Skip to content

BUG/PERF: perf issues in object groupby aggregations (GH7555) #7568

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
jreback merged 1 commit into from
Jun 25, 2014
Merged

BUG/PERF: perf issues in object groupby aggregations (GH7555) #7568

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 2 additions & 1 deletion pandas/core/groupby.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1414,10 +1414,11 @@ def aggregate(self, values, how, axis=0):
else:
is_numeric = issubclass(values.dtype.type, (np.datetime64,
np.timedelta64))
out_dtype = 'float64'
if is_numeric:
out_dtype = 'float64'
values = values.view('int64')
else:
out_dtype = 'object'
values = values.astype(object)

# will be filled in Cython function
Expand Down
2 changes: 1 addition & 1 deletion pandas/src/generate_code.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2234,7 +2234,7 @@ def generate_put_template(template, use_ints=True, use_floats=True,
date_like_list = [
('int64', 'int64_t', 'float64_t', 'np.float64'),
]
object_list = [('object', 'object', 'float64_t', 'np.float64')]
object_list = [('object', 'object', 'object', 'np.object_')]
function_list = []
if use_floats:
function_list.extend(floats_list)
Expand Down
4 changes: 2 additions & 2 deletions pandas/src/generated.pyx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6697,7 +6697,7 @@ def group_count_float32(ndarray[float32_t, ndim=2] out,

@cython.boundscheck(False)
@cython.wraparound(False)
def group_count_object(ndarray[float64_t, ndim=2] out,
def group_count_object(ndarray[object, ndim=2] out,
ndarray[int64_t] counts,
ndarray[object, ndim=2] values,
ndarray[int64_t] labels):
Expand Down Expand Up @@ -6838,7 +6838,7 @@ def group_count_bin_float32(ndarray[float32_t, ndim=2] out,

@cython.boundscheck(False)
@cython.wraparound(False)
def group_count_bin_object(ndarray[float64_t, ndim=2] out,
def group_count_bin_object(ndarray[object, ndim=2] out,
ndarray[int64_t] counts,
ndarray[object, ndim=2] values,
ndarray[int64_t] bins):
Expand Down
28 changes: 25 additions & 3 deletions vb_suite/groupby.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -232,24 +232,46 @@ def f():
labels = labels.take(np.random.permutation(len(labels)))
"""

groupby_first = Benchmark('data.groupby(labels).first()', setup,
groupby_first_float64 = Benchmark('data.groupby(labels).first()', setup,
start_date=datetime(2012, 5, 1))

groupby_first_float32 = Benchmark('data2.groupby(labels).first()', setup,
start_date=datetime(2013, 1, 1))

groupby_last = Benchmark('data.groupby(labels).last()', setup,
groupby_last_float64 = Benchmark('data.groupby(labels).last()', setup,
start_date=datetime(2012, 5, 1))

groupby_last_float32 = Benchmark('data2.groupby(labels).last()', setup,
start_date=datetime(2013, 1, 1))

groupby_nth_float64 = Benchmark('data.groupby(labels).nth(0)', setup,
start_date=datetime(2012, 5, 1))

groupby_nth_float32 = Benchmark('data2.groupby(labels).nth(0)', setup,
start_date=datetime(2013, 1, 1))

# with datetimes (GH7555)
setup = common_setup + """
df = DataFrame({'a' : date_range('1/1/2011',periods=100000,freq='s'),'b' : range(100000)})
"""

groupby_mixed_first = Benchmark('df.groupby("b").first()', setup,
groupby_first_datetimes = Benchmark('df.groupby("b").first()', setup,
start_date=datetime(2013, 5, 1))
groupby_last_datetimes = Benchmark('df.groupby("b").last()', setup,
start_date=datetime(2013, 5, 1))
groupby_nth_datetimes = Benchmark('df.groupby("b").nth(0)', setup,
start_date=datetime(2013, 5, 1))

# with object
setup = common_setup + """
df = DataFrame({'a' : ['foo']*100000,'b' : range(100000)})
"""

groupby_first_object = Benchmark('df.groupby("b").first()', setup,
start_date=datetime(2013, 5, 1))
groupby_last_object = Benchmark('df.groupby("b").last()', setup,
start_date=datetime(2013, 5, 1))
groupby_nth_object = Benchmark('df.groupby("b").nth(0)', setup,
start_date=datetime(2013, 5, 1))

#----------------------------------------------------------------------
Expand Down