Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Implemented bindings for ceil timestamp operation #9141

Merged
merged 7 commits into from
Sep 4, 2021
Merged
Show file tree
Hide file tree
Changes from 5 commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
11 changes: 11 additions & 0 deletions python/cudf/cudf/_lib/cpp/datetime.pxd
Original file line number Diff line number Diff line change
Expand Up @@ -12,6 +12,17 @@ cdef extern from "cudf/datetime.hpp" namespace "cudf::datetime" nogil:
cdef unique_ptr[column] extract_hour(const column_view& column) except +
cdef unique_ptr[column] extract_minute(const column_view& column) except +
cdef unique_ptr[column] extract_second(const column_view& column) except +
cdef unique_ptr[column] ceil_day(const column_view& column) except +
cdef unique_ptr[column] ceil_hour(const column_view& column) except +
cdef unique_ptr[column] ceil_minute(const column_view& column) except +
cdef unique_ptr[column] ceil_second(const column_view& column) except +
cdef unique_ptr[column] ceil_millisecond(
const column_view& column
) except +
cdef unique_ptr[column] ceil_microsecond(
const column_view& column
) except +
cdef unique_ptr[column] ceil_nanosecond(const column_view& column) except +
cdef unique_ptr[column] add_calendrical_months(
const column_view& timestamps,
const column_view& months
Expand Down
27 changes: 27 additions & 0 deletions python/cudf/cudf/_lib/datetime.pyx
Original file line number Diff line number Diff line change
Expand Up @@ -59,6 +59,33 @@ def extract_datetime_component(Column col, object field):
return result


def ceil_datetime(Column col, object field):
cdef unique_ptr[column] c_result
cdef column_view col_view = col.view()

with nogil:
# https://pandas.pydata.org/pandas-docs/version/0.25.0/reference/api/pandas.Timedelta.resolution.html
if field == "D":
c_result = move(libcudf_datetime.ceil_day(col_view))
elif field == "H":
c_result = move(libcudf_datetime.ceil_hour(col_view))
elif field == "T":
c_result = move(libcudf_datetime.ceil_minute(col_view))
elif field == "S":
c_result = move(libcudf_datetime.ceil_second(col_view))
elif field == "L":
c_result = move(libcudf_datetime.ceil_millisecond(col_view))
elif field == "U":
c_result = move(libcudf_datetime.ceil_microsecond(col_view))
elif field == "N":
c_result = move(libcudf_datetime.ceil_nanosecond(col_view))
else:
raise ValueError(f"Invalid resolution: '{field}'")

result = Column.from_unique_ptr(move(c_result))
return result


def is_leap_year(Column col):
"""Returns a boolean indicator whether the year of the date is a leap year
"""
Expand Down
3 changes: 3 additions & 0 deletions python/cudf/cudf/core/column/datetime.py
Original file line number Diff line number Diff line change
Expand Up @@ -170,6 +170,9 @@ def values(self):
def get_dt_field(self, field: str) -> ColumnBase:
return libcudf.datetime.extract_datetime_component(self, field)

def ceil_datetime(self, field: str) -> ColumnBase:
shaneding marked this conversation as resolved.
Show resolved Hide resolved
return libcudf.datetime.ceil_datetime(self, field)

def normalize_binop_value(self, other: DatetimeLikeScalar) -> ScalarLike:
if isinstance(other, cudf.Scalar):
return other
Expand Down
7 changes: 7 additions & 0 deletions python/cudf/cudf/core/series.py
Original file line number Diff line number Diff line change
Expand Up @@ -5923,6 +5923,13 @@ def _get_dt_field(self, field):
data=out_column, index=self.series._index, name=self.series.name
)

def ceil(self, field):
out_column = self.series._column.ceil_datetime(field)

return Series(
data=out_column, index=self.series._index, name=self.series.name
)

def strftime(self, date_format, *args, **kwargs):
"""
Convert to Series using specified ``date_format``.
Expand Down
28 changes: 28 additions & 0 deletions python/cudf/cudf/tests/test_datetime.py
Original file line number Diff line number Diff line change
Expand Up @@ -1551,3 +1551,31 @@ def test_error_values():
match="DateTime Arrays is not yet implemented in cudf",
):
s.values


@pytest.mark.parametrize(
"data",
[
(
[
"2020-05-31 08:00:00",
"1999-12-31 18:40:10",
"2000-12-31 04:00:05",
"1900-02-28 07:00:06",
"1800-03-14 07:30:20",
"2100-03-14 07:30:20",
"1970-01-01 00:00:09",
"1969-12-31 12:59:10",
]
)
],
)
@pytest.mark.parametrize("resolution", ["D", "H", "T", "S", "L", "U", "N"])
def test_ceil(data, resolution):
for time_type in DATETIME_TYPES:
shaneding marked this conversation as resolved.
Show resolved Hide resolved
ps = pd.Series(data, dtype=time_type)
gs = cudf.from_pandas(ps)

expect = ps.dt.ceil(resolution)
got = gs.dt.ceil(resolution)
assert_eq(expect, got)