DEPR: Deprecate range-based PeriodIndex construction

doc/source/whatsnew/v0.24.0.rst

@@ -1142,7 +1142,7 @@ Deprecations
 - Timezone converting a tz-aware ``datetime.datetime`` or :class:`Timestamp` with :class:`Timestamp` and the ``tz`` argument is now deprecated. Instead, use :meth:`Timestamp.tz_convert` (:issue:`23579`)
 - :func:`pandas.api.types.is_period` is deprecated in favor of `pandas.api.types.is_period_dtype` (:issue:`23917`)
 - :func:`pandas.api.types.is_datetimetz` is deprecated in favor of `pandas.api.types.is_datetime64tz` (:issue:`23917`)
-- Creating a :class:`TimedeltaIndex` or :class:`DatetimeIndex` by passing range arguments `start`, `end`, and `periods` is deprecated in favor of :func:`timedelta_range` and :func:`date_range` (:issue:`23919`)
+- Creating a :class:`TimedeltaIndex`, :class:`DatetimeIndex`, or :class:`PeriodIndex` by passing range arguments `start`, `end`, and `periods` is deprecated in favor of :func:`timedelta_range`, :func:`date_range`, or :func:`period_range` (:issue:`23919`)
 - Passing a string alias like ``'datetime64[ns, UTC]'`` as the ``unit`` parameter to :class:`DatetimeTZDtype` is deprecated. Use :class:`DatetimeTZDtype.construct_from_string` instead (:issue:`23990`).
 - In :meth:`Series.where` with Categorical data, providing an ``other`` that is not present in the categories is deprecated. Convert the categorical to a different dtype or add the ``other`` to the categories first (:issue:`24077`).
 - :meth:`Series.clip_lower`, :meth:`Series.clip_upper`, :meth:`DataFrame.clip_lower` and :meth:`DataFrame.clip_upper` are deprecated and will be removed in a future version. Use ``Series.clip(lower=threshold)``, ``Series.clip(upper=threshold)`` and the equivalent ``DataFrame`` methods (:issue:`24203`)
@@ -1313,6 +1313,7 @@ Datetimelike
 - Bug in :meth:`Series.combine_first` not properly aligning categoricals, so that missing values in ``self`` where not filled by valid values from ``other`` (:issue:`24147`)
 - Bug in :func:`DataFrame.combine` with datetimelike values raising a TypeError (:issue:`23079`)
 - Bug in :func:`date_range` with frequency of ``Day`` or higher where dates sufficiently far in the future could wrap around to the past instead of raising ``OutOfBoundsDatetime`` (:issue:`14187`)
+- Bug in :func:`period_range` ignoring the frequency of ``start`` and ``end`` when those are provided as :class:`Period` objects (:issue:`20535`).
 - Bug in :class:`PeriodIndex` with attribute ``freq.n`` greater than 1 where adding a :class:`DateOffset` object would return incorrect results (:issue:`23215`)
 - Bug in :class:`Series` that interpreted string indices as lists of characters when setting datetimelike values (:issue:`23451`)
 - Bug in :class:`Timestamp` constructor which would drop the frequency of an input :class:`Timestamp` (:issue:`22311`)

pandas/core/indexes/datetimes.py

@@ -81,15 +81,27 @@ class DatetimeIndex(DatetimeArray, DatetimeIndexOpsMixin, Int64Index):
     start : starting value, datetime-like, optional
         If data is None, start is used as the start point in generating regular
         timestamp data.
+
+        .. deprecated:: 0.24.0
+
     periods  : int, optional, > 0
         Number of periods to generate, if generating index. Takes precedence
         over end argument
-    end   : end time, datetime-like, optional
+
+        .. deprecated:: 0.24.0
+
+    end : end time, datetime-like, optional
         If periods is none, generated index will extend to first conforming
         time on or just past end argument
+
+        .. deprecated:: 0.24.0
+
     closed : string or None, default None
         Make the interval closed with respect to the given frequency to
         the 'left', 'right', or both sides (None)
+
+        .. deprecated:: 0.24. 0
+
     tz : pytz.timezone or dateutil.tz.tzfile
     ambiguous : 'infer', bool-ndarray, 'NaT', default 'raise'
         When clocks moved backward due to DST, ambiguous times may arise.
@@ -166,12 +178,16 @@ class DatetimeIndex(DatetimeArray, DatetimeIndexOpsMixin, Int64Index):
     To learn more about the frequency strings, please see `this link
     <http://pandas.pydata.org/pandas-docs/stable/timeseries.html#offset-aliases>`__.
 
+    Creating a DatetimeIndex based on `start`, `periods`, and `end` has
+    been deprecated in favor of :func:`date_range`.
+
     See Also
     ---------
     Index : The base pandas Index type.
     TimedeltaIndex : Index of timedelta64 data.
     PeriodIndex : Index of Period data.
-    pandas.to_datetime : Convert argument to datetime.
+    to_datetime : Convert argument to datetime.
+    date_range : Create a fixed-frequency DatetimeIndex.
     """
     _typ = 'datetimeindex'
     _join_precedence = 10
@@ -223,13 +239,13 @@ def __new__(cls, data=None,
             verify_integrity = True
 
         if data is None:
+            result = cls._generate_range(start, end, periods,
+                                         freq=freq, tz=tz, normalize=normalize,
+                                         closed=closed, ambiguous=ambiguous)
             warnings.warn("Creating a DatetimeIndex by passing range "
                           "endpoints is deprecated.  Use "
                           "`pandas.date_range` instead.",
                           FutureWarning, stacklevel=2)
-            result = cls._generate_range(start, end, periods,
-                                         freq=freq, tz=tz, normalize=normalize,
-                                         closed=closed, ambiguous=ambiguous)
             result.name = name
             return result
 

pandas/core/indexes/period.py

@@ -92,12 +92,21 @@ class PeriodIndex(DatelikeOps, DatetimeIndexOpsMixin, Int64Index,
     start : starting value, period-like, optional
         If data is None, used as the start point in generating regular
         period data.
+
+        .. deprecated:: 0.24.0
+
     periods : int, optional, > 0
         Number of periods to generate, if generating index. Takes precedence
         over end argument
+
+        .. deprecated:: 0.24.0
+
     end : end value, period-like, optional
         If periods is none, generated index will extend to first conforming
         period on or just past end argument
+
+        .. deprecated:: 0.24.0
+
     year : int, array, or Series, default None
     month : int, array, or Series, default None
     quarter : int, array, or Series, default None
@@ -138,18 +147,22 @@ class PeriodIndex(DatelikeOps, DatetimeIndexOpsMixin, Int64Index,
     strftime
     to_timestamp
 
+    Notes
+    -----
+    Creating a PeriodIndex based on `start`, `periods`, and `end` has
+    been deprecated in favor of :func:`period_range`.
+
     Examples
     --------
     >>> idx = pd.PeriodIndex(year=year_arr, quarter=q_arr)
 
-    >>> idx2 = pd.PeriodIndex(start='2000', end='2010', freq='A')
-
     See Also
     ---------
     Index : The base pandas Index type.
     Period : Represents a period of time.
     DatetimeIndex : Index with datetime64 data.
     TimedeltaIndex : Index of timedelta64 data.
+    period_range : Create a fixed-frequency PeriodIndex.
     """
     _typ = 'periodindex'
     _attributes = ['name', 'freq']
@@ -183,6 +196,14 @@ def __new__(cls, data=None, ordinal=None, freq=None, start=None, end=None,
             # range-based.
             data, freq = PeriodArray._generate_range(start, end, periods,
                                                      freq, fields)
+            # PeriodArray._generate range does validate that fields is
+            # empty when really using the range-based constructor.
+            if not fields:
+                warnings.warn("Creating a PeriodIndex by passing range "
+                              "endpoints is deprecated.  Use "
+                              "`pandas.period_range` instead.",
+                              FutureWarning, stacklevel=2)
+
             data = PeriodArray(data, freq=freq)
         else:
             freq = validate_dtype_freq(dtype, freq)
@@ -983,7 +1004,7 @@ def base(self):
 PeriodIndex._add_datetimelike_methods()
 
 
-def period_range(start=None, end=None, periods=None, freq='D', name=None):
+def period_range(start=None, end=None, periods=None, freq=None, name=None):
     """
     Return a fixed frequency PeriodIndex, with day (calendar) as the default
     frequency
@@ -996,8 +1017,11 @@ def period_range(start=None, end=None, periods=None, freq='D', name=None):
         Right bound for generating periods
     periods : integer, default None
         Number of periods to generate
-    freq : string or DateOffset, default 'D'
-        Frequency alias
+    freq : string or DateOffset, optional
+        Frequency alias. By default the freq is taken from `start` or `end`
+        if those are Period objects. Otherwise, the default is ``"D"`` for
+        daily frequency.
+
     name : string, default None
         Name of the resulting PeriodIndex
 
@@ -1035,5 +1059,7 @@ def period_range(start=None, end=None, periods=None, freq='D', name=None):
         raise ValueError('Of the three parameters: start, end, and periods, '
                          'exactly two must be specified')
 
-    return PeriodIndex(start=start, end=end, periods=periods,
-                       freq=freq, name=name)
+    data, freq = PeriodArray._generate_range(start, end, periods, freq,
+                                             fields={})
+    data = PeriodArray(data, freq=freq)
+    return PeriodIndex(data, name=name)

pandas/core/indexes/timedeltas.py

@@ -68,15 +68,27 @@ class TimedeltaIndex(TimedeltaArray, DatetimeIndexOpsMixin,
     start : starting value, timedelta-like, optional
         If data is None, start is used as the start point in generating regular
         timedelta data.
+
+        .. deprecated:: 0.24.0
+
     periods  : int, optional, > 0
         Number of periods to generate, if generating index. Takes precedence
         over end argument
-    end   : end time, timedelta-like, optional
+
+        .. deprecated:: 0.24.0
+
+    end : end time, timedelta-like, optional
         If periods is none, generated index will extend to first conforming
         time on or just past end argument
+
+        .. deprecated:: 0.24. 0
+
     closed : string or None, default None
         Make the interval closed with respect to the given frequency to
         the 'left', 'right', or both sides (None)
+
+        .. deprecated:: 0.24. 0
+
     name : object
         Name to be stored in the index
 
@@ -104,12 +116,15 @@ class TimedeltaIndex(TimedeltaArray, DatetimeIndexOpsMixin,
     Timedelta : Represents a duration between two dates or times.
     DatetimeIndex : Index of datetime64 data.
     PeriodIndex : Index of Period data.
+    timedelta_range : Create a fixed-frequency TimedeltaIndex.
 
     Notes
     -----
-
     To learn more about the frequency strings, please see `this link
     <http://pandas.pydata.org/pandas-docs/stable/timeseries.html#offset-aliases>`__.
+
+    Creating a TimedeltaIndex based on `start`, `periods`, and `end` has
+    been deprecated in favor of :func:`timedelta_range`.
     """
 
     _typ = 'timedeltaindex'

pandas/tests/indexes/period/test_arithmetic.py

@@ -4,7 +4,7 @@
 import pytest
 
 import pandas as pd
-from pandas import PeriodIndex
+from pandas import PeriodIndex, period_range
 import pandas.util.testing as tm
 
 
@@ -26,36 +26,36 @@ def test_pi_shift_ndarray(self):
         tm.assert_index_equal(result, expected)
 
     def test_shift(self):
-        pi1 = PeriodIndex(freq='A', start='1/1/2001', end='12/1/2009')
-        pi2 = PeriodIndex(freq='A', start='1/1/2002', end='12/1/2010')
+        pi1 = period_range(freq='A', start='1/1/2001', end='12/1/2009')
+        pi2 = period_range(freq='A', start='1/1/2002', end='12/1/2010')
 
         tm.assert_index_equal(pi1.shift(0), pi1)
 
         assert len(pi1) == len(pi2)
         tm.assert_index_equal(pi1.shift(1), pi2)
 
-        pi1 = PeriodIndex(freq='A', start='1/1/2001', end='12/1/2009')
-        pi2 = PeriodIndex(freq='A', start='1/1/2000', end='12/1/2008')
+        pi1 = period_range(freq='A', start='1/1/2001', end='12/1/2009')
+        pi2 = period_range(freq='A', start='1/1/2000', end='12/1/2008')
         assert len(pi1) == len(pi2)
         tm.assert_index_equal(pi1.shift(-1), pi2)
 
-        pi1 = PeriodIndex(freq='M', start='1/1/2001', end='12/1/2009')
-        pi2 = PeriodIndex(freq='M', start='2/1/2001', end='1/1/2010')
+        pi1 = period_range(freq='M', start='1/1/2001', end='12/1/2009')
+        pi2 = period_range(freq='M', start='2/1/2001', end='1/1/2010')
         assert len(pi1) == len(pi2)
         tm.assert_index_equal(pi1.shift(1), pi2)
 
-        pi1 = PeriodIndex(freq='M', start='1/1/2001', end='12/1/2009')
-        pi2 = PeriodIndex(freq='M', start='12/1/2000', end='11/1/2009')
+        pi1 = period_range(freq='M', start='1/1/2001', end='12/1/2009')
+        pi2 = period_range(freq='M', start='12/1/2000', end='11/1/2009')
         assert len(pi1) == len(pi2)
         tm.assert_index_equal(pi1.shift(-1), pi2)
 
-        pi1 = PeriodIndex(freq='D', start='1/1/2001', end='12/1/2009')
-        pi2 = PeriodIndex(freq='D', start='1/2/2001', end='12/2/2009')
+        pi1 = period_range(freq='D', start='1/1/2001', end='12/1/2009')
+        pi2 = period_range(freq='D', start='1/2/2001', end='12/2/2009')
         assert len(pi1) == len(pi2)
         tm.assert_index_equal(pi1.shift(1), pi2)
 
-        pi1 = PeriodIndex(freq='D', start='1/1/2001', end='12/1/2009')
-        pi2 = PeriodIndex(freq='D', start='12/31/2000', end='11/30/2009')
+        pi1 = period_range(freq='D', start='1/1/2001', end='12/1/2009')
+        pi2 = period_range(freq='D', start='12/31/2000', end='11/30/2009')
         assert len(pi1) == len(pi2)
         tm.assert_index_equal(pi1.shift(-1), pi2)
 
@@ -100,7 +100,7 @@ def test_shift_gh8083(self):
 
     def test_shift_periods(self):
         # GH #22458 : argument 'n' was deprecated in favor of 'periods'
-        idx = PeriodIndex(freq='A', start='1/1/2001', end='12/1/2009')
+        idx = period_range(freq='A', start='1/1/2001', end='12/1/2009')
         tm.assert_index_equal(idx.shift(periods=0), idx)
         tm.assert_index_equal(idx.shift(0), idx)
         with tm.assert_produces_warning(FutureWarning,

pandas/tests/indexes/period/test_asfreq.py

@@ -2,20 +2,20 @@
 import pytest
 
 import pandas as pd
-from pandas import DataFrame, PeriodIndex, Series
+from pandas import DataFrame, PeriodIndex, Series, period_range
 from pandas.util import testing as tm
 
 
 class TestPeriodIndex(object):
 
     def test_asfreq(self):
-        pi1 = PeriodIndex(freq='A', start='1/1/2001', end='1/1/2001')
-        pi2 = PeriodIndex(freq='Q', start='1/1/2001', end='1/1/2001')
-        pi3 = PeriodIndex(freq='M', start='1/1/2001', end='1/1/2001')
-        pi4 = PeriodIndex(freq='D', start='1/1/2001', end='1/1/2001')
-        pi5 = PeriodIndex(freq='H', start='1/1/2001', end='1/1/2001 00:00')
-        pi6 = PeriodIndex(freq='Min', start='1/1/2001', end='1/1/2001 00:00')
-        pi7 = PeriodIndex(freq='S', start='1/1/2001', end='1/1/2001 00:00:00')
+        pi1 = period_range(freq='A', start='1/1/2001', end='1/1/2001')
+        pi2 = period_range(freq='Q', start='1/1/2001', end='1/1/2001')
+        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')
+        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
@@ -70,7 +70,7 @@ def test_asfreq(self):
         pytest.raises(ValueError, pi7.asfreq, 'T', 'foo')
         result1 = pi1.asfreq('3M')
         result2 = pi1.asfreq('M')
-        expected = PeriodIndex(freq='M', start='2001-12', end='2001-12')
+        expected = period_range(freq='M', start='2001-12', end='2001-12')
         tm.assert_numpy_array_equal(result1.asi8, expected.asi8)
         assert result1.freqstr == '3M'
         tm.assert_numpy_array_equal(result2.asi8, expected.asi8)
@@ -126,7 +126,7 @@ def test_asfreq_combined_pi(self):
             assert result.freq == exp.freq
 
     def test_asfreq_ts(self):
-        index = PeriodIndex(freq='A', start='1/1/2001', end='12/31/2010')
+        index = period_range(freq='A', start='1/1/2001', end='12/31/2010')
         ts = Series(np.random.randn(len(index)), index=index)
         df = DataFrame(np.random.randn(len(index), 3), index=index)