Add isocalendar API support

python/cudf/cudf/core/index.py

@@ -1616,6 +1616,40 @@ def quarter(self):
         res = extract_quarter(self._values)
         return Int8Index(res, dtype="int8")
 
+    def isocalendar(self):
+        """
+        Returns a DataFrame with the year, week, and day
+        calculated according to the ISO 8601 standard.
+        Returns
+        -------
+        DataFrame
+            with columns year, week and day
+        Examples
+        --------
+
+        >>> gIndex = cudf.DatetimeIndex(["2020-05-31 08:00:00",
+        ...    "1999-12-31 18:40:00"])
+        >>> gIndex.isocalendar()
+                             year  week  day
+        2020-05-31 08:00:00  2020    22    7
+        1999-12-31 18:40:00  1999    52    5
+        """
+        indexSeries = cudf.core.tools.datetimes.to_iso_calendar(self)
+
+        @property
+        def day(self):
+            return indexSeries["day"]
+
+        @property
+        def week(self):
+            return indexSeries["week"]
+
+        @property
+        def year(self):
+            return indexSeries["year"]
+
+        return indexSeries
+
     def to_pandas(self):
         nanos = self._values.astype("datetime64[ns]")
         return pd.DatetimeIndex(nanos.to_pandas(), name=self.name)

python/cudf/cudf/core/series.py

@@ -5503,6 +5503,61 @@ def quarter(self):
             {None: res}, index=self.series._index, name=self.series.name,
         )
 
+    def isocalendar(self):
+        """
+        Returns a DataFrame with the year, week, and day
+        calculated according to the ISO 8601 standard.
+        Returns
+        -------
+        DataFrame
+            with columns year, week and day
+        Examples
+        --------
+        >>> ser = cudf.Series(pd.date_range(start="2021-07-25",
+        ... end="2021-07-30"))
+        >>> ser.dt.isocalendar()
+        year  week  day
+        0  2021    29    7
+        1  2021    30    1
+        2  2021    30    2
+        3  2021    30    3
+        4  2021    30    4
+        5  2021    30    5
+        >>> ser.dt.isocalendar().week
+        0    29
+        1    30
+        2    30
+        3    30
+        4    30
+        5    30
+        Name: week, dtype: object
+
+        >>> serIndex = cudf.to_datetime(pd.Series(["2010-01-01", pd.NaT]))
+        >>> serIndex.dt.isocalendar()
+           year  week  day
+        0  2009    53     5
+        1  <NA>  <NA>  <NA>
+        >>> serIndex.dt.isocalendar().year
+        0    2009
+        1    <NA>
+        Name: year, dtype: object
+        """
+        isoSeries = cudf.core.tools.datetimes.to_iso_calendar(self)
+
+        @property
+        def day(self):
+            return isoSeries["day"]
+
+        @property
+        def week(self):
+            return isoSeries["week"]
+
+        @property
+        def year(self):
+            return isoSeries["year"]
+
+        return isoSeries
+
     @property
     def is_month_start(self):
         """
@@ -5806,9 +5861,8 @@ def strftime(self, date_format, *args, **kwargs):
         Notes
         -----
 
-        The following date format identifiers are not yet supported: ``%a``,
-        ``%A``, ``%w``, ``%b``, ``%B``, ``%U``, ``%W``, ``%c``, ``%x``,
-        ``%X``, ``%G``, ``%u``, ``%V``
+        The following date format identifiers are not yet
+        supported: ``%c``, ``%x``,``%X``
 
         Examples
         --------
@@ -5847,19 +5901,9 @@ def strftime(self, date_format, *args, **kwargs):
         # once https://github.com/rapidsai/cudf/issues/5991
         # is implemented
         not_implemented_formats = {
-            "%a",
-            "%A",
-            "%w",
-            "%b",
-            "%B",
-            "%U",
-            "%W",
             "%c",
             "%x",
             "%X",
-            "%G",
-            "%u",
-            "%V",
         }
         for d_format in not_implemented_formats:
             if d_format in date_format:
@@ -5869,7 +5913,6 @@ def strftime(self, date_format, *args, **kwargs):
                     f"https://github.com/rapidsai/cudf/issues/5991 "
                     f"for tracking purposes."
                 )
-
         str_col = self.series._column.as_string_column(
             dtype="str", format=date_format
         )

python/cudf/cudf/core/tools/datetimes.py

@@ -4,6 +4,7 @@
 from typing import Sequence, Union
 
 import numpy as np
+import pandas as pd
 from pandas.core.tools.datetimes import _unit_map
 
 import cudf
@@ -221,8 +222,8 @@ def to_datetime(
                 format=format,
             )
             return as_index(col, name=arg.name)
-        elif isinstance(arg, cudf.Series):
-            col = arg._column
+        elif isinstance(arg, (cudf.Series, pd.Series)):
+            col = column.as_column(arg)
             col = _process_col(
                 col=col,
                 unit=unit,
@@ -652,3 +653,26 @@ def _isin_datetimelike(
 
     res = lhs._obtain_isin_result(rhs)
     return res
+
+
+def to_iso_calendar(self):
+    formats = ["%G", "%V", "%u"]
+    if isinstance(
+        self,
+        (
+            cudf.core.series.DatetimeProperties,
+            pd.core.indexes.accessors.DatetimeProperties,
+        ),
+    ):
+        iso_params = [self.strftime(fmt) for fmt in formats]
+        index = None
+    elif isinstance(self, (cudf.Index, pd.Index)):
+        iso_params = [
+            self._values.as_string_column(self._values.dtype, fmt)
+            for fmt in formats
+        ]
+        index = self._values
+    data = dict(zip(["year", "week", "day"], iso_params))
+    isoSeries = cudf.DataFrame(data, index=index, dtype=np.int32)
+
+    return isoSeries

python/cudf/cudf/tests/test_datetime.py

@@ -1143,24 +1143,7 @@ def test_datetime_strftime(data, dtype, date_format):
     assert_eq(expected, actual)
 
 
-@pytest.mark.parametrize(
-    "date_format",
-    [
-        "%a",
-        "%A",
-        "%w",
-        "%b",
-        "%B",
-        "%U",
-        "%W",
-        "%c",
-        "%x",
-        "%X",
-        "%G",
-        "%u",
-        "%V",
-    ],
-)
+@pytest.mark.parametrize("date_format", ["%c", "%x", "%X"])
 def test_datetime_strftime_not_implemented_formats(date_format):
     gsr = cudf.Series([1, 2, 3], dtype="datetime64[ms]")
 
@@ -1334,6 +1317,57 @@ def test_quarter():
     assert_eq(expect2.values, got2.values, check_dtype=False)
 
 
+def test_isocalendar():
+    data = [
+        "2020-05-31 08:00:00",
+        "1999-12-31 18:40:00",
+        "2000-12-31 04:00:00",
+        "1900-02-28 07:00:00",
+        "1800-03-14 07:30:00",
+        "2100-03-14 07:30:00",
+        "1970-01-01 00:00:00",
+        "1969-12-31 12:59:00",
+    ]
+    dtype = "datetime64[s]"
+
+    # Series
+    ps = pd.Series(data, dtype=dtype)
+    gs = cudf.from_pandas(ps)
+
+    expect = ps.dt.isocalendar()
+    got = gs.dt.isocalendar()
+
+    assert_eq(expect, got, check_dtype=False)
+
+    # Series day
+    expectday = ps.dt.isocalendar().day
+    gotday = gs.dt.isocalendar().day
+
+    assert_eq(expectday, gotday, check_dtype=False)
+
+    # Series week
+    expectweek = ps.dt.isocalendar().week
+    gotweek = gs.dt.isocalendar().week
+
+    assert_eq(expectweek, gotweek, check_dtype=False)
+
+    # Series year
+    expectyear = ps.dt.isocalendar().year
+    gotyear = gs.dt.isocalendar().year
+
+    assert_eq(expectyear, gotyear, check_dtype=False)
+
+    # DatetimeIndex
+    pIndex = pd.DatetimeIndex(data)
+    gIndex = cudf.from_pandas(pIndex)
+
+    expect2 = pIndex.isocalendar()
+    got2 = gIndex.isocalendar()
+
+    # assert isinstance(got2, cudf.Int8Index)
+    assert_eq(expect2.values, got2.values, check_dtype=False)
+
+
 @pytest.mark.parametrize("dtype", DATETIME_TYPES)
 def test_days_in_months(dtype):
     nrows = 1000