pandas-dev#26065 Fix Type Annotations in pandas.core.arrays

Author: gwrome

pandas/core/arrays/array_.py

@@ -1,4 +1,4 @@
-from typing import Optional, Sequence, Union
+from typing import Optional, Sequence, Union, cast
 
 import numpy as np
 
@@ -229,7 +229,7 @@ def array(data: Sequence[object],
         dtype = registry.find(dtype) or dtype
 
     if is_extension_array_dtype(dtype):
-        cls = dtype.construct_array_type()
+        cls = cast(ExtensionDtype, dtype).construct_array_type()
         return cls._from_sequence(data, dtype=dtype, copy=copy)
 
     if dtype is None:

pandas/core/arrays/datetimelike.py

@@ -1,7 +1,7 @@
 # -*- coding: utf-8 -*-
 from datetime import datetime, timedelta
 import operator
-from typing import Any, Sequence, Tuple, Type, Union
+from typing import Any, Sequence, Type, Union, cast
 import warnings
 
 import numpy as np
@@ -40,6 +40,7 @@
 
 
 class AttributesMixin(object):
+    _data = None  # type: np.ndarray
 
     @property
     def _attributes(self):
@@ -57,7 +58,8 @@ def _get_attributes_dict(self):
         return {k: getattr(self, k, None) for k in self._attributes}
 
     @property
-    def _scalar_type(self) -> Union[Type, Tuple[Type]]:
+    def _scalar_type(self) -> Union[Type[Period], Type[Timestamp],
+                                    Type[Timedelta]]:
         """The scalar associated with this datelike
 
         * PeriodArray : Period
@@ -479,12 +481,13 @@ def __setitem__(
                 raise ValueError("setting an array element with a sequence.")
 
             if (not is_slice
-                    and len(key) != len(value)
+                    and len(cast(Sequence, key)) != len(value)
                     and not com.is_bool_indexer(key)):
                 msg = ("shape mismatch: value array of length '{}' does not "
                        "match indexing result of length '{}'.")
-                raise ValueError(msg.format(len(key), len(value)))
-            if not is_slice and len(key) == 0:
+                raise ValueError(msg.format(
+                    len(cast(Sequence, key)), len(value)))
+            if not is_slice and len(cast(Sequence, key)) == 0:
                 return
 
             value = type(self)._from_sequence(value, dtype=self.dtype)

pandas/core/arrays/integer.py

@@ -1,5 +1,6 @@
 import copy
 import sys
+from typing import Type
 import warnings
 
 import numpy as np
@@ -31,9 +32,9 @@ class _IntegerDtype(ExtensionDtype):
 
     The attributes name & type are set when these subclasses are created.
     """
-    name = None
+    name = None  # type: str
     base = None
-    type = None
+    type = None  # type: Type
     na_value = np.nan
 
     def __repr__(self):

pandas/core/arrays/interval.py

@@ -943,7 +943,7 @@ def mid(self):
         else False
         """
 
-    @property
+    @property  # type: ignore # Mypy does not support decorated properties
     @Appender(_interval_shared_docs['is_non_overlapping_monotonic']
               % _shared_docs_kwargs)
     def is_non_overlapping_monotonic(self):

pandas/core/arrays/period.py

@@ -1,7 +1,7 @@
 # -*- coding: utf-8 -*-
 from datetime import timedelta
 import operator
-from typing import Any, Callable, Optional, Sequence, Union
+from typing import Any, Callable, List, Optional, Sequence, Union
 
 import numpy as np
 
@@ -24,7 +24,7 @@
 from pandas.core.dtypes.missing import isna, notna
 
 import pandas.core.algorithms as algos
-from pandas.core.arrays import ExtensionArray, datetimelike as dtl
+from pandas.core.arrays import datetimelike as dtl
 import pandas.core.common as com
 
 from pandas.tseries import frequencies
@@ -95,7 +95,7 @@ class PeriodArray(dtl.DatetimeLikeArrayMixin, dtl.DatelikeOps):
 
     Parameters
     ----------
-    values : Union[PeriodArray, Series[period], ndarary[int], PeriodIndex]
+    values : Union[PeriodArray, Series[period], ndarray[int], PeriodIndex]
         The data to store. These should be arrays that can be directly
         converted to ordinals without inference or copy (PeriodArray,
         ndarray[int64]), or a box around such an array (Series[period],
@@ -136,7 +136,7 @@ class PeriodArray(dtl.DatetimeLikeArrayMixin, dtl.DatelikeOps):
     _scalar_type = Period
 
     # Names others delegate to us
-    _other_ops = []
+    _other_ops = []  # type: List[str]
     _bool_ops = ['is_leap_year']
     _object_ops = ['start_time', 'end_time', 'freq']
     _field_ops = ['year', 'month', 'day', 'hour', 'minute', 'second',
@@ -190,7 +190,8 @@ def _from_sequence(
             copy: bool = False,
     ) -> ABCPeriodArray:
         if dtype:
-            freq = dtype.freq
+            freq = dtype.freq  # type: ignore
+            # freq is generated dynamically in PeriodDtype's __new__ method
         else:
             freq = None
 
@@ -277,7 +278,7 @@ def _check_compatible_with(self, other):
     def dtype(self):
         return self._dtype
 
-    @property
+    @property  # type: ignore # read-only property overwriting read/write
     def freq(self):
         """
         Return the frequency object for this PeriodArray.
@@ -539,17 +540,20 @@ def _sub_period(self, other):
 
         return new_data
 
+    Other_types = Union[ABCPeriodArray, ABCSeries, ABCPeriodIndex,
+                        np.ndarray]
+
     @Appender(dtl.DatetimeLikeArrayMixin._addsub_int_array.__doc__)
     def _addsub_int_array(
             self,
-            other: Union[ExtensionArray, np.ndarray, ABCIndexClass],
+            other: Other_types,
             op: Callable[[Any], Any]
     ) -> ABCPeriodArray:
         assert op in [operator.add, operator.sub]
         if op is operator.sub:
             other = -other
-        res_values = algos.checked_add_with_arr(self.asi8, other,
-                                                arr_mask=self._isnan)
+            res_values = algos.checked_add_with_arr(self.asi8, other,
+                                                    arr_mask=self._isnan)
         res_values = res_values.view('i8')
         res_values[self._isnan] = iNaT
         return type(self)(res_values, freq=self.freq)
@@ -782,7 +786,7 @@ def period_array(
     data = np.asarray(data)
 
     if freq:
-        dtype = PeriodDtype(freq)
+        dtype = PeriodDtype(freq)  # type: Optional[PeriodDtype]
     else:
         dtype = None
 

pandas/core/arrays/timedeltas.py

@@ -1,6 +1,7 @@
 # -*- coding: utf-8 -*-
 from datetime import timedelta
 import textwrap
+from typing import List
 import warnings
 
 import numpy as np
@@ -131,8 +132,8 @@ class TimedeltaArray(dtl.DatetimeLikeArrayMixin, dtl.TimelikeOps):
     _scalar_type = Timedelta
     __array_priority__ = 1000
     # define my properties & methods for delegation
-    _other_ops = []
-    _bool_ops = []
+    _other_ops = []  # type: List[str]
+    _bool_ops = []  # type: List[str]
     _object_ops = ['freq']
     _field_ops = ['days', 'seconds', 'microseconds', 'nanoseconds']
     _datetimelike_ops = _field_ops + _object_ops + _bool_ops