Merge remote-tracking branch 'upstream/master' into df_set_index_warn

Author: h-vetinari

azure-pipelines.yml

@@ -18,8 +18,8 @@ jobs:
 - template: ci/azure/windows.yml
  parameters:
  name: Windows
- vmImage: vs2017-win2017
+ vmImage: vs2017-win2016
 - template: ci/azure/windows-py27.yml
  parameters:
  name: WindowsPy27
- vmImage: vs2017-win2017
+ vmImage: vs2017-win2016

doc/source/whatsnew/v0.24.0.txt

@@ -532,6 +532,35 @@ Current Behavior:
  ...
  OverflowError: Trying to coerce negative values to unsigned integers
 
+.. _whatsnew_0240.api.crosstab_dtypes
+
+Crosstab Preserves Dtypes
+^^^^^^^^^^^^^^^^^^^^^^^^^
+
+:func:`crosstab` will preserve now dtypes in some cases that previously would
+cast from integer dtype to floating dtype (:issue:`22019`)
+
+Previous Behavior:
+
+.. code-block:: ipython
+
+ In [3]: df = pd.DataFrame({'a': [1, 2, 2, 2, 2], 'b': [3, 3, 4, 4, 4],
+ ...: 'c': [1, 1, np.nan, 1, 1]})
+ In [4]: pd.crosstab(df.a, df.b, normalize='columns')
+ Out[4]:
+ b 3 4
+ a
+ 1 0.5 0.0
+ 2 0.5 1.0
+
+Current Behavior:
+
+.. code-block:: ipython
+
+ In [3]: df = pd.DataFrame({'a': [1, 2, 2, 2, 2], 'b': [3, 3, 4, 4, 4],
+ ...: 'c': [1, 1, np.nan, 1, 1]})
+ In [4]: pd.crosstab(df.a, df.b, normalize='columns')
+
 Datetimelike API Changes
 ^^^^^^^^^^^^^^^^^^^^^^^^
 
@@ -667,7 +696,7 @@ Timedelta
 - Bug in :class:`Index` with numeric dtype when multiplying or dividing an array with dtype ``timedelta64`` (:issue:`22390`)
 - Bug in :class:`TimedeltaIndex` incorrectly allowing indexing with ``Timestamp`` object (:issue:`20464`)
 - Fixed bug where subtracting :class:`Timedelta` from an object-dtyped array would raise ``TypeError`` (:issue:`21980`)
--
+- Fixed bug in adding a :class:`DataFrame` with all-`timedelta64[ns]` dtypes to a :class:`DataFrame` with all-integer dtypes returning incorrect results instead of raising ``TypeError`` (:issue:`22696`)
 -
 
 Timezones
@@ -835,4 +864,3 @@ Other
 - :meth:`DataFrame.nlargest` and :meth:`DataFrame.nsmallest` now returns the correct n values when keep != 'all' also when tied on the first columns (:issue:`22752`)
 - :meth:`~pandas.io.formats.style.Styler.bar` now also supports tablewise application (in addition to rowwise and columnwise) with ``axis=None`` and setting clipping range with ``vmin`` and ``vmax`` (:issue:`21548` and :issue:`21526`). ``NaN`` values are also handled properly.
 - Logical operations ``&, |, ^`` between :class:`Series` and :class:`Index` will no longer raise ``ValueError`` (:issue:`22092`)
--

pandas/core/arrays/interval.py

@@ -108,12 +108,7 @@ class IntervalArray(IntervalMixin, ExtensionArray):
  _na_value = _fill_value = np.nan
 
  def __new__(cls, data, closed=None, dtype=None, copy=False,
- fastpath=False, verify_integrity=True):
-
- if fastpath:
- return cls._simple_new(data.left, data.right, closed,
- copy=copy, dtype=dtype,
- verify_integrity=False)
+ verify_integrity=True):
 
  if isinstance(data, ABCSeries) and is_interval_dtype(data):
  data = data.values

pandas/core/arrays/period.py

@@ -264,7 +264,7 @@ def asfreq(self, freq=None, how='E'):
  if self.hasnans:
  new_data[self._isnan] = iNaT
 
- return self._simple_new(new_data, self.name, freq=freq)
+ return self._shallow_copy(new_data, freq=freq)
 
  # ------------------------------------------------------------------
  # Arithmetic Methods

pandas/core/computation/pytables.py

@@ -411,7 +411,7 @@ def visit_Subscript(self, node, **kwargs):
  slobj = self.visit(node.slice)
  try:
  value = value.value
- except:
+ except AttributeError:
  pass
 
  try:

pandas/core/dtypes/common.py

@@ -467,7 +467,7 @@ def is_timedelta64_dtype(arr_or_dtype):
  return False
  try:
  tipo = _get_dtype_type(arr_or_dtype)
- except:
+ except (TypeError, ValueError, SyntaxError):
  return False
  return issubclass(tipo, np.timedelta64)
 

pandas/core/dtypes/dtypes.py

@@ -358,11 +358,11 @@ def construct_from_string(cls, string):
  try:
  if string == 'category':
  return cls()
- except:
+ else:
+ raise TypeError("cannot construct a CategoricalDtype")
+ except AttributeError:
  pass
 
- raise TypeError("cannot construct a CategoricalDtype")
-
  @staticmethod
  def validate_ordered(ordered):
  """
@@ -519,7 +519,7 @@ def __new__(cls, unit=None, tz=None):
  if m is not None:
  unit = m.groupdict()['unit']
  tz = m.groupdict()['tz']
- except:
+ except TypeError:
  raise ValueError("could not construct DatetimeTZDtype")
 
  elif isinstance(unit, compat.string_types):

pandas/core/frame.py

@@ -3262,7 +3262,7 @@ def _ensure_valid_index(self, value):
  if not len(self.index) and is_list_like(value):
  try:
  value = Series(value)
- except:
+ except (ValueError, NotImplementedError, TypeError):
  raise ValueError('Cannot set a frame with no defined index '
  'and a value that cannot be converted to a '
  'Series')
@@ -3631,7 +3631,8 @@ def align(self, other, join='outer', axis=None, level=None, copy=True,
  fill_axis=fill_axis,
  broadcast_axis=broadcast_axis)
 
- @Appender(_shared_docs['reindex'] % _shared_doc_kwargs)
+ @Substitution(**_shared_doc_kwargs)
+ @Appender(NDFrame.reindex.__doc__)
  @rewrite_axis_style_signature('labels', [('method', None),
  ('copy', True),
  ('level', None),
@@ -4500,7 +4501,8 @@ def f(vals):
  # ----------------------------------------------------------------------
  # Sorting
 
- @Appender(_shared_docs['sort_values'] % _shared_doc_kwargs)
+ @Substitution(**_shared_doc_kwargs)
+ @Appender(NDFrame.sort_values.__doc__)
  def sort_values(self, by, axis=0, ascending=True, inplace=False,
  kind='quicksort', na_position='last'):
  inplace = validate_bool_kwarg(inplace, 'inplace')
@@ -4542,7 +4544,8 @@ def sort_values(self, by, axis=0, ascending=True, inplace=False,
  else:
  return self._constructor(new_data).__finalize__(self)
 
- @Appender(_shared_docs['sort_index'] % _shared_doc_kwargs)
+ @Substitution(**_shared_doc_kwargs)
+ @Appender(NDFrame.sort_index.__doc__)
  def sort_index(self, axis=0, level=None, ascending=True, inplace=False,
  kind='quicksort', na_position='last', sort_remaining=True,
  by=None):
@@ -4907,7 +4910,7 @@ def _arith_op(left, right):
  left, right = ops.fill_binop(left, right, fill_value)
  return func(left, right)
 
- if this._is_mixed_type or other._is_mixed_type:
+ if ops.should_series_dispatch(this, other, func):
  # iterate over columns
  return ops.dispatch_to_series(this, other, _arith_op)
  else:
@@ -4917,7 +4920,6 @@ def _arith_op(left, right):
  copy=False)
 
  def _combine_match_index(self, other, func, level=None):
- assert isinstance(other, Series)
  left, right = self.align(other, join='outer', axis=0, level=level,
  copy=False)
  assert left.index.equals(right.index)
@@ -4937,11 +4939,7 @@ def _combine_match_columns(self, other, func, level=None, try_cast=True):
  left, right = self.align(other, join='outer', axis=1, level=level,
  copy=False)
  assert left.columns.equals(right.index)
-
- new_data = left._data.eval(func=func, other=right,
- axes=[left.columns, self.index],
- try_cast=try_cast)
- return self._constructor(new_data)
+ return ops.dispatch_to_series(left, right, func, axis="columns")
 
  def _combine_const(self, other, func, errors='raise', try_cast=True):
  if lib.is_scalar(other) or np.ndim(other) == 0:
@@ -7768,7 +7766,7 @@ def convert(v):
  values = np.array([convert(v) for v in values])
  else:
  values = convert(values)
- except:
+ except (ValueError, TypeError):
  values = convert(values)
 
  else: