Merge pull request #5554 from jreback/apply_bug

BUG: Bug fix in apply when using custom function and objects are not mutated (GH5545)

Author: jreback

doc/source/release.rst

@@ -812,6 +812,7 @@ Bug Fixes
  length to the indexer (:issue:`5508`)
  - Bug in getitem with a multi-index and ``iloc`` (:issue:`5528`)
  - Bug in delitem on a Series (:issue:`5542`)
+ - Bug fix in apply when using custom function and objects are not mutated (:issue:`5545`)
 
 pandas 0.12.0
 -------------

pandas/core/groupby.py

@@ -543,13 +543,13 @@ def head(self, n=5):
 
  >>> df = DataFrame([[1, 2], [1, 4], [5, 6]],
  columns=['A', 'B'])
- >>> df.groupby('A', as_index=False).head(1) 
+ >>> df.groupby('A', as_index=False).head(1)
  A B
  0 1 2
  2 5 6
  >>> df.groupby('A').head(1)
  A B
- A 
+ A
  1 0 1 2
  5 2 5 6
 
@@ -572,16 +572,16 @@ def tail(self, n=5):
 
  >>> df = DataFrame([[1, 2], [1, 4], [5, 6]],
  columns=['A', 'B'])
- >>> df.groupby('A', as_index=False).tail(1) 
+ >>> df.groupby('A', as_index=False).tail(1)
  A B
  0 1 2
  2 5 6
  >>> df.groupby('A').head(1)
  A B
- A 
+ A
  1 0 1 2
  5 2 5 6
- 
+
  """
  rng = np.arange(0, -self.grouper._max_groupsize, -1, dtype='int64')
  in_tail = self._cumcount_array(rng, ascending=False) > -n
@@ -2149,6 +2149,12 @@ def _wrap_applied_output(self, keys, values, not_indexed_same=False):
  keys, values, not_indexed_same=not_indexed_same
  )
 
+ # still a series
+ # path added as of GH 5545
+ elif all_indexed_same:
+ from pandas.tools.merge import concat
+ return concat(values)
+
  if not all_indexed_same:
  return self._concat_objects(
  keys, values, not_indexed_same=not_indexed_same

pandas/src/reduce.pyx

@@ -541,7 +541,7 @@ def apply_frame_axis0(object frame, object f, object names,
  # I'm paying the price for index-sharing, ugh
  try:
  if piece.index is slider.dummy.index:
- piece.index = piece.index.copy()
+ piece = piece.copy()
  else:
  mutated = True
  except AttributeError:

pandas/tests/test_groupby.py

@@ -1214,7 +1214,7 @@ def test_groupby_as_index_apply(self):
  res_not_as_apply = g_not_as.apply(lambda x: x.head(2)).index
 
  # apply doesn't maintain the original ordering
- exp_not_as_apply = Index([0, 2, 1, 4]) 
+ exp_not_as_apply = Index([0, 2, 1, 4])
  exp_as_apply = MultiIndex.from_tuples([(1, 0), (1, 2), (2, 1), (3, 4)])
 
  assert_index_equal(res_as_apply, exp_as_apply)
@@ -1845,6 +1845,28 @@ def test_apply_corner(self):
  expected = self.tsframe * 2
  assert_frame_equal(result, expected)
 
+ def test_apply_without_copy(self):
+ # GH 5545
+ # returning a non-copy in an applied function fails
+
+ data = DataFrame({'id_field' : [100, 100, 200, 300], 'category' : ['a','b','c','c'], 'value' : [1,2,3,4]})
+
+ def filt1(x):
+ if x.shape[0] == 1:
+ return x.copy()
+ else:
+ return x[x.category == 'c']
+
+ def filt2(x):
+ if x.shape[0] == 1:
+ return x
+ else:
+ return x[x.category == 'c']
+
+ expected = data.groupby('id_field').apply(filt1)
+ result = data.groupby('id_field').apply(filt2)
+ assert_frame_equal(result,expected)
+
  def test_apply_use_categorical_name(self):
  from pandas import qcut
  cats = qcut(self.df.C, 4)
@@ -2638,7 +2660,7 @@ def test_cumcount_mi(self):
  expected = Series([0, 1, 2, 0, 3], index=mi)
 
  assert_series_equal(expected, g.cumcount())
- assert_series_equal(expected, sg.cumcount()) 
+ assert_series_equal(expected, sg.cumcount())
 
  def test_cumcount_groupby_not_col(self):
  df = DataFrame([['a'], ['a'], ['a'], ['b'], ['a']], columns=['A'], index=[0] * 5)
@@ -2895,7 +2917,7 @@ def test_filter_maintains_ordering(self):
  def test_filter_and_transform_with_non_unique_int_index(self):
  # GH4620
  index = [1, 1, 1, 2, 1, 1, 0, 1]
- df = DataFrame({'pid' : [1,1,1,2,2,3,3,3], 
+ df = DataFrame({'pid' : [1,1,1,2,2,3,3,3],
  'tag' : [23,45,62,24,45,34,25,62]}, index=index)
  grouped_df = df.groupby('tag')
  ser = df['pid']
@@ -2923,7 +2945,7 @@ def test_filter_and_transform_with_non_unique_int_index(self):
  # ^ made manually because this can get confusing!
  assert_series_equal(actual, expected)
 
- # Transform Series 
+ # Transform Series
  actual = grouped_ser.transform(len)
  expected = Series([1, 2, 2, 1, 2, 1, 1, 2], index)
  assert_series_equal(actual, expected)