Revert "Revert FrozenList changes (doc build slowdown, pandas-dev#15559)"

This reverts commit d652485.

Author: jreback

doc/source/groupby.rst

@@ -126,6 +126,16 @@ We could naturally group by either the ``A`` or ``B`` columns or both:
    grouped = df.groupby('A')
    grouped = df.groupby(['A', 'B'])
 
+.. versionadded:: 0.20
+
+If we also have a MultiIndex on columns ``A`` and ``B``, we can group by all
+but the specified columns.
+
+.. ipython:: python
+
+   df2 = df.set_index(['A', 'B'])
+   grouped = df2.groupby(level=df2.index.names.difference(['B'])
+
 These will split the DataFrame on its index (rows). We could also split by the
 columns:
 

doc/source/whatsnew/v0.20.0.txt

@@ -29,6 +29,7 @@ New features
 
 - Integration with the ``feather-format``, including a new top-level ``pd.read_feather()`` and ``DataFrame.to_feather()`` method, see :ref:`here <io.feather>`.
 - ``.str.replace`` now accepts a callable, as replacement, which is passed to ``re.sub`` (:issue:`15055`)
+- ``FrozenList`` has gained the ``.difference()`` setop method (:issue:`15475`)
 
 
 
@@ -600,6 +601,7 @@ Deprecations
 - ``Series.sortlevel`` and ``DataFrame.sortlevel`` have been deprecated in favor of ``Series.sort_index`` and ``DataFrame.sort_index`` (:issue:`15099`)
 - importing ``concat`` from ``pandas.tools.merge`` has been deprecated in favor of imports from the ``pandas`` namespace. This should only affect explict imports (:issue:`15358`)
 - ``Series/DataFrame/Panel.consolidate()`` been deprecated as a public method. (:issue:`15483`)
+- ``FrozenList`` addition (new object and inplace) have been deprecated in favor of the ``.union()`` method. (:issue: `15475`)
 - The following top-level pandas functions have been deprecated and will be removed in a future version (:issue:`13790`)
   * ``pd.pnow()``, replaced by ``Period.now()``
   * ``pd.Term``, is removed, as it is not applicable to user code. Instead use in-line string expressions in the where clause when searching in HDFStore

pandas/core/panel.py

@@ -940,9 +940,9 @@ def construct_index_parts(idx, major=True):
             minor_labels, minor_levels, minor_names = construct_index_parts(
                 self.minor_axis, major=False)
 
-        levels = major_levels + minor_levels
-        labels = major_labels + minor_labels
-        names = major_names + minor_names
+        levels = list(major_levels) + list(minor_levels)
+        labels = list(major_labels) + list(minor_labels)
+        names = list(major_names) + list(minor_names)
 
         index = MultiIndex(levels=levels, labels=labels, names=names,
                            verify_integrity=False)

pandas/core/reshape.py

@@ -216,8 +216,8 @@ def get_new_columns(self):
         width = len(self.value_columns)
         propagator = np.repeat(np.arange(width), stride)
         if isinstance(self.value_columns, MultiIndex):
-            new_levels = self.value_columns.levels + (self.removed_level,)
-            new_names = self.value_columns.names + (self.removed_name,)
+            new_levels = self.value_columns.levels.union((self.removed_level,))
+            new_names = self.value_columns.names.union((self.removed_name,))
 
             new_labels = [lab.take(propagator)
                           for lab in self.value_columns.labels]
@@ -806,7 +806,7 @@ def melt(frame, id_vars=None, value_vars=None, var_name=None,
     for col in id_vars:
         mdata[col] = np.tile(frame.pop(col).values, K)
 
-    mcolumns = id_vars + var_name + [value_name]
+    mcolumns = list(id_vars) + list(var_name) + list([value_name])
 
     mdata[value_name] = frame.values.ravel('F')
     for i, col in enumerate(var_name):

pandas/core/strings.py

@@ -787,7 +787,7 @@ def str_extractall(arr, pat, flags=0):
     if 0 < len(index_list):
         from pandas import MultiIndex
         index = MultiIndex.from_tuples(
-            index_list, names=arr.index.names + ["match"])
+            index_list, names=arr.index.names.union(["match"]))
     else:
         index = None
     result = arr._constructor_expanddim(match_list, index=index,

pandas/indexes/frozen.py

@@ -13,6 +13,8 @@
 from pandas.types.cast import _coerce_indexer_dtype
 from pandas.formats.printing import pprint_thing
 
+import warnings
+
 
 class FrozenList(PandasObject, list):
 
@@ -25,11 +27,14 @@ class FrozenList(PandasObject, list):
     #           typechecks
 
     def __add__(self, other):
+        warnings.warn("__add__ is deprecated, use union(...)", FutureWarning)
+        return self.union(other)
+
+    def __iadd__(self, other):
+        warnings.warn("__iadd__ is deprecated, use union(...)", FutureWarning)
         if isinstance(other, tuple):
             other = list(other)
-        return self.__class__(super(FrozenList, self).__add__(other))
-
-    __iadd__ = __add__
+        return super(FrozenList, self).__iadd__(other)
 
     # Python 2 compat
     def __getslice__(self, i, j):
@@ -80,6 +85,19 @@ def __repr__(self):
     __setitem__ = __setslice__ = __delitem__ = __delslice__ = _disabled
     pop = append = extend = remove = sort = insert = _disabled
 
+    def union(self, other):
+        """Returns a FrozenList with other concatenated to the end of self"""
+        if isinstance(other, tuple):
+            other = list(other)
+        return self.__class__(super(FrozenList, self).__add__(other))
+
+    def difference(self, other):
+        """Returns a FrozenList with the same elements as self, but with elements
+        that are also in other removed."""
+        other = set(other)
+        temp = [x for x in self if x not in other]
+        return self.__class__(temp)
+
 
 class FrozenNDArray(PandasObject, np.ndarray):
 

pandas/tests/groupby/test_value_counts.py

@@ -28,7 +28,7 @@ def check_value_counts(df, keys, bins):
 
             gr = df.groupby(keys, sort=isort)
             right = gr['3rd'].apply(Series.value_counts, **kwargs)
-            right.index.names = right.index.names[:-1] + ['3rd']
+            right.index.names = right.index.names[:-1].union(['3rd'])
 
             # have to sort on index because of unstable sort on values
             left, right = map(rebuild_index, (left, right))  # xref GH9212

pandas/tests/indexes/test_frozen.py

@@ -15,20 +15,35 @@ def setUp(self):
         self.klass = FrozenList
 
     def test_add(self):
-        result = self.container + (1, 2, 3)
+        q = FrozenList([1])
+        with tm.assert_produces_warning(FutureWarning,
+                                        check_stacklevel=False):
+            q = q + [2, 3]
+        expected = FrozenList([1, 2, 3])
+        self.check_result(q, expected)
+
+    def test_iadd(self):
+        q = FrozenList([1])
+        with tm.assert_produces_warning(FutureWarning,
+                                        check_stacklevel=False):
+            q += [2, 3]
+        expected = FrozenList([1, 2, 3])
+        self.check_result(q, expected)
+
+    def test_union(self):
+        result = self.container.union((1, 2, 3))
         expected = FrozenList(self.lst + [1, 2, 3])
         self.check_result(result, expected)
 
-        result = (1, 2, 3) + self.container
-        expected = FrozenList([1, 2, 3] + self.lst)
+    def test_difference(self):
+        result = self.container.difference([2])
+        expected = FrozenList([1, 3, 4, 5])
         self.check_result(result, expected)
 
-    def test_inplace(self):
-        q = r = self.container
-        q += [5]
-        self.check_result(q, self.lst + [5])
-        # other shouldn't be mutated
-        self.check_result(r, self.lst)
+    def test_difference_dupe(self):
+        result = FrozenList([1, 2, 3, 2]).difference([2])
+        expected = FrozenList([1, 3])
+        self.check_result(result, expected)
 
 
 class TestFrozenNDArray(CheckImmutable, CheckStringMixin, tm.TestCase):

pandas/tools/concat.py

@@ -574,7 +574,7 @@ def _make_concat_multiindex(indexes, keys, levels=None, names=None):
                                      " not have the same number of levels")
 
             # also copies
-            names = names + _get_consensus_names(indexes)
+            names = list(names) + list(_get_consensus_names(indexes))
 
         return MultiIndex(levels=levels, labels=label_list, names=names,
                           verify_integrity=False)

test_fast.sh

@@ -5,4 +5,4 @@
 # https://github.com/pytest-dev/pytest/issues/1075
 export PYTHONHASHSEED=$(python -c 'import random; print(random.randint(1, 4294967295))')
 
-pytest pandas --skip-slow --skip-network -m "not single" -n 4
+pytest pandas --skip-slow --skip-network -m "not single" -n 4 $@