Port pandas 3.0 change to Day offset to cftime version

Author: spencerkclark

doc/whats-new.rst

@@ -16,6 +16,15 @@ New Features
 
 Breaking changes
 ~~~~~~~~~~~~~~~~
+- Following pandas 3.0 (`pandas-dev/pandas#61985
+  <https://github.com/pandas-dev/pandas/pull/61985>`_), ``Day`` is no longer
+  considered a ``Tick``-like frequency. Therefore non-``None`` values of
+  ``offset`` and non-``"start_day"`` values of ``origin`` will have no effect
+  when resampling to a daily frequency for objects indexed by a
+  :py:class:`xarray.CFTimeIndex`. As in `pandas-dev/pandas#62101
+  <https://github.com/pandas-dev/pandas/pull/62101>`_ warnings will be emitted
+  if non default values are provided in this context (:issue:`10640`,
+  :pull:`10650`). By `Spencer Clark <https://github.com/spencerkclark>`_.
 
 
 Deprecations

xarray/coding/cftime_offsets.py

@@ -609,14 +609,19 @@ def rollback(self, date):
             return date - YearEnd(month=self.month)
 
 
-class Day(Tick):
-    _freq = "D"
+class Day(BaseCFTimeOffset):
+    """Day offset following definition in pandas/_libs/tslibs/offsets.pyx"""
 
-    def as_timedelta(self) -> timedelta:
-        return timedelta(days=self.n)
+    _freq = "D"
 
     def __apply__(self, other):
-        return other + self.as_timedelta()
+        if isinstance(other, Day):
+            return Day(self.n + other.n)
+        else:
+            return other + timedelta(days=self.n)
+
+    def onOffset(self, date) -> bool:
+        return True
 
 
 class Hour(Tick):
@@ -719,7 +724,7 @@ def _generate_anchored_offsets(
 
 # pandas defines these offsets as "Tick" objects, which for instance have
 # distinct behavior from monthly or longer frequencies in resample.
-CFTIME_TICKS = (Day, Hour, Minute, Second)
+CFTIME_TICKS = (Hour, Minute, Second)
 
 
 def _generate_anchored_deprecated_frequencies(

xarray/coding/cftimeindex.py

@@ -707,16 +707,22 @@ def freq(self):
 
     def _round_via_method(self, freq, method):
         """Round dates using a specified method."""
-        from xarray.coding.cftime_offsets import CFTIME_TICKS, to_offset
+        from xarray.coding.cftime_offsets import CFTIME_TICKS, Day, to_offset
 
         if not self._data.size:
             return CFTimeIndex(np.array(self))
 
         offset = to_offset(freq)
-        if not isinstance(offset, CFTIME_TICKS):
+        if isinstance(offset, Day):
+            # Following pandas, "In the 'round' context, Day unambiguously
+            # means 24h, not calendar-day"
+            offset_as_timedelta = timedelta(days=offset.n)
+        elif isinstance(offset, CFTIME_TICKS):
+            offset_as_timedelta = offset.as_timedelta()
+        else:
             raise ValueError(f"{offset} is a non-fixed frequency")
 
-        unit = _total_microseconds(offset.as_timedelta())
+        unit = _total_microseconds(offset_as_timedelta)
         values = self.asi8
         rounded = method(values, unit)
         return _cftimeindex_from_i8(rounded, self.date_type, self.name)

xarray/core/resample_cftime.py

@@ -45,6 +45,7 @@
 import pandas as pd
 
 from xarray.coding.cftime_offsets import (
+    CFTIME_TICKS,
     BaseCFTimeOffset,
     MonthEnd,
     QuarterEnd,
@@ -56,6 +57,7 @@
 )
 from xarray.coding.cftimeindex import CFTimeIndex
 from xarray.core.types import SideOptions
+from xarray.core.utils import emit_user_level_warning
 
 if typing.TYPE_CHECKING:
     from xarray.core.types import CFTimeDatetime, ResampleCompatible
@@ -84,6 +86,22 @@ def __init__(
         self.freq = to_offset(freq)
         self.origin = origin
 
+        if not isinstance(self.freq, CFTIME_TICKS):
+            if offset is not None:
+                message = (
+                    "The 'offset' keyword does not take effect when "
+                    "resampling with a 'freq' that is not Tick-like (h, m, s, "
+                    "ms, us)"
+                )
+                emit_user_level_warning(message, category=RuntimeWarning)
+            if origin != "start_day":
+                message = (
+                    "The 'origin' keyword does not take effect when "
+                    "resampling with a 'freq' that is not Tick-like (h, m, s, "
+                    "ms, us)"
+                )
+                emit_user_level_warning(message, category=RuntimeWarning)
+
         if isinstance(self.freq, MonthEnd | QuarterEnd | YearEnd) or self.origin in [
             "end",
             "end_day",

xarray/tests/test_cftime_offsets.py

@@ -445,7 +445,6 @@ def test_eq(a, b):
     (Second(), 3, Second(n=3)),
     (Millisecond(), 3, Millisecond(n=3)),
     (Microsecond(), 3, Microsecond(n=3)),
-    (Day(), 0.5, Hour(n=12)),
     (Hour(), 0.5, Minute(n=30)),
     (Hour(), -0.5, Minute(n=-30)),
     (Minute(), 0.5, Second(n=30)),
@@ -534,6 +533,20 @@ def test_add_sub_monthly(offset, expected_date_args, calendar):
     assert result == expected
 
 
+def test_add_daily_offsets() -> None:
+    offset = Day(n=2)
+    expected = Day(n=4)
+    result = offset + offset
+    assert result == expected
+
+
+def test_subtract_daily_offsets() -> None:
+    offset = Day(n=2)
+    expected = Day(n=0)
+    result = offset - offset
+    assert result == expected
+
+
 @pytest.mark.parametrize(("offset", "expected_date_args"), _ADD_TESTS, ids=_id_func)
 def test_radd_sub_monthly(offset, expected_date_args, calendar):
     date_type = get_date_type(calendar)

xarray/tests/test_cftimeindex.py

@@ -830,7 +830,6 @@ def test_cftimeindex_add_timedeltaindex(calendar) -> None:
 @pytest.mark.parametrize(
     "freq,units",
     [
-        ("D", "D"),
         ("h", "h"),
         ("min", "min"),
         ("s", "s"),

xarray/tests/test_cftimeindex_resample.py

@@ -6,12 +6,17 @@
 import numpy as np
 import pandas as pd
 import pytest
-from packaging.version import Version
 
 import xarray as xr
-from xarray.coding.cftime_offsets import _new_to_legacy_freq
+from xarray.coding.cftime_offsets import (
+    CFTIME_TICKS,
+    Day,
+    _new_to_legacy_freq,
+    to_offset,
+)
 from xarray.coding.cftimeindex import CFTimeIndex
 from xarray.core.resample_cftime import CFTimeGrouper
+from xarray.tests import has_pandas_3
 
 cftime = pytest.importorskip("cftime")
 
@@ -54,6 +59,20 @@
 ]
 
 
+def has_tick_resample_freq(freqs):
+    resample_freq, _ = freqs
+    resample_freq_as_offset = to_offset(resample_freq)
+    return isinstance(resample_freq_as_offset, CFTIME_TICKS)
+
+
+def has_non_tick_resample_freq(freqs):
+    return not has_tick_resample_freq(freqs)
+
+
+FREQS_WITH_TICK_RESAMPLE_FREQ = list(filter(has_tick_resample_freq, FREQS))
+FREQS_WITH_NON_TICK_RESAMPLE_FREQ = list(filter(has_non_tick_resample_freq, FREQS))
+
+
 def compare_against_pandas(
     da_datetimeindex,
     da_cftimeindex,
@@ -110,22 +129,14 @@ def da(index) -> xr.DataArray:
     )
 
 
-@pytest.mark.parametrize("freqs", FREQS, ids=lambda x: "{}->{}".format(*x))
+@pytest.mark.parametrize(
+    "freqs", FREQS_WITH_TICK_RESAMPLE_FREQ, ids=lambda x: "{}->{}".format(*x)
+)
 @pytest.mark.parametrize("closed", [None, "left", "right"])
 @pytest.mark.parametrize("label", [None, "left", "right"])
 @pytest.mark.parametrize("offset", [None, "5s"], ids=lambda x: f"{x}")
-def test_resample(freqs, closed, label, offset) -> None:
+def test_resample_with_tick_resample_freq(freqs, closed, label, offset) -> None:
     initial_freq, resample_freq = freqs
-    if (
-        resample_freq == "4001D"
-        and closed == "right"
-        and Version(pd.__version__) < Version("2.2")
-    ):
-        pytest.skip(
-            "Pandas fixed a bug in this test case in version 2.2, which we "
-            "ported to xarray, so this test no longer produces the same "
-            "result as pandas for earlier pandas versions."
-        )
     start = "2000-01-01T12:07:01"
     origin = "start"
 
@@ -149,6 +160,43 @@ def test_resample(freqs, closed, label, offset) -> None:
     )
 
 
+@pytest.mark.parametrize(
+    "freqs", FREQS_WITH_NON_TICK_RESAMPLE_FREQ, ids=lambda x: "{}->{}".format(*x)
+)
+@pytest.mark.parametrize("closed", [None, "left", "right"])
+@pytest.mark.parametrize("label", [None, "left", "right"])
+def test_resample_with_non_tick_resample_freq(freqs, closed, label) -> None:
+    initial_freq, resample_freq = freqs
+    resample_freq_as_offset = to_offset(resample_freq)
+    if isinstance(resample_freq_as_offset, Day) and not has_pandas_3:
+        pytest.skip("Only valid for pandas >= 3.0")
+    start = "2000-01-01T12:07:01"
+
+    # Set offset and origin to their default values since they have no effect
+    # on resampling data with a non-tick resample frequency.
+    offset = None
+    origin = "start_day"
+
+    datetime_index = pd.date_range(
+        start=start, periods=5, freq=_new_to_legacy_freq(initial_freq)
+    )
+    cftime_index = xr.date_range(
+        start=start, periods=5, freq=initial_freq, use_cftime=True
+    )
+    da_datetimeindex = da(datetime_index)
+    da_cftimeindex = da(cftime_index)
+
+    compare_against_pandas(
+        da_datetimeindex,
+        da_cftimeindex,
+        resample_freq,
+        closed=closed,
+        label=label,
+        offset=offset,
+        origin=origin,
+    )
+
+
 @pytest.mark.parametrize(
     ("freq", "expected"),
     [
@@ -225,19 +273,28 @@ def test_origin(closed, origin) -> None:
 
 @pytest.mark.parametrize("offset", ["foo", "5MS", 10])
 def test_invalid_offset_error(offset: str | int) -> None:
-    cftime_index = xr.date_range("2000", periods=5, use_cftime=True)
+    cftime_index = xr.date_range("2000", periods=5, freq="h", use_cftime=True)
     da_cftime = da(cftime_index)
     with pytest.raises(ValueError, match="offset must be"):
-        da_cftime.resample(time="2D", offset=offset)  # type: ignore[arg-type]
+        da_cftime.resample(time="2h", offset=offset)  # type: ignore[arg-type]
 
 
 def test_timedelta_offset() -> None:
     timedelta = datetime.timedelta(seconds=5)
     string = "5s"
 
-    cftime_index = xr.date_range("2000", periods=5, use_cftime=True)
+    cftime_index = xr.date_range("2000", periods=5, freq="h", use_cftime=True)
     da_cftime = da(cftime_index)
 
-    timedelta_result = da_cftime.resample(time="2D", offset=timedelta).mean()
-    string_result = da_cftime.resample(time="2D", offset=string).mean()
+    timedelta_result = da_cftime.resample(time="2h", offset=timedelta).mean()
+    string_result = da_cftime.resample(time="2h", offset=string).mean()
     xr.testing.assert_identical(timedelta_result, string_result)
+
+
+@pytest.mark.parametrize(("option", "value"), [("offset", "5s"), ("origin", "start")])
+def test_non_tick_option_warning(option, value) -> None:
+    cftime_index = xr.date_range("2000", periods=5, use_cftime=True)
+    da_cftime = da(cftime_index)
+    kwargs = {option: value}
+    with pytest.warns(RuntimeWarning, match=option):
+        da_cftime.resample(time="ME", **kwargs)