DEPR: remove ix (pandas-dev#27620)

Author: jbrockmendel

asv_bench/benchmarks/frame_methods.py

@@ -321,10 +321,9 @@ class Dropna:
 
     def setup(self, how, axis):
         self.df = DataFrame(np.random.randn(10000, 1000))
-        with warnings.catch_warnings(record=True):
-            self.df.ix[50:1000, 20:50] = np.nan
-            self.df.ix[2000:3000] = np.nan
-            self.df.ix[:, 60:70] = np.nan
+        self.df.iloc[50:1000, 20:50] = np.nan
+        self.df.iloc[2000:3000] = np.nan
+        self.df.iloc[:, 60:70] = np.nan
         self.df_mixed = self.df.copy()
         self.df_mixed["foo"] = "bar"
 
@@ -342,10 +341,9 @@ class Count:
 
     def setup(self, axis):
         self.df = DataFrame(np.random.randn(10000, 1000))
-        with warnings.catch_warnings(record=True):
-            self.df.ix[50:1000, 20:50] = np.nan
-            self.df.ix[2000:3000] = np.nan
-            self.df.ix[:, 60:70] = np.nan
+        self.df.iloc[50:1000, 20:50] = np.nan
+        self.df.iloc[2000:3000] = np.nan
+        self.df.iloc[:, 60:70] = np.nan
         self.df_mixed = self.df.copy()
         self.df_mixed["foo"] = "bar"
 

asv_bench/benchmarks/indexing.py

@@ -67,22 +67,6 @@ def time_iloc_scalar(self, index, index_structure):
     def time_iloc_slice(self, index, index_structure):
         self.data.iloc[:800000]
 
-    def time_ix_array(self, index, index_structure):
-        with warnings.catch_warnings(record=True):
-            self.data.ix[self.array]
-
-    def time_ix_list_like(self, index, index_structure):
-        with warnings.catch_warnings(record=True):
-            self.data.ix[[800000]]
-
-    def time_ix_scalar(self, index, index_structure):
-        with warnings.catch_warnings(record=True):
-            self.data.ix[800000]
-
-    def time_ix_slice(self, index, index_structure):
-        with warnings.catch_warnings(record=True):
-            self.data.ix[:800000]
-
     def time_loc_array(self, index, index_structure):
         self.data.loc[self.array]
 
@@ -148,10 +132,6 @@ def setup(self):
         self.bool_indexer = self.df[self.col_scalar] > 0
         self.bool_obj_indexer = self.bool_indexer.astype(object)
 
-    def time_ix(self):
-        with warnings.catch_warnings(record=True):
-            self.df.ix[self.idx_scalar, self.col_scalar]
-
     def time_loc(self):
         self.df.loc[self.idx_scalar, self.col_scalar]
 
@@ -228,14 +208,6 @@ def setup(self):
         self.idx = IndexSlice[20000:30000, 20:30, 35:45, 30000:40000]
         self.mdt = self.mdt.set_index(["A", "B", "C", "D"]).sort_index()
 
-    def time_series_ix(self):
-        with warnings.catch_warnings(record=True):
-            self.s.ix[999]
-
-    def time_frame_ix(self):
-        with warnings.catch_warnings(record=True):
-            self.df.ix[999]
-
     def time_index_slice(self):
         self.mdt.loc[self.idx, :]
 
@@ -310,10 +282,6 @@ def setup_cache(self):
     def time_lookup_iloc(self, s):
         s.iloc
 
-    def time_lookup_ix(self, s):
-        with warnings.catch_warnings(record=True):
-            s.ix
-
     def time_lookup_loc(self, s):
         s.loc
 

doc/source/reference/index.rst

@@ -49,7 +49,6 @@ public functions related to data types in pandas.
 
         api/pandas.DataFrame.blocks
         api/pandas.DataFrame.as_matrix
-        api/pandas.DataFrame.ix
         api/pandas.Index.asi8
         api/pandas.Index.data
         api/pandas.Index.flags
@@ -60,7 +59,6 @@ public functions related to data types in pandas.
         api/pandas.Series.asobject
         api/pandas.Series.blocks
         api/pandas.Series.from_array
-        api/pandas.Series.ix
         api/pandas.Series.imag
         api/pandas.Series.real
 

doc/source/whatsnew/v1.0.0.rst

@@ -535,6 +535,7 @@ or ``matplotlib.Axes.plot``. See :ref:`plotting.formatters` for more.
 - :meth:`DataFrame.hist` and :meth:`Series.hist` no longer allows ``figsize="default"``, specify figure size by passinig a tuple instead (:issue:`30003`)
 - Floordiv of integer-dtyped array by :class:`Timedelta` now raises ``TypeError`` (:issue:`21036`)
 - :func:`pandas.api.types.infer_dtype` argument ``skipna`` defaults to ``True`` instead of ``False`` (:issue:`24050`)
+- Removed the previously deprecated :attr:`Series.ix` and :attr:`DataFrame.ix` (:issue:`26438`)
 - Removed the previously deprecated :meth:`Index.summary` (:issue:`18217`)
 - Removed the previously deprecated "fastpath" keyword from the :class:`Index` constructor (:issue:`23110`)
 - Removed the previously deprecated :meth:`Series.get_value`, :meth:`Series.set_value`, :meth:`DataFrame.get_value`, :meth:`DataFrame.set_value` (:issue:`17739`)

pandas/core/indexing.py

@@ -1,6 +1,4 @@
-import textwrap
 from typing import Tuple
-import warnings
 
 import numpy as np
 
@@ -10,10 +8,8 @@
 from pandas.util._decorators import Appender
 
 from pandas.core.dtypes.common import (
-    ensure_platform_int,
     is_float,
     is_integer,
-    is_integer_dtype,
     is_iterator,
     is_list_like,
     is_numeric_dtype,
@@ -34,7 +30,6 @@
 def get_indexers_list():
 
     return [
-        ("ix", _IXIndexer),
         ("iloc", _iLocIndexer),
         ("loc", _LocIndexer),
         ("at", _AtIndexer),
@@ -112,9 +107,7 @@ def __call__(self, axis=None):
         new_self.axis = axis
         return new_self
 
-    def __iter__(self):
-        raise NotImplementedError("ix is not iterable")
-
+    # TODO: remove once geopandas no longer needs this
     def __getitem__(self, key):
         # Used in ix and downstream in geopandas _CoordinateIndexer
         if type(key) is tuple:
@@ -921,9 +914,6 @@ def _getitem_lowerdim(self, tup: Tuple):
         if len(tup) > self.ndim:
             raise IndexingError("Too many indexers. handle elsewhere")
 
-        # to avoid wasted computation
-        # df.ix[d1:d2, 0] -> columns first (True)
-        # df.ix[0, ['C', 'B', A']] -> rows first (False)
         for i, key in enumerate(tup):
             if is_label_like(key) or isinstance(key, tuple):
                 section = self._getitem_axis(key, axis=i)
@@ -1004,6 +994,7 @@ def _getitem_nested_tuple(self, tup: Tuple):
 
         return obj
 
+    # TODO: remove once geopandas no longer needs __getitem__
     def _getitem_axis(self, key, axis: int):
         if is_iterator(key):
             key = list(key)
@@ -1292,106 +1283,6 @@ def _get_slice_axis(self, slice_obj: slice, axis: int):
         return self._slice(indexer, axis=axis, kind="iloc")
 
 
-class _IXIndexer(_NDFrameIndexer):
-    """
-    A primarily label-location based indexer, with integer position fallback.
-
-    Warning: Starting in 0.20.0, the .ix indexer is deprecated, in
-    favor of the more strict .iloc and .loc indexers.
-
-    ``.ix[]`` supports mixed integer and label based access. It is
-    primarily label based, but will fall back to integer positional
-    access unless the corresponding axis is of integer type.
-
-    ``.ix`` is the most general indexer and will support any of the
-    inputs in ``.loc`` and ``.iloc``. ``.ix`` also supports floating
-    point label schemes. ``.ix`` is exceptionally useful when dealing
-    with mixed positional and label based hierarchical indexes.
-
-    However, when an axis is integer based, ONLY label based access
-    and not positional access is supported. Thus, in such cases, it's
-    usually better to be explicit and use ``.iloc`` or ``.loc``.
-
-    See more at :ref:`Advanced Indexing <advanced>`.
-    """
-
-    _ix_deprecation_warning = textwrap.dedent(
-        """
-        .ix is deprecated. Please use
-        .loc for label based indexing or
-        .iloc for positional indexing
-
-        See the documentation here:
-        http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#ix-indexer-is-deprecated"""  # noqa: E501
-    )
-
-    def __init__(self, name, obj):
-        warnings.warn(self._ix_deprecation_warning, FutureWarning, stacklevel=2)
-        super().__init__(name, obj)
-
-    @Appender(_NDFrameIndexer._validate_key.__doc__)
-    def _validate_key(self, key, axis: int) -> bool:
-        """
-        Returns
-        -------
-        bool
-        """
-        if isinstance(key, slice):
-            return True
-
-        elif com.is_bool_indexer(key):
-            return True
-
-        elif is_list_like_indexer(key):
-            return True
-
-        else:
-
-            self._convert_scalar_indexer(key, axis)
-
-        return True
-
-    def _convert_for_reindex(self, key, axis: int):
-        """
-        Transform a list of keys into a new array ready to be used as axis of
-        the object we return (e.g. including NaNs).
-
-        Parameters
-        ----------
-        key : list-like
-            Targeted labels.
-        axis: int
-            Where the indexing is being made.
-
-        Returns
-        -------
-        list-like of labels.
-        """
-        labels = self.obj._get_axis(axis)
-
-        if com.is_bool_indexer(key):
-            key = check_bool_indexer(labels, key)
-            return labels[key]
-
-        if isinstance(key, Index):
-            keyarr = labels._convert_index_indexer(key)
-        else:
-            # asarray can be unsafe, NumPy strings are weird
-            keyarr = com.asarray_tuplesafe(key)
-
-        if is_integer_dtype(keyarr):
-            # Cast the indexer to uint64 if possible so
-            # that the values returned from indexing are
-            # also uint64.
-            keyarr = labels._convert_arr_indexer(keyarr)
-
-            if not labels.is_integer():
-                keyarr = ensure_platform_int(keyarr)
-                return labels.take(keyarr)
-
-        return keyarr
-
-
 class _LocationIndexer(_NDFrameIndexer):
     def __getitem__(self, key):
         if type(key) is tuple: