DEPR: resample with PeriodIndex (#55968)

* DEPR: resample with PeriodIndex

* update docstring

* code block in 0.21.0 whatsnew

* Suppress warning

* deprecate convention keyword

doc/source/whatsnew/v0.21.0.rst

@@ -635,17 +635,22 @@ Previous behavior:
 
 New behavior:
 
-.. ipython:: python
+.. code-block:: ipython
 
-   pi = pd.period_range('2017-01', periods=12, freq='M')
+   In [1]: pi = pd.period_range('2017-01', periods=12, freq='M')
 
-   s = pd.Series(np.arange(12), index=pi)
+   In [2]: s = pd.Series(np.arange(12), index=pi)
 
-   resampled = s.resample('2Q').mean()
+   In [3]: resampled = s.resample('2Q').mean()
 
-   resampled
+   In [4]: resampled
+   Out[4]:
+   2017Q1    2.5
+   2017Q3    8.5
+   Freq: 2Q-DEC, dtype: float64
 
-   resampled.index
+   In [5]: resampled.index
+   Out[5]: PeriodIndex(['2017Q1', '2017Q3'], dtype='period[2Q-DEC]')
 
 Upsampling and calling ``.ohlc()`` previously returned a ``Series``, basically identical to calling ``.asfreq()``. OHLC upsampling now returns a DataFrame with columns ``open``, ``high``, ``low`` and ``close`` (:issue:`13083`). This is consistent with downsampling and ``DatetimeIndex`` behavior.
 

doc/source/whatsnew/v2.2.0.rst

@@ -441,6 +441,7 @@ Other Deprecations
 - Deprecated :meth:`.DataFrameGroupBy.fillna` and :meth:`.SeriesGroupBy.fillna`; use :meth:`.DataFrameGroupBy.ffill`, :meth:`.DataFrameGroupBy.bfill` for forward and backward filling or :meth:`.DataFrame.fillna` to fill with a single value (or the Series equivalents) (:issue:`55718`)
 - Deprecated :meth:`Index.format`, use ``index.astype(str)`` or ``index.map(formatter)`` instead (:issue:`55413`)
 - Deprecated :meth:`Series.ravel`, the underlying array is already 1D, so ravel is not necessary (:issue:`52511`)
+- Deprecated :meth:`Series.resample` and :meth:`DataFrame.resample` with a :class:`PeriodIndex` (and the 'convention' keyword), convert to :class:`DatetimeIndex` (with ``.to_timestamp()``) before resampling instead (:issue:`53481`)
 - Deprecated :meth:`Series.view`, use :meth:`Series.astype` instead to change the dtype (:issue:`20251`)
 - Deprecated ``core.internals`` members ``Block``, ``ExtensionBlock``, and ``DatetimeTZBlock``, use public APIs instead (:issue:`55139`)
 - Deprecated ``year``, ``month``, ``quarter``, ``day``, ``hour``, ``minute``, and ``second`` keywords in the :class:`PeriodIndex` constructor, use :meth:`PeriodIndex.from_fields` instead (:issue:`55960`)

pandas/core/generic.py

@@ -9333,7 +9333,7 @@ def resample(
         axis: Axis | lib.NoDefault = lib.no_default,
         closed: Literal["right", "left"] | None = None,
         label: Literal["right", "left"] | None = None,
-        convention: Literal["start", "end", "s", "e"] = "start",
+        convention: Literal["start", "end", "s", "e"] | lib.NoDefault = lib.no_default,
         kind: Literal["timestamp", "period"] | None | lib.NoDefault = lib.no_default,
         on: Level | None = None,
         level: Level | None = None,
@@ -9371,6 +9371,9 @@ def resample(
         convention : {{'start', 'end', 's', 'e'}}, default 'start'
             For `PeriodIndex` only, controls whether to use the start or
             end of `rule`.
+
+            .. deprecated:: 2.2.0
+                Convert PeriodIndex to DatetimeIndex before resampling instead.
         kind : {{'timestamp', 'period'}}, optional, default None
             Pass 'timestamp' to convert the resulting index to a
             `DateTimeIndex` or 'period' to convert it to a `PeriodIndex`.
@@ -9535,55 +9538,6 @@ def resample(
         2000-01-01 00:06:00    26
         Freq: 3min, dtype: int64
 
-        For a Series with a PeriodIndex, the keyword `convention` can be
-        used to control whether to use the start or end of `rule`.
-
-        Resample a year by quarter using 'start' `convention`. Values are
-        assigned to the first quarter of the period.
-
-        >>> s = pd.Series([1, 2], index=pd.period_range('2012-01-01',
-        ...                                             freq='Y',
-        ...                                             periods=2))
-        >>> s
-        2012    1
-        2013    2
-        Freq: Y-DEC, dtype: int64
-        >>> s.resample('Q', convention='start').asfreq()
-        2012Q1    1.0
-        2012Q2    NaN
-        2012Q3    NaN
-        2012Q4    NaN
-        2013Q1    2.0
-        2013Q2    NaN
-        2013Q3    NaN
-        2013Q4    NaN
-        Freq: Q-DEC, dtype: float64
-
-        Resample quarters by month using 'end' `convention`. Values are
-        assigned to the last month of the period.
-
-        >>> q = pd.Series([1, 2, 3, 4], index=pd.period_range('2018-01-01',
-        ...                                                   freq='Q',
-        ...                                                   periods=4))
-        >>> q
-        2018Q1    1
-        2018Q2    2
-        2018Q3    3
-        2018Q4    4
-        Freq: Q-DEC, dtype: int64
-        >>> q.resample('M', convention='end').asfreq()
-        2018-03    1.0
-        2018-04    NaN
-        2018-05    NaN
-        2018-06    2.0
-        2018-07    NaN
-        2018-08    NaN
-        2018-09    3.0
-        2018-10    NaN
-        2018-11    NaN
-        2018-12    4.0
-        Freq: M, dtype: float64
-
         For DataFrame objects, the keyword `on` can be used to specify the
         column instead of the index for resampling.
 
@@ -9748,6 +9702,18 @@ def resample(
         else:
             kind = None
 
+        if convention is not lib.no_default:
+            warnings.warn(
+                f"The 'convention' keyword in {type(self).__name__}.resample is "
+                "deprecated and will be removed in a future version. "
+                "Explicitly cast PeriodIndex to DatetimeIndex before resampling "
+                "instead.",
+                FutureWarning,
+                stacklevel=find_stack_level(),
+            )
+        else:
+            convention = "start"
+
         return get_resampler(
             cast("Series | DataFrame", self),
             freq=rule,

pandas/core/resample.py

@@ -1876,6 +1876,12 @@ class PeriodIndexResampler(DatetimeIndexResampler):
 
     @property
     def _resampler_for_grouping(self):
+        warnings.warn(
+            "Resampling a groupby with a PeriodIndex is deprecated. "
+            "Cast to DatetimeIndex before resampling instead.",
+            FutureWarning,
+            stacklevel=find_stack_level(),
+        )
         return PeriodIndexResamplerGroupby
 
     def _get_binner_for_time(self):
@@ -2225,6 +2231,21 @@ def _get_resampler(self, obj: NDFrame, kind=None) -> Resampler:
                 gpr_index=ax,
             )
         elif isinstance(ax, PeriodIndex) or kind == "period":
+            if isinstance(ax, PeriodIndex):
+                # GH#53481
+                warnings.warn(
+                    "Resampling with a PeriodIndex is deprecated. "
+                    "Cast index to DatetimeIndex before resampling instead.",
+                    FutureWarning,
+                    stacklevel=find_stack_level(),
+                )
+            else:
+                warnings.warn(
+                    "Resampling with kind='period' is deprecated.  "
+                    "Use datetime paths instead.",
+                    FutureWarning,
+                    stacklevel=find_stack_level(),
+                )
             return PeriodIndexResampler(
                 obj,
                 timegrouper=self,

pandas/plotting/_matplotlib/timeseries.py

@@ -86,9 +86,12 @@ def maybe_resample(series: Series, ax: Axes, kwargs: dict[str, Any]):
             )
             freq = ax_freq
         elif _is_sup(freq, ax_freq):  # one is weekly
-            how = "last"
-            series = getattr(series.resample("D"), how)().dropna()
-            series = getattr(series.resample(ax_freq), how)().dropna()
+            # Resampling with PeriodDtype is deprecated, so we convert to
+            #  DatetimeIndex, resample, then convert back.
+            ser_ts = series.to_timestamp()
+            ser_d = ser_ts.resample("D").last().dropna()
+            ser_freq = ser_d.resample(ax_freq).last().dropna()
+            series = ser_freq.to_period(ax_freq)
             freq = ax_freq
         elif is_subperiod(freq, ax_freq) or _is_sub(freq, ax_freq):
             _upsample_others(ax, freq, kwargs)

pandas/tests/resample/test_base.py

@@ -85,8 +85,13 @@ def test_asfreq_fill_value(series, create_index):
 def test_resample_interpolate(frame):
     # GH#12925
     df = frame
-    result = df.resample("1min").asfreq().interpolate()
-    expected = df.resample("1min").interpolate()
+    warn = None
+    if isinstance(df.index, PeriodIndex):
+        warn = FutureWarning
+    msg = "Resampling with a PeriodIndex is deprecated"
+    with tm.assert_produces_warning(warn, match=msg):
+        result = df.resample("1min").asfreq().interpolate()
+        expected = df.resample("1min").interpolate()
     tm.assert_frame_equal(result, expected)
 
 
@@ -118,7 +123,13 @@ def test_resample_empty_series(freq, empty_series_dti, resample_method):
     elif freq == "ME" and isinstance(ser.index, PeriodIndex):
         # index is PeriodIndex, so convert to corresponding Period freq
         freq = "M"
-    rs = ser.resample(freq)
+
+    warn = None
+    if isinstance(ser.index, PeriodIndex):
+        warn = FutureWarning
+    msg = "Resampling with a PeriodIndex is deprecated"
+    with tm.assert_produces_warning(warn, match=msg):
+        rs = ser.resample(freq)
     result = getattr(rs, resample_method)()
 
     if resample_method == "ohlc":
@@ -150,7 +161,10 @@ def test_resample_nat_index_series(freq, series, resample_method):
 
     ser = series.copy()
     ser.index = PeriodIndex([NaT] * len(ser), freq=freq)
-    rs = ser.resample(freq)
+
+    msg = "Resampling with a PeriodIndex is deprecated"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        rs = ser.resample(freq)
     result = getattr(rs, resample_method)()
 
     if resample_method == "ohlc":
@@ -182,7 +196,13 @@ def test_resample_count_empty_series(freq, empty_series_dti, resample_method):
     elif freq == "ME" and isinstance(ser.index, PeriodIndex):
         # index is PeriodIndex, so convert to corresponding Period freq
         freq = "M"
-    rs = ser.resample(freq)
+
+    warn = None
+    if isinstance(ser.index, PeriodIndex):
+        warn = FutureWarning
+    msg = "Resampling with a PeriodIndex is deprecated"
+    with tm.assert_produces_warning(warn, match=msg):
+        rs = ser.resample(freq)
 
     result = getattr(rs, resample_method)()
 
@@ -210,7 +230,13 @@ def test_resample_empty_dataframe(empty_frame_dti, freq, resample_method):
     elif freq == "ME" and isinstance(df.index, PeriodIndex):
         # index is PeriodIndex, so convert to corresponding Period freq
         freq = "M"
-    rs = df.resample(freq, group_keys=False)
+
+    warn = None
+    if isinstance(df.index, PeriodIndex):
+        warn = FutureWarning
+    msg = "Resampling with a PeriodIndex is deprecated"
+    with tm.assert_produces_warning(warn, match=msg):
+        rs = df.resample(freq, group_keys=False)
     result = getattr(rs, resample_method)()
     if resample_method == "ohlc":
         # TODO: no tests with len(df.columns) > 0
@@ -253,7 +279,14 @@ def test_resample_count_empty_dataframe(freq, empty_frame_dti):
     elif freq == "ME" and isinstance(empty_frame_dti.index, PeriodIndex):
         # index is PeriodIndex, so convert to corresponding Period freq
         freq = "M"
-    result = empty_frame_dti.resample(freq).count()
+
+    warn = None
+    if isinstance(empty_frame_dti.index, PeriodIndex):
+        warn = FutureWarning
+    msg = "Resampling with a PeriodIndex is deprecated"
+    with tm.assert_produces_warning(warn, match=msg):
+        rs = empty_frame_dti.resample(freq)
+    result = rs.count()
 
     index = _asfreq_compat(empty_frame_dti.index, freq)
 
@@ -280,7 +313,14 @@ def test_resample_size_empty_dataframe(freq, empty_frame_dti):
     elif freq == "ME" and isinstance(empty_frame_dti.index, PeriodIndex):
         # index is PeriodIndex, so convert to corresponding Period freq
         freq = "M"
-    result = empty_frame_dti.resample(freq).size()
+
+    msg = "Resampling with a PeriodIndex"
+    warn = None
+    if isinstance(empty_frame_dti.index, PeriodIndex):
+        warn = FutureWarning
+    with tm.assert_produces_warning(warn, match=msg):
+        rs = empty_frame_dti.resample(freq)
+    result = rs.size()
 
     index = _asfreq_compat(empty_frame_dti.index, freq)
 
@@ -298,12 +338,21 @@ def test_resample_size_empty_dataframe(freq, empty_frame_dti):
     ],
 )
 @pytest.mark.parametrize("dtype", [float, int, object, "datetime64[ns]"])
+@pytest.mark.filterwarnings(r"ignore:PeriodDtype\[B\] is deprecated:FutureWarning")
 def test_resample_empty_dtypes(index, dtype, resample_method):
     # Empty series were sometimes causing a segfault (for the functions
     # with Cython bounds-checking disabled) or an IndexError.  We just run
     # them to ensure they no longer do.  (GH #10228)
+    warn = None
+    if isinstance(index, PeriodIndex):
+        # GH#53511
+        index = PeriodIndex([], freq="B", name=index.name)
+        warn = FutureWarning
+    msg = "Resampling with a PeriodIndex is deprecated"
+
     empty_series_dti = Series([], index, dtype)
-    rs = empty_series_dti.resample("d", group_keys=False)
+    with tm.assert_produces_warning(warn, match=msg):
+        rs = empty_series_dti.resample("d", group_keys=False)
     try:
         getattr(rs, resample_method)()
     except DataError:
@@ -329,8 +378,18 @@ def test_apply_to_empty_series(empty_series_dti, freq):
     elif freq == "ME" and isinstance(empty_series_dti.index, PeriodIndex):
         # index is PeriodIndex, so convert to corresponding Period freq
         freq = "M"
-    result = ser.resample(freq, group_keys=False).apply(lambda x: 1)
-    expected = ser.resample(freq).apply("sum")
+
+    msg = "Resampling with a PeriodIndex"
+    warn = None
+    if isinstance(empty_series_dti.index, PeriodIndex):
+        warn = FutureWarning
+
+    with tm.assert_produces_warning(warn, match=msg):
+        rs = ser.resample(freq, group_keys=False)
+
+    result = rs.apply(lambda x: 1)
+    with tm.assert_produces_warning(warn, match=msg):
+        expected = ser.resample(freq).apply("sum")
 
     tm.assert_series_equal(result, expected, check_dtype=False)
 
@@ -340,8 +399,16 @@ def test_resampler_is_iterable(series):
     # GH 15314
     freq = "h"
     tg = Grouper(freq=freq, convention="start")
-    grouped = series.groupby(tg)
-    resampled = series.resample(freq)
+    msg = "Resampling with a PeriodIndex"
+    warn = None
+    if isinstance(series.index, PeriodIndex):
+        warn = FutureWarning
+
+    with tm.assert_produces_warning(warn, match=msg):
+        grouped = series.groupby(tg)
+
+    with tm.assert_produces_warning(warn, match=msg):
+        resampled = series.resample(freq)
     for (rk, rv), (gk, gv) in zip(resampled, grouped):
         assert rk == gk
         tm.assert_series_equal(rv, gv)
@@ -353,6 +420,12 @@ def test_resample_quantile(series):
     ser = series
     q = 0.75
     freq = "h"
-    result = ser.resample(freq).quantile(q)
-    expected = ser.resample(freq).agg(lambda x: x.quantile(q)).rename(ser.name)
+
+    msg = "Resampling with a PeriodIndex"
+    warn = None
+    if isinstance(series.index, PeriodIndex):
+        warn = FutureWarning
+    with tm.assert_produces_warning(warn, match=msg):
+        result = ser.resample(freq).quantile(q)
+        expected = ser.resample(freq).agg(lambda x: x.quantile(q)).rename(ser.name)
     tm.assert_series_equal(result, expected)

pandas/tests/resample/test_datetime_index.py

@@ -184,6 +184,9 @@ def test_resample_basic_grouper(series, unit):
     tm.assert_series_equal(result, expected)
 
 
+@pytest.mark.filterwarnings(
+    "ignore:The 'convention' keyword in Series.resample:FutureWarning"
+)
 @pytest.mark.parametrize(
     "_index_start,_index_end,_index_name",
     [("1/1/2000 00:00:00", "1/1/2000 00:13:00", "index")],
@@ -1055,7 +1058,10 @@ def test_period_with_agg():
     )
 
     expected = s2.to_timestamp().resample("D").mean().to_period()
-    result = s2.resample("D").agg(lambda x: x.mean())
+    msg = "Resampling with a PeriodIndex is deprecated"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        rs = s2.resample("D")
+    result = rs.agg(lambda x: x.mean())
     tm.assert_series_equal(result, expected)