CLN: enforce deprecation of frequencies deprecated for offsets (#57986)

* enforce deprecation of offset deprecated freqstr * fix tests * fix mypy error * add a note to v3.0.0 * remove c_REVERSE_OFFSET_REMOVED_FREQSTR, correct tests * fix test_to_period_offsets_not_supported * correct to_offset, fix tests * add dict PERIOD_TO_OFFSET_FREQSTR, corect meth to_offset, fix tests * add a comment * create dictionary PERIOD_AND_OFFSET_ALIASES * correct def to_offset * fixup * fixup * replace c_OFFSET_RENAMED_FREQSTR with c_PERIOD_TO_OFFSET_FREQSTR * add cimport c_PERIOD_TO_OFFSET_FREQSTR * add a helper function for error reporting * minor whatsnew fix --------- Co-authored-by: Marco Edward Gorelli <[email protected]> Co-authored-by: Marco Gorelli <[email protected]>
pandas-dev · Jun 7, 2024 · c95716a · c95716a
1 parent ee127fc
commit c95716a
Show file tree

Hide file tree

Showing 17 changed files with 274 additions and 301 deletions.
diff --git a/doc/source/whatsnew/v3.0.0.rst b/doc/source/whatsnew/v3.0.0.rst
@@ -280,6 +280,34 @@ Other Deprecations
 
 Removal of prior version deprecations/changes
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+Enforced deprecation of aliases ``M``, ``Q``, ``Y``, etc. in favour of ``ME``, ``QE``, ``YE``, etc. for offsets
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+Renamed the following offset aliases (:issue:`57986`):
+
++-------------------------------+------------------+------------------+
+| offset                        | removed alias    | new alias        |
++===============================+==================+==================+
+|:class:`MonthEnd`              |      ``M``       |     ``ME``       |
++-------------------------------+------------------+------------------+
+|:class:`BusinessMonthEnd`      |      ``BM``      |     ``BME``      |
++-------------------------------+------------------+------------------+
+|:class:`SemiMonthEnd`          |      ``SM``      |     ``SME``      |
++-------------------------------+------------------+------------------+
+|:class:`CustomBusinessMonthEnd`|      ``CBM``     |     ``CBME``     |
++-------------------------------+------------------+------------------+
+|:class:`QuarterEnd`            |      ``Q``       |     ``QE``       |
++-------------------------------+------------------+------------------+
+|:class:`BQuarterEnd`           |      ``BQ``      |     ``BQE``      |
++-------------------------------+------------------+------------------+
+|:class:`YearEnd`               |      ``Y``       |     ``YE``       |
++-------------------------------+------------------+------------------+
+|:class:`BYearEnd`              |      ``BY``      |     ``BYE``      |
++-------------------------------+------------------+------------------+
+
+Other Removals
+^^^^^^^^^^^^^^
 - :class:`.DataFrameGroupBy.idxmin`, :class:`.DataFrameGroupBy.idxmax`, :class:`.SeriesGroupBy.idxmin`, and :class:`.SeriesGroupBy.idxmax` will now raise a ``ValueError`` when used with ``skipna=False`` and an NA value is encountered (:issue:`10694`)
 - :func:`concat` no longer ignores empty objects when determining output dtypes (:issue:`39122`)
 - :func:`concat` with all-NA entries no longer ignores the dtype of those entries when determining the result dtype (:issue:`40893`)
@@ -343,7 +371,7 @@ Removal of prior version deprecations/changes
 - Enforced deprecation of string ``A`` denoting frequency in :class:`YearEnd` and strings ``A-DEC``, ``A-JAN``, etc. denoting annual frequencies with various fiscal year ends (:issue:`57699`)
 - Enforced deprecation of string ``BAS`` denoting frequency in :class:`BYearBegin` and strings ``BAS-DEC``, ``BAS-JAN``, etc. denoting annual frequencies with various fiscal year starts (:issue:`57793`)
 - Enforced deprecation of string ``BA`` denoting frequency in :class:`BYearEnd` and strings ``BA-DEC``, ``BA-JAN``, etc. denoting annual frequencies with various fiscal year ends (:issue:`57793`)
-- Enforced deprecation of strings ``T``, ``L``, ``U``, and ``N`` denoting frequencies in :class:`Minute`, :class:`Second`, :class:`Milli`, :class:`Micro`, :class:`Nano` (:issue:`57627`)
+- Enforced deprecation of strings ``T``, ``L``, ``U``, and ``N`` denoting frequencies in :class:`Minute`, :class:`Milli`, :class:`Micro`, :class:`Nano` (:issue:`57627`)
 - Enforced deprecation of strings ``T``, ``L``, ``U``, and ``N`` denoting units in :class:`Timedelta` (:issue:`57627`)
 - Enforced deprecation of the behavior of :func:`concat` when ``len(keys) != len(objs)`` would truncate to the shorter of the two. Now this raises a ``ValueError`` (:issue:`43485`)
 - Enforced deprecation of values "pad", "ffill", "bfill", and "backfill" for :meth:`Series.interpolate` and :meth:`DataFrame.interpolate` (:issue:`57869`)

diff --git a/pandas/_libs/tslibs/dtypes.pxd b/pandas/_libs/tslibs/dtypes.pxd
@@ -12,9 +12,10 @@ cdef NPY_DATETIMEUNIT get_supported_reso(NPY_DATETIMEUNIT reso)
 cdef bint is_supported_unit(NPY_DATETIMEUNIT reso)
 
 cdef dict c_OFFSET_TO_PERIOD_FREQSTR
-cdef dict c_OFFSET_DEPR_FREQSTR
-cdef dict c_REVERSE_OFFSET_DEPR_FREQSTR
+cdef dict c_PERIOD_TO_OFFSET_FREQSTR
+cdef dict c_OFFSET_RENAMED_FREQSTR
 cdef dict c_DEPR_ABBREVS
+cdef dict c_PERIOD_AND_OFFSET_DEPR_FREQSTR
 cdef dict attrname_to_abbrevs
 cdef dict npy_unit_to_attrname
 cdef dict attrname_to_npy_unit

diff --git a/pandas/_libs/tslibs/dtypes.pyx b/pandas/_libs/tslibs/dtypes.pyx
@@ -176,6 +176,10 @@ OFFSET_TO_PERIOD_FREQSTR: dict = {
     "EOM": "M",
     "BME": "M",
     "SME": "M",
+    "BMS": "M",
+    "CBME": "M",
+    "CBMS": "M",
+    "SMS": "M",
     "BQS": "Q",
     "QS": "Q",
     "BQE": "Q",
@@ -228,7 +232,6 @@ OFFSET_TO_PERIOD_FREQSTR: dict = {
     "YE-NOV": "Y-NOV",
     "W": "W",
     "ME": "M",
-    "Y": "Y",
     "BYE": "Y",
     "BYE-DEC": "Y-DEC",
     "BYE-JAN": "Y-JAN",
@@ -245,7 +248,7 @@ OFFSET_TO_PERIOD_FREQSTR: dict = {
     "YS": "Y",
     "BYS": "Y",
 }
-cdef dict c_OFFSET_DEPR_FREQSTR = {
+cdef dict c_OFFSET_RENAMED_FREQSTR = {
     "M": "ME",
     "Q": "QE",
     "Q-DEC": "QE-DEC",
@@ -303,10 +306,37 @@ cdef dict c_OFFSET_DEPR_FREQSTR = {
     "BQ-OCT": "BQE-OCT",
     "BQ-NOV": "BQE-NOV",
 }
-cdef dict c_OFFSET_TO_PERIOD_FREQSTR = OFFSET_TO_PERIOD_FREQSTR
-cdef dict c_REVERSE_OFFSET_DEPR_FREQSTR = {
-    v: k for k, v in c_OFFSET_DEPR_FREQSTR.items()
+PERIOD_TO_OFFSET_FREQSTR = {
+    "M": "ME",
+    "Q": "QE",
+    "Q-DEC": "QE-DEC",
+    "Q-JAN": "QE-JAN",
+    "Q-FEB": "QE-FEB",
+    "Q-MAR": "QE-MAR",
+    "Q-APR": "QE-APR",
+    "Q-MAY": "QE-MAY",
+    "Q-JUN": "QE-JUN",
+    "Q-JUL": "QE-JUL",
+    "Q-AUG": "QE-AUG",
+    "Q-SEP": "QE-SEP",
+    "Q-OCT": "QE-OCT",
+    "Q-NOV": "QE-NOV",
+    "Y": "YE",
+    "Y-DEC": "YE-DEC",
+    "Y-JAN": "YE-JAN",
+    "Y-FEB": "YE-FEB",
+    "Y-MAR": "YE-MAR",
+    "Y-APR": "YE-APR",
+    "Y-MAY": "YE-MAY",
+    "Y-JUN": "YE-JUN",
+    "Y-JUL": "YE-JUL",
+    "Y-AUG": "YE-AUG",
+    "Y-SEP": "YE-SEP",
+    "Y-OCT": "YE-OCT",
+    "Y-NOV": "YE-NOV",
 }
+cdef dict c_OFFSET_TO_PERIOD_FREQSTR = OFFSET_TO_PERIOD_FREQSTR
+cdef dict c_PERIOD_TO_OFFSET_FREQSTR = PERIOD_TO_OFFSET_FREQSTR
 
 # Map deprecated resolution abbreviations to correct resolution abbreviations
 cdef dict c_DEPR_ABBREVS = {
@@ -316,6 +346,11 @@ cdef dict c_DEPR_ABBREVS = {
     "S": "s",
 }
 
+cdef dict c_PERIOD_AND_OFFSET_DEPR_FREQSTR = {
+    "w": "W",
+    "MIN": "min",
+}
+
 
 class FreqGroup(Enum):
     # Mirrors c_FreqGroup in the .pxd file

diff --git a/pandas/_libs/tslibs/offsets.pyx b/pandas/_libs/tslibs/offsets.pyx
@@ -57,8 +57,10 @@ from pandas._libs.tslibs.ccalendar cimport (
 from pandas._libs.tslibs.conversion cimport localize_pydatetime
 from pandas._libs.tslibs.dtypes cimport (
     c_DEPR_ABBREVS,
-    c_OFFSET_DEPR_FREQSTR,
-    c_REVERSE_OFFSET_DEPR_FREQSTR,
+    c_OFFSET_RENAMED_FREQSTR,
+    c_OFFSET_TO_PERIOD_FREQSTR,
+    c_PERIOD_AND_OFFSET_DEPR_FREQSTR,
+    c_PERIOD_TO_OFFSET_FREQSTR,
     periods_per_day,
 )
 from pandas._libs.tslibs.nattype cimport (
@@ -4711,6 +4713,34 @@ INVALID_FREQ_ERR_MSG = "Invalid frequency: {0}"
 _offset_map = {}
 
 
+def _validate_to_offset_alias(alias: str, is_period: bool) -> None:
+    if not is_period:
+        if alias.upper() in c_OFFSET_RENAMED_FREQSTR:
+            raise ValueError(
+                f"\'{alias}\' is no longer supported for offsets. Please "
+                f"use \'{c_OFFSET_RENAMED_FREQSTR.get(alias.upper())}\' "
+                f"instead."
+            )
+        if (alias.upper() != alias and
+                alias.lower() not in {"s", "ms", "us", "ns"} and
+                alias.upper().split("-")[0].endswith(("S", "E"))):
+            raise ValueError(INVALID_FREQ_ERR_MSG.format(alias))
+    if (is_period and
+            alias.upper() in c_OFFSET_TO_PERIOD_FREQSTR and
+            alias != "ms" and
+            alias.upper().split("-")[0].endswith(("S", "E"))):
+        if (alias.upper().startswith("B") or
+                alias.upper().startswith("S") or
+                alias.upper().startswith("C")):
+            raise ValueError(INVALID_FREQ_ERR_MSG.format(alias))
+        else:
+            alias_msg = "".join(alias.upper().split("E", 1))
+            raise ValueError(
+                f"for Period, please use \'{alias_msg}\' "
+                f"instead of \'{alias}\'"
+            )
+
+
 # TODO: better name?
 def _get_offset(name: str) -> BaseOffset:
     """
@@ -4850,54 +4880,26 @@ cpdef to_offset(freq, bint is_period=False):
 
             tups = zip(split[0::4], split[1::4], split[2::4])
             for n, (sep, stride, name) in enumerate(tups):
-                if not is_period and name.upper() in c_OFFSET_DEPR_FREQSTR:
-                    warnings.warn(
-                        f"\'{name}\' is deprecated and will be removed "
-                        f"in a future version, please use "
-                        f"\'{c_OFFSET_DEPR_FREQSTR.get(name.upper())}\' instead.",
-                        FutureWarning,
-                        stacklevel=find_stack_level(),
-                    )
-                    name = c_OFFSET_DEPR_FREQSTR[name.upper()]
-                if (not is_period and
-                        name != name.upper() and
-                        name.lower() not in {"s", "ms", "us", "ns"} and
-                        name.upper().split("-")[0].endswith(("S", "E"))):
-                    warnings.warn(
-                        f"\'{name}\' is deprecated and will be removed "
-                        f"in a future version, please use "
-                        f"\'{name.upper()}\' instead.",
-                        FutureWarning,
-                        stacklevel=find_stack_level(),
-                    )
-                    name = name.upper()
-                if is_period and name.upper() in c_REVERSE_OFFSET_DEPR_FREQSTR:
-                    if name.upper().startswith("Y"):
-                        raise ValueError(
-                            f"for Period, please use \'Y{name.upper()[2:]}\' "
-                            f"instead of \'{name}\'"
-                        )
-                    if (name.upper().startswith("B") or
-                            name.upper().startswith("S") or
-                            name.upper().startswith("C")):
-                        raise ValueError(INVALID_FREQ_ERR_MSG.format(name))
-                    else:
-                        raise ValueError(
-                            f"for Period, please use "
-                            f"\'{c_REVERSE_OFFSET_DEPR_FREQSTR.get(name.upper())}\' "
-                            f"instead of \'{name}\'"
-                        )
-                elif is_period and name.upper() in c_OFFSET_DEPR_FREQSTR:
-                    if name.upper() != name:
+                _validate_to_offset_alias(name, is_period)
+                if is_period:
+                    if name.upper() in c_PERIOD_TO_OFFSET_FREQSTR:
+                        if name.upper() != name:
+                            raise ValueError(
+                                f"\'{name}\' is no longer supported, "
+                                f"please use \'{name.upper()}\' instead.",
+                            )
+                        name = c_PERIOD_TO_OFFSET_FREQSTR.get(name.upper())
+
+                    if name in c_PERIOD_AND_OFFSET_DEPR_FREQSTR:
                         warnings.warn(
-                            f"\'{name}\' is deprecated and will be removed in "
-                            f"a future version, please use \'{name.upper()}\' "
-                            f"instead.",
+                            f"\'{name}\' is deprecated and will be removed "
+                            f"in a future version, please use "
+                            f"\'{c_PERIOD_AND_OFFSET_DEPR_FREQSTR.get(name)}\' "
+                            f" instead.",
                             FutureWarning,
                             stacklevel=find_stack_level(),
-                        )
-                    name = c_OFFSET_DEPR_FREQSTR.get(name.upper())
-
+                            )
+                        name = c_PERIOD_AND_OFFSET_DEPR_FREQSTR.get(name)
                 if sep != "" and not sep.isspace():
                     raise ValueError("separator must be spaces")
                 prefix = _lite_rule_alias.get(name) or name

diff --git a/pandas/tests/arrays/test_datetimes.py b/pandas/tests/arrays/test_datetimes.py
@@ -764,29 +764,14 @@ def test_iter_zoneinfo_fold(self, tz):
         assert left.utcoffset() == right2.utcoffset()
 
     @pytest.mark.parametrize(
-        "freq, freq_depr",
-        [
-            ("2ME", "2M"),
-            ("2SME", "2SM"),
-            ("2SME", "2sm"),
-            ("2QE", "2Q"),
-            ("2QE-SEP", "2Q-SEP"),
-            ("1YE", "1Y"),
-            ("2YE-MAR", "2Y-MAR"),
-            ("2ME", "2m"),
-            ("2QE-SEP", "2q-sep"),
-            ("2YE", "2y"),
-        ],
+        "freq",
+        ["2M", "2SM", "2sm", "2Q", "2Q-SEP", "1Y", "2Y-MAR", "2m", "2q-sep", "2y"],
     )
-    def test_date_range_frequency_M_Q_Y_A_deprecated(self, freq, freq_depr):
-        # GH#9586, GH#54275
-        depr_msg = f"'{freq_depr[1:]}' is deprecated and will be removed "
-        f"in a future version, please use '{freq[1:]}' instead."
+    def test_date_range_frequency_M_Q_Y_raises(self, freq):
+        msg = f"Invalid frequency: {freq}"
 
-        expected = pd.date_range("1/1/2000", periods=4, freq=freq)
-        with tm.assert_produces_warning(FutureWarning, match=depr_msg):
-            result = pd.date_range("1/1/2000", periods=4, freq=freq_depr)
-        tm.assert_index_equal(result, expected)
+        with pytest.raises(ValueError, match=msg):
+            pd.date_range("1/1/2000", periods=4, freq=freq)
 
     @pytest.mark.parametrize("freq_depr", ["2H", "2CBH", "2MIN", "2S", "2mS", "2Us"])
     def test_date_range_uppercase_frequency_deprecated(self, freq_depr):
@@ -800,7 +785,7 @@ def test_date_range_uppercase_frequency_deprecated(self, freq_depr):
         tm.assert_index_equal(result, expected)
 
     @pytest.mark.parametrize(
-        "freq_depr",
+        "freq",
         [
             "2ye-mar",
             "2ys",
@@ -811,17 +796,21 @@ def test_date_range_uppercase_frequency_deprecated(self, freq_depr):
             "2bms",
             "2cbme",
             "2me",
-            "2w",
         ],
     )
-    def test_date_range_lowercase_frequency_deprecated(self, freq_depr):
+    def test_date_range_lowercase_frequency_raises(self, freq):
+        msg = f"Invalid frequency: {freq}"
+
+        with pytest.raises(ValueError, match=msg):
+            pd.date_range("1/1/2000", periods=4, freq=freq)
+
+    def test_date_range_lowercase_frequency_deprecated(self):
         # GH#9586, GH#54939
-        depr_msg = f"'{freq_depr[1:]}' is deprecated and will be removed in a "
-        f"future version, please use '{freq_depr.upper()[1:]}' instead."
+        depr_msg = "'w' is deprecated and will be removed in a future version"
 
-        expected = pd.date_range("1/1/2000", periods=4, freq=freq_depr.upper())
+        expected = pd.date_range("1/1/2000", periods=4, freq="2W")
         with tm.assert_produces_warning(FutureWarning, match=depr_msg):
-            result = pd.date_range("1/1/2000", periods=4, freq=freq_depr)
+            result = pd.date_range("1/1/2000", periods=4, freq="2w")
         tm.assert_index_equal(result, expected)
 
     @pytest.mark.parametrize("freq", ["1A", "2A-MAR", "2a-mar"])

diff --git a/pandas/tests/frame/methods/test_asfreq.py b/pandas/tests/frame/methods/test_asfreq.py
@@ -236,32 +236,30 @@ def test_asfreq_2ME(self, freq, freq_half):
         "freq, freq_depr",
         [
             ("2ME", "2M"),
+            ("2ME", "2m"),
             ("2QE", "2Q"),
             ("2QE-SEP", "2Q-SEP"),
             ("1BQE", "1BQ"),
             ("2BQE-SEP", "2BQ-SEP"),
-            ("1YE", "1Y"),
+            ("2BQE-SEP", "2bq-sep"),
+            ("1YE", "1y"),
             ("2YE-MAR", "2Y-MAR"),
         ],
     )
-    def test_asfreq_frequency_M_Q_Y_deprecated(self, freq, freq_depr):
-        # GH#9586, #55978
-        depr_msg = f"'{freq_depr[1:]}' is deprecated and will be removed "
-        f"in a future version, please use '{freq[1:]}' instead."
+    def test_asfreq_frequency_M_Q_Y_raises(self, freq, freq_depr):
+        msg = f"Invalid frequency: {freq_depr}"
 
         index = date_range("1/1/2000", periods=4, freq=f"{freq[1:]}")
         df = DataFrame({"s": Series([0.0, 1.0, 2.0, 3.0], index=index)})
-        expected = df.asfreq(freq=freq)
-        with tm.assert_produces_warning(FutureWarning, match=depr_msg):
-            result = df.asfreq(freq=freq_depr)
-        tm.assert_frame_equal(result, expected)
+        with pytest.raises(ValueError, match=msg):
+            df.asfreq(freq=freq_depr)
 
     @pytest.mark.parametrize(
         "freq, error_msg",
         [
             (
                 "2MS",
-                "MS is not supported as period frequency",
+                "Invalid frequency: 2MS",
             ),
             (
                 offsets.MonthBegin(),