DEPR: lower/uppercase strings such as 'y', 'q', 'H', 'MIN', etc. denoting freqs/units for time series, period, and timedelta

doc/source/user_guide/10min.rst

@@ -612,7 +612,7 @@ financial applications. See the :ref:`Time Series section <timeseries>`.
 
    rng = pd.date_range("1/1/2012", periods=100, freq="s")
    ts = pd.Series(np.random.randint(0, 500, len(rng)), index=rng)
-   ts.resample("5Min").sum()
+   ts.resample("5min").sum()
 
 :meth:`Series.tz_localize` localizes a time series to a time zone:
 

doc/source/user_guide/timeseries.rst

@@ -1645,7 +1645,7 @@ Basics
 
    ts = pd.Series(np.random.randint(0, 500, len(rng)), index=rng)
 
-   ts.resample("5Min").sum()
+   ts.resample("5min").sum()
 
 The ``resample`` function is very flexible and allows you to specify many
 different parameters to control the frequency conversion and resampling
@@ -1657,31 +1657,31 @@ a method of the returned object, including ``sum``, ``mean``, ``std``, ``sem``,
 
 .. ipython:: python
 
-   ts.resample("5Min").mean()
+   ts.resample("5min").mean()
 
-   ts.resample("5Min").ohlc()
+   ts.resample("5min").ohlc()
 
-   ts.resample("5Min").max()
+   ts.resample("5min").max()
 
 
 For downsampling, ``closed`` can be set to 'left' or 'right' to specify which
 end of the interval is closed:
 
 .. ipython:: python
 
-   ts.resample("5Min", closed="right").mean()
+   ts.resample("5min", closed="right").mean()
 
-   ts.resample("5Min", closed="left").mean()
+   ts.resample("5min", closed="left").mean()
 
 Parameters like ``label`` are used to manipulate the resulting labels.
 ``label`` specifies whether the result is labeled with the beginning or
 the end of the interval.
 
 .. ipython:: python
 
-   ts.resample("5Min").mean()  # by default label='left'
+   ts.resample("5min").mean()  # by default label='left'
 
-   ts.resample("5Min", label="left").mean()
+   ts.resample("5min", label="left").mean()
 
 .. warning::
 

doc/source/whatsnew/v2.2.0.rst

@@ -619,6 +619,41 @@ For example:
 
     pd.date_range('2020-01-01', periods=3, freq='QE-NOV')
 
+Deprecate lowercase strings ``'w'``, ``'m'``, ``'q'``, etc. and uppercase strings ``'H'``, ``'MIN'``, ``'S'``, etc. for time series, period, and timedelta
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+Previously, both ``'h'`` and ``'H'`` were allowed for 'hour' offset/period alias. We now require the case to be correct - check the :ref:`offset aliases <timeseries.offset_aliases>` and the :ref:`period aliases <timeseries.period_aliases>` parts of the docs and make sure you're using the correct one (:issue:`56346`)
+
+Deprecated lowercase strings in favour of uppercase strings denoting:
+
+- period aliases weekly, monthly and bigger frequency
+- offsets aliases representing :class:`DateOffset` subclasses that are a week or bigger (``Week``, ``MonthBegin``, ``MonthEnd``, etc.)
+- timedelta units for week, month or bigger
+
+Deprecated uppercase strings in favour of lowercase strings denoting:
+
+- period aliases hourly, minutely or smaller frequency
+- offsets aliases representing :class:`DateOffset` subclasses that are an hour or smaller (``Hour``, ``Minute``, etc.)
+- timedelta units for hour, minute or smaller
+
+For example:
+
+*Previous behavior*:
+
+.. code-block:: ipython
+
+    In [9]: pd.date_range('2020-01-01', periods=3, freq='H')
+    Out[9]:
+    DatetimeIndex(['2020-01-01 00:00:00', '2020-01-01 01:00:00',
+                   '2020-01-01 02:00:00'],
+                  dtype='datetime64[ns]', freq='H')
+
+*Future behavior*:
+
+.. ipython:: python
+
+    pd.date_range('2020-01-01', periods=3, freq='h')
+
 Deprecated automatic downcasting
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 

pandas/_libs/tslibs/dtypes.pyx

@@ -407,8 +407,10 @@ cdef dict c_DEPR_ABBREVS = {
     "L": "ms",
     "l": "ms",
     "U": "us",
+    "US": "us",
     "u": "us",
     "N": "ns",
+    "NS": "ns",
     "n": "ns",
 }
 
@@ -509,7 +511,7 @@ class Resolution(Enum):
                 warnings.warn(
                     f"\'{freq}\' is deprecated and will be removed in a future "
                     f"version. Please use \'{abbrev}\' "
-                    "instead of \'{freq}\'.",
+                    f"instead of \'{freq}\'.",
                     FutureWarning,
                     stacklevel=find_stack_level(),
                 )

pandas/_libs/tslibs/offsets.pyx

@@ -1796,7 +1796,7 @@ cdef class BusinessDay(BusinessMixin):
                     s -= hrs * 3600
                 mts = int(s / 60)
                 if mts != 0:
-                    off_str += str(mts) + "Min"
+                    off_str += str(mts) + "min"
                     s -= mts * 60
                 if s != 0:
                     off_str += str(s) + "s"
@@ -4702,14 +4702,13 @@ _lite_rule_alias = {
     "BYE": "BYE-DEC",    # BYearEnd(month=12),
     "BYS": "BYS-JAN",  # BYearBegin(month=1),
 
-    "Min": "min",
     "min": "min",
     "ms": "ms",
     "us": "us",
     "ns": "ns",
 }
 
-_dont_uppercase = _dont_uppercase = {"h", "bh", "cbh", "MS", "ms", "s"}
+_dont_uppercase = {"h", "bh", "cbh", "MS", "ms", "s"}
 
 
 INVALID_FREQ_ERR_MSG = "Invalid frequency: {0}"
@@ -4913,7 +4912,15 @@ cpdef to_offset(freq, bint is_period=False):
                             stacklevel=find_stack_level(),
                         )
                     name = c_OFFSET_DEPR_FREQSTR.get(name.upper())
-
+                if name == "w"or name == "d":
+                    warnings.warn(
+                        f"\'{name}\' is deprecated and will be removed in "
+                        f"a future version, please use \'{name.upper()}\' "
+                        f"instead.",
+                        FutureWarning,
+                        stacklevel=find_stack_level(),
+                    )
+                    name = name.upper()
                 if sep != "" and not sep.isspace():
                     raise ValueError("separator must be spaces")
                 prefix = _lite_rule_alias.get(name) or name

pandas/_libs/tslibs/timedeltas.pyx

@@ -719,15 +719,24 @@ cpdef inline str parse_timedelta_unit(str unit):
         return "ns"
     elif unit == "M":
         return unit
-    elif unit in c_DEPR_ABBREVS:
+    elif unit.upper() == "MIN" and unit.lower() != unit:
         warnings.warn(
             f"\'{unit}\' is deprecated and will be removed in a "
-            f"future version. Please use \'{c_DEPR_ABBREVS.get(unit)}\' "
+            f"future version. Please use \'{unit.lower()}\' "
             f"instead of \'{unit}\'.",
             FutureWarning,
             stacklevel=find_stack_level(),
         )
-        unit = c_DEPR_ABBREVS[unit]
+        unit = unit.lower()
+    elif unit.upper() in c_DEPR_ABBREVS and unit != c_DEPR_ABBREVS.get(unit.upper()):
+        warnings.warn(
+            f"\'{unit}\' is deprecated and will be removed in a "
+            f"future version. Please use \'{c_DEPR_ABBREVS.get(unit.upper())}\' "
+            f"instead of \'{unit}\'.",
+            FutureWarning,
+            stacklevel=find_stack_level(),
+        )
+        unit = c_DEPR_ABBREVS[unit.upper()]
     try:
         return timedelta_abbrevs[unit.lower()]
     except KeyError:

pandas/tests/frame/methods/test_astype.py

@@ -716,7 +716,7 @@ def test_astype_ignores_errors_for_extension_dtypes(self, data, dtype, errors):
 
     def test_astype_tz_conversion(self):
         # GH 35973
-        val = {"tz": date_range("2020-08-30", freq="d", periods=2, tz="Europe/London")}
+        val = {"tz": date_range("2020-08-30", freq="D", periods=2, tz="Europe/London")}
         df = DataFrame(val)
         result = df.astype({"tz": "datetime64[ns, Europe/Berlin]"})
 
@@ -727,7 +727,7 @@ def test_astype_tz_conversion(self):
     @pytest.mark.parametrize("tz", ["UTC", "Europe/Berlin"])
     def test_astype_tz_object_conversion(self, tz):
         # GH 35973
-        val = {"tz": date_range("2020-08-30", freq="d", periods=2, tz="Europe/London")}
+        val = {"tz": date_range("2020-08-30", freq="D", periods=2, tz="Europe/London")}
         expected = DataFrame(val)
 
         # convert expected to object dtype from other tz str (independently tested)

pandas/tests/frame/methods/test_at_time.py

@@ -58,7 +58,7 @@ def test_at_time_midnight(self, frame_or_series):
 
     def test_at_time_nonexistent(self, frame_or_series):
         # time doesn't exist
-        rng = date_range("1/1/2012", freq="23Min", periods=384)
+        rng = date_range("1/1/2012", freq="23min", periods=384)
         ts = DataFrame(np.random.default_rng(2).standard_normal(len(rng)), rng)
         ts = tm.get_obj(ts, frame_or_series)
         rs = ts.at_time("16:00")

pandas/tests/frame/methods/test_reindex.py

@@ -787,7 +787,9 @@ def test_reindex_axes(self):
             index=[datetime(2012, 1, 1), datetime(2012, 1, 2), datetime(2012, 1, 3)],
             columns=["a", "b", "c"],
         )
-        time_freq = date_range("2012-01-01", "2012-01-03", freq="d")
+        msg = "'d' is deprecated and will be removed in a future version."
+        with tm.assert_produces_warning(FutureWarning, match=msg):
+            time_freq = date_range("2012-01-01", "2012-01-03", freq="d")
         some_cols = ["a", "b"]
 
         index_freq = df.reindex(index=time_freq).index.freq

pandas/tests/indexes/datetimes/methods/test_repeat.py

@@ -63,7 +63,7 @@ def test_repeat(self, tz_naive_fixture, unit):
         reps = 2
         msg = "the 'axis' parameter is not supported"
 
-        rng = date_range(start="2016-01-01", periods=2, freq="30Min", tz=tz, unit=unit)
+        rng = date_range(start="2016-01-01", periods=2, freq="30min", tz=tz, unit=unit)
 
         expected_rng = DatetimeIndex(
             [

pandas/tests/indexes/datetimes/methods/test_round.py

@@ -42,7 +42,7 @@ def test_round_invalid(self, freq, error_msg):
 
     def test_round(self, tz_naive_fixture, unit):
         tz = tz_naive_fixture
-        rng = date_range(start="2016-01-01", periods=5, freq="30Min", tz=tz, unit=unit)
+        rng = date_range(start="2016-01-01", periods=5, freq="30min", tz=tz, unit=unit)
         elt = rng[1]
 
         expected_rng = DatetimeIndex(
@@ -104,7 +104,7 @@ def test_round4(self, tz_naive_fixture):
     def test_no_rounding_occurs(self, tz_naive_fixture):
         # GH 21262
         tz = tz_naive_fixture
-        rng = date_range(start="2016-01-01", periods=5, freq="2Min", tz=tz)
+        rng = date_range(start="2016-01-01", periods=5, freq="2min", tz=tz)
 
         expected_rng = DatetimeIndex(
             [

pandas/tests/indexes/period/methods/test_asfreq.py

@@ -19,43 +19,43 @@ def test_asfreq(self):
         pi3 = period_range(freq="M", start="1/1/2001", end="1/1/2001")
         pi4 = period_range(freq="D", start="1/1/2001", end="1/1/2001")
         pi5 = period_range(freq="h", start="1/1/2001", end="1/1/2001 00:00")
-        pi6 = period_range(freq="Min", start="1/1/2001", end="1/1/2001 00:00")
+        pi6 = period_range(freq="min", start="1/1/2001", end="1/1/2001 00:00")
         pi7 = period_range(freq="s", start="1/1/2001", end="1/1/2001 00:00:00")
 
         assert pi1.asfreq("Q", "s") == pi2
         assert pi1.asfreq("Q", "s") == pi2
         assert pi1.asfreq("M", "start") == pi3
         assert pi1.asfreq("D", "StarT") == pi4
         assert pi1.asfreq("h", "beGIN") == pi5
-        assert pi1.asfreq("Min", "s") == pi6
+        assert pi1.asfreq("min", "s") == pi6
         assert pi1.asfreq("s", "s") == pi7
 
         assert pi2.asfreq("Y", "s") == pi1
         assert pi2.asfreq("M", "s") == pi3
         assert pi2.asfreq("D", "s") == pi4
         assert pi2.asfreq("h", "s") == pi5
-        assert pi2.asfreq("Min", "s") == pi6
+        assert pi2.asfreq("min", "s") == pi6
         assert pi2.asfreq("s", "s") == pi7
 
         assert pi3.asfreq("Y", "s") == pi1
         assert pi3.asfreq("Q", "s") == pi2
         assert pi3.asfreq("D", "s") == pi4
         assert pi3.asfreq("h", "s") == pi5
-        assert pi3.asfreq("Min", "s") == pi6
+        assert pi3.asfreq("min", "s") == pi6
         assert pi3.asfreq("s", "s") == pi7
 
         assert pi4.asfreq("Y", "s") == pi1
         assert pi4.asfreq("Q", "s") == pi2
         assert pi4.asfreq("M", "s") == pi3
         assert pi4.asfreq("h", "s") == pi5
-        assert pi4.asfreq("Min", "s") == pi6
+        assert pi4.asfreq("min", "s") == pi6
         assert pi4.asfreq("s", "s") == pi7
 
         assert pi5.asfreq("Y", "s") == pi1
         assert pi5.asfreq("Q", "s") == pi2
         assert pi5.asfreq("M", "s") == pi3
         assert pi5.asfreq("D", "s") == pi4
-        assert pi5.asfreq("Min", "s") == pi6
+        assert pi5.asfreq("min", "s") == pi6
         assert pi5.asfreq("s", "s") == pi7
 
         assert pi6.asfreq("Y", "s") == pi1
@@ -70,7 +70,7 @@ def test_asfreq(self):
         assert pi7.asfreq("M", "s") == pi3
         assert pi7.asfreq("D", "s") == pi4
         assert pi7.asfreq("h", "s") == pi5
-        assert pi7.asfreq("Min", "s") == pi6
+        assert pi7.asfreq("min", "s") == pi6
 
         msg = "How must be one of S or E"
         with pytest.raises(ValueError, match=msg):

pandas/tests/indexes/period/test_constructors.py

@@ -531,7 +531,7 @@ def test_period_range_length(self):
         pi = period_range(freq="h", start="1/1/2001", end="12/31/2001 23:00")
         assert len(pi) == 365 * 24
 
-        pi = period_range(freq="Min", start="1/1/2001", end="1/1/2001 23:59")
+        pi = period_range(freq="min", start="1/1/2001", end="1/1/2001 23:59")
         assert len(pi) == 24 * 60
 
         pi = period_range(freq="s", start="1/1/2001", end="1/1/2001 23:59:59")

pandas/tests/indexes/period/test_period.py

@@ -87,7 +87,7 @@ def test_values(self):
             period_range(freq="M", start="1/1/2001", end="1/1/2002"),
             period_range(freq="D", start="12/1/2001", end="6/1/2001"),
             period_range(freq="h", start="12/31/2001", end="1/1/2002 23:00"),
-            period_range(freq="Min", start="12/31/2001", end="1/1/2002 00:20"),
+            period_range(freq="min", start="12/31/2001", end="1/1/2002 00:20"),
             period_range(
                 freq="s", start="12/31/2001 00:00:00", end="12/31/2001 00:05:00"
             ),

pandas/tests/indexes/period/test_period_range.py

@@ -215,7 +215,7 @@ def test_constructor_U(self):
         ],
     )
     def test_a_deprecated_from_time_series(self, freq, freq_depr):
-        # GH#52536
+        # GH#52536,GH#56346
         msg = f"'{freq_depr[1:]}' is deprecated and will be removed in a "
         f"future version. Please use '{freq[1:]}' instead."
 

pandas/tests/indexes/timedeltas/test_scalar_compat.py

@@ -49,7 +49,7 @@ def test_tdi_total_seconds_all_nat(self):
         tm.assert_series_equal(result, expected)
 
     def test_tdi_round(self):
-        td = timedelta_range(start="16801 days", periods=5, freq="30Min")
+        td = timedelta_range(start="16801 days", periods=5, freq="30min")
         elt = td[1]
 
         expected_rng = TimedeltaIndex(

pandas/tests/indexes/timedeltas/test_timedelta_range.py

@@ -88,6 +88,15 @@ def test_timedelta_range_H_T_deprecated(self, freq, msg_freq):
             expected = timedelta_range(start="0 days", end="4 days", freq=freq)
         tm.assert_index_equal(result, expected)
 
+    @pytest.mark.parametrize("freq_depr", ["H", "MIN", "Min", "S", "Us", "NS"])
+    def test_timedelta_range_uppercase_freq_deprecated(self, freq_depr):
+        # GH#56346
+        expected = to_timedelta(np.arange(5), unit=freq_depr.lower())
+        to_timedelta(np.arange(5), unit="m")
+        with tm.assert_produces_warning(FutureWarning, match=freq_depr):
+            result = to_timedelta(np.arange(5), unit=freq_depr)
+            tm.assert_index_equal(result, expected)
+
     def test_errors(self):
         # not enough params
         msg = (

pandas/tests/io/formats/test_to_csv.py

@@ -221,7 +221,9 @@ def test_to_csv_na_rep_nullable_string(self, nullable_string_dtype):
     def test_to_csv_date_format(self):
         # GH 10209
         df_sec = DataFrame({"A": pd.date_range("20130101", periods=5, freq="s")})
-        df_day = DataFrame({"A": pd.date_range("20130101", periods=5, freq="d")})
+        msg = "'d' is deprecated and will be removed in a future version."
+        with tm.assert_produces_warning(FutureWarning, match=msg):
+            df_day = DataFrame({"A": pd.date_range("20130101", periods=5, freq="d")})
 
         expected_rows = [
             ",A",