map(f)`
	`1942`	`+ array = self._data.array`
	`1943`	`+`
	`1944`	`+ if hasattr(array, "_str_zfill"):`
	`1945`	`+ result = array._str_zfill(width)`
	`1946`	`+ else:`
	`1947`	`+ result = array._str_map(f)`
	`1948`	`+`
`1943`	`1949`	`return self._wrap_result(result)`
`1944`	`1950`
`1945`	`1951`	`def slice(self, start=None, stop=None, step=None):`

Lines changed: 16 additions & 0 deletions

Original file line number	Diff line number	Diff line change
`@@ -110,3 +110,19 @@ def test_string_array_extract(nullable_string_dtype):`
`110`	`110`
`111`	`111`	`result = result.astype(object)`
`112`	`112`	`tm.assert_equal(result, expected)`
	`113`	`+`
	`114`	`+`
	`115`	`+@pytest.mark.parametrize(`
	`116`	`+ "values, width, expected",`
	`117`	`+ [`
	`118`	`+ (["a", "ab", "abc", None], 4, ["000a", "00ab", "0abc", None]),`
	`119`	`+ (["1", "-1", "+1", None], 4, ["0001", "-001", "+001", None]),`
	`120`	`+ (["1234", "-1234"], 3, ["1234", "-1234"]),`
	`121`	`+ ],`
	`122`	`+)`
	`123`	`+def test_string_array_zfill(nullable_string_dtype, values, width, expected):`
	`124`	`+ # GH #61485`
	`125`	`+ s = Series(values, dtype=nullable_string_dtype)`
	`126`	`+ result = s.str.zfill(width)`
	`127`	`+ expected = Series(expected, dtype=nullable_string_dtype)`
	`128`	`+ tm.assert_series_equal(result, expected)`

Comments

(0)