Simplify Period/Datetime Array/Index constructors

Author: jbrockmendel

pandas/core/arrays/datetimelike.py

@@ -18,6 +18,7 @@
 from pandas.tseries.offsets import Tick, DateOffset
 
 from pandas.core.dtypes.common import (
+    pandas_dtype,
     needs_i8_conversion,
     is_list_like,
     is_offsetlike,
@@ -911,3 +912,34 @@ def validate_tz_from_dtype(dtype, tz):
         except TypeError:
             pass
     return tz
+
+
+def validate_dtype_freq(dtype, freq):
+    """
+    If both a dtype and a freq are available, ensure they match.  If only
+    dtype is available, extract the implied freq.
+
+    Parameters
+    ----------
+    dtype : dtype
+    freq : DateOffset or None
+
+    Returns
+    -------
+    freq : DateOffset
+
+    Raises
+    ------
+    ValueError : non-period dtype
+    IncompatibleFrequency : mismatch between dtype and freq
+    """
+    if dtype is not None:
+        dtype = pandas_dtype(dtype)
+        if not is_period_dtype(dtype):
+            raise ValueError('dtype must be PeriodDtype')
+        if freq is None:
+            freq = dtype.freq
+        elif freq != dtype.freq:
+            raise IncompatibleFrequency('specified freq and dtype '
+                                        'are different')
+    return freq

pandas/core/arrays/datetimes.py

@@ -222,6 +222,12 @@ def __new__(cls, values, freq=None, tz=None, dtype=None):
     @classmethod
     def _generate_range(cls, start, end, periods, freq, tz=None,
                         normalize=False, ambiguous='raise', closed=None):
+
+        periods = dtl.validate_periods(periods)
+        if freq is None and any(x is None for x in [periods, start, end]):
+            raise ValueError('Must provide freq argument if no data is '
+                             'supplied')
+
         if com.count_not_none(start, end, periods, freq) != 3:
             raise ValueError('Of the four parameters: start, end, periods, '
                              'and freq, exactly three must be specified')
@@ -264,7 +270,7 @@ def _generate_range(cls, start, end, periods, freq, tz=None,
         if freq is not None:
             if cls._use_cached_range(freq, _normalized, start, end):
                 # Currently always False; never hit
-                # Should be reimplemented as apart of GH 17914
+                # Should be reimplemented as a part of GH#17914
                 index = cls._cached_range(start, end, periods=periods,
                                           freq=freq)
             else:

pandas/core/arrays/period.py

@@ -27,6 +27,7 @@
 from pandas.tseries import frequencies
 from pandas.tseries.offsets import Tick, DateOffset
 
+from pandas.core.arrays import datetimelike as dtl
 from pandas.core.arrays.datetimelike import DatetimeLikeArrayMixin
 
 
@@ -132,7 +133,7 @@ def __new__(cls, values, freq=None, **kwargs):
             # TODO: what if it has tz?
             values = dt64arr_to_periodarr(values, freq)
 
-        return cls._simple_new(values, freq, **kwargs)
+        return cls._simple_new(values, freq=freq, **kwargs)
 
     @classmethod
     def _simple_new(cls, values, freq=None, **kwargs):
@@ -141,21 +142,27 @@ def _simple_new(cls, values, freq=None, **kwargs):
         Ordinals in an ndarray are fastpath-ed to `_from_ordinals`
         """
 
+        if is_period_dtype(values):
+            freq = dtl.validate_dtype_freq(values.dtype, freq)
+            values = values.asi8
+
         if not is_integer_dtype(values):
             values = np.array(values, copy=False)
             if len(values) > 0 and is_float_dtype(values):
                 raise TypeError("{cls} can't take floats"
                                 .format(cls=cls.__name__))
-            return cls(values, freq=freq)
+            return cls(values, freq=freq, **kwargs)
 
-        return cls._from_ordinals(values, freq)
+        return cls._from_ordinals(values, freq=freq, **kwargs)
 
     @classmethod
-    def _from_ordinals(cls, values, freq=None):
+    def _from_ordinals(cls, values, freq=None, **kwargs):
         """
         Values should be int ordinals
         `__new__` & `_simple_new` cooerce to ordinals and call this method
         """
+        # **kwargs are included so that the signature matches PeriodIndex,
+        #  letting us share _simple_new
 
         values = np.array(values, dtype='int64', copy=False)
 

pandas/core/arrays/timedeltas.py

@@ -131,15 +131,6 @@ def __new__(cls, values, freq=None, start=None, end=None, periods=None,
 
         freq, freq_infer = dtl.maybe_infer_freq(freq)
 
-        if values is None:
-            # TODO: Remove this block and associated kwargs; GH#20535
-            if freq is None and com._any_none(periods, start, end):
-                raise ValueError('Must provide freq argument if no data is '
-                                 'supplied')
-            periods = dtl.validate_periods(periods)
-            return cls._generate_range(start, end, periods, freq,
-                                       closed=closed)
-
         result = cls._simple_new(values, freq=freq)
         if freq_infer:
             inferred = result.inferred_freq
@@ -151,6 +142,12 @@ def __new__(cls, values, freq=None, start=None, end=None, periods=None,
     @classmethod
     def _generate_range(cls, start, end, periods, freq, closed=None, **kwargs):
         # **kwargs are for compat with TimedeltaIndex, which includes `name`
+
+        periods = dtl.validate_periods(periods)
+        if freq is None and any(x is None for x in [periods, start, end]):
+            raise ValueError('Must provide freq argument if no data is '
+                             'supplied')
+
         if com.count_not_none(start, end, periods, freq) != 3:
             raise ValueError('Of the four parameters: start, end, periods, '
                              'and freq, exactly three must be specified')

pandas/core/indexes/datetimes.py

@@ -294,10 +294,6 @@ def __new__(cls, data=None,
 
         if data is None:
             # TODO: Remove this block and associated kwargs; GH#20535
-            if freq is None and com._any_none(periods, start, end):
-                raise ValueError('Must provide freq argument if no data is '
-                                 'supplied')
-            periods = dtl.validate_periods(periods)
             return cls._generate_range(start, end, periods, name, freq,
                                        tz=tz, normalize=normalize,
                                        closed=closed, ambiguous=ambiguous)

pandas/core/indexes/period.py

@@ -181,15 +181,7 @@ def __new__(cls, data=None, ordinal=None, freq=None, start=None, end=None,
         if name is None and hasattr(data, 'name'):
             name = data.name
 
-        if dtype is not None:
-            dtype = pandas_dtype(dtype)
-            if not is_period_dtype(dtype):
-                raise ValueError('dtype must be PeriodDtype')
-            if freq is None:
-                freq = dtype.freq
-            elif freq != dtype.freq:
-                msg = 'specified freq and dtype are different'
-                raise IncompatibleFrequency(msg)
+        freq = dtl.validate_dtype_freq(dtype, freq)
 
         # coerce freq to freq object, otherwise it can be coerced elementwise
         # which is slow
@@ -218,7 +210,7 @@ def __new__(cls, data=None, ordinal=None, freq=None, start=None, end=None,
         # not array / index
         if not isinstance(data, (np.ndarray, PeriodIndex,
                                  DatetimeIndex, Int64Index)):
-            if is_scalar(data) or isinstance(data, Period):
+            if is_scalar(data):
                 cls._scalar_data_error(data)
 
             # other iterable of some kind
@@ -248,21 +240,7 @@ def _engine(self):
         return self._engine_type(lambda: self, len(self))
 
     @classmethod
-    def _simple_new(cls, values, name=None, freq=None, **kwargs):
-        """
-        Values can be any type that can be coerced to Periods.
-        Ordinals in an ndarray are fastpath-ed to `_from_ordinals`
-        """
-        if not is_integer_dtype(values):
-            values = np.array(values, copy=False)
-            if len(values) > 0 and is_float_dtype(values):
-                raise TypeError("PeriodIndex can't take floats")
-            return cls(values, name=name, freq=freq, **kwargs)
-
-        return cls._from_ordinals(values, name, freq, **kwargs)
-
-    @classmethod
-    def _from_ordinals(cls, values, name=None, freq=None, **kwargs):
+    def _from_ordinals(cls, values, freq=None, name=None):
         """
         Values should be int ordinals
         `__new__` & `_simple_new` cooerce to ordinals and call this method

pandas/tests/indexes/period/test_construction.py

@@ -264,20 +264,20 @@ def test_constructor_mixed(self):
 
     def test_constructor_simple_new(self):
         idx = period_range('2007-01', name='p', periods=2, freq='M')
-        result = idx._simple_new(idx, 'p', freq=idx.freq)
+        result = idx._simple_new(idx, name='p', freq=idx.freq)
         tm.assert_index_equal(result, idx)
 
-        result = idx._simple_new(idx.astype('i8'), 'p', freq=idx.freq)
+        result = idx._simple_new(idx.astype('i8'), name='p', freq=idx.freq)
         tm.assert_index_equal(result, idx)
 
         result = idx._simple_new([pd.Period('2007-01', freq='M'),
                                   pd.Period('2007-02', freq='M')],
-                                 'p', freq=idx.freq)
+                                 name='p', freq=idx.freq)
         tm.assert_index_equal(result, idx)
 
         result = idx._simple_new(np.array([pd.Period('2007-01', freq='M'),
                                            pd.Period('2007-02', freq='M')]),
-                                 'p', freq=idx.freq)
+                                 name='p', freq=idx.freq)
         tm.assert_index_equal(result, idx)
 
     def test_constructor_simple_new_empty(self):