Results ¶

arc_distance (source code) ¶

Plot for arc_distance

	Rank	Function name	Cold time (s)	Warm time (s): best (stddev)	Speedup
	#1	arc_distance_pythran_nested_for_loops	0.084	0.079 (0.003)	86.9
	#2	arc_distance_parakeet_for_loops	0.378	0.082 (0.001)	84.0
	#3	arc_distance_theano_broadcast	0.090	0.086 (0.000)	79.6
	#4	arc_distance_parakeet_comprehensions	0.446	0.092 (0.002)	74.9
	#5	arc_distance_numpy_broadcast	N/A	0.093 (0.000)	73.4
	#6	arc_distance_theano_alloc	0.116	0.107 (0.001)	64.1
	#7	arc_distance_numpy_tile	N/A	0.174 (0.000)	39.3
	#8	arc_distance_python_nested_for_loops	N/A	6.857 (0.000)	1.0

There were 1 import error(s) while running this benchmark.

growcut (source code) ¶

Plot for growcut

	Rank	Function name	Cold time (s)	Warm time (s): best (stddev)	Speedup
	#1	growcut_parakeet	0.421	0.005 (0.000)	2147.5
	#2	growcut_cython	0.009	0.006 (0.001)	1615.0
	#3	growcut_pythran	0.012	0.010 (0.002)	1027.8
	#4	growcut_python	N/A	10.016 (0.000)	1.0

There were 1 import error(s) while running this benchmark.

julia (source code) ¶

Plot for julia

	Rank	Function name	Cold time (s)	Warm time (s): best (stddev)	Speedup
	#1	julia_pythran_for_loops	0.003	0.002 (0.000)	1865.0
	#2	julia_parakeet_for_loops	0.433	0.003 (0.000)	1341.1
	#3	julia_cython_for_loops	0.004	0.004 (0.000)	1194.4
	#4	julia_pyopencl	N/A	0.010 (0.000)	433.0
	#5	julia_python_numpy	N/A	0.187 (0.000)	24.1
	#6	julia_python_for_loops	N/A	4.497 (0.000)	1.0

There were 1 import error(s) while running this benchmark.

pairwise (source code) ¶

Plot for pairwise

	Rank	Function name	Cold time (s)	Warm time (s): best (stddev)	Speedup
	#1	pairwise_theano_blas_float32	0.001	0.001 (0.000)	25834.8
	#2	pairwise_theano_blas_float64	0.001	0.001 (0.000)	20523.4
	#3	pairwise_python_numpy_dot	N/A	0.002 (0.000)	9637.2
	#4	pairwise_pyopencl_cpu	N/A	0.004 (0.000)	5671.6
	#5	pairwise_cython_for_loops	0.013	0.012 (0.000)	1697.3
	#6	pairwise_parakeet_inner_numpy	0.246	0.013 (0.002)	1624.9
	#7	pairwise_parakeet_nested_for_loops	0.183	0.013 (0.000)	1610.8
	#8	pairwise_parakeet_comprehensions	0.299	0.014 (0.001)	1565.5
	#9	pairwise_pythran_nested_for_loops	0.018	0.016 (0.002)	1353.4
	#10	pairwise_theano_broadcast_float32	0.053	0.034 (0.002)	631.5
	#11	pairwise_theano_broadcast_float64	0.045	0.042 (0.001)	501.1
	#12	pairwise_python_broadcast_numpy	N/A	0.159 (0.000)	133.0
	#13	pairwise_python_inner_numpy	N/A	1.843 (0.000)	11.5
	#14	pairwise_python_nested_for_loops	N/A	21.182 (0.000)	1.0

There were 1 import error(s) while running this benchmark.

rosen_der (source code) ¶

Plot for rosen_der

	Rank	Function name	Cold time (s)	Warm time (s): best (stddev)	Speedup
	#1	rosen_der_loops_parakeet	0.261	0.004 (0.000)	2459.3
	#2	rosen_der_pythran	0.005	0.005 (0.001)	1990.5
	#3	rosen_der_cython	0.008	0.005 (0.000)	1888.2
	#4	rosen_der_theano_float32	0.012	0.006 (0.001)	1587.4
	#5	rosen_der_theano_float64	0.015	0.008 (0.001)	1285.0
	#6	rosen_der_numpy_parakeet	0.374	0.009 (0.000)	1124.9
	#7	rosen_der_numpy	N/A	0.045 (0.000)	227.3
	#8	rosen_der_python	N/A	10.181 (0.000)	1.0

There were 1 import error(s) while running this benchmark.

Error Summary ¶

arc_distance ¶

Benchmark loading errors ¶

arc_distance_numba ¶

AttributeError: 'NoneType' object has no attribute 'core'
Traceback (most recent call last):
File "run_benchmarks.py", line 98, in find_benchmarks
module = __import__(abs_module_name, fromlist="dummy")
File "/Users/ogrisel/code/python-benchmarks/arc_distance/arc_distance_numba.py", line 5, in 
from numba import autojit
File "/usr/local/lib/python2.7/site-packages/numba/__init__.py", line 18, in 
from numba import utils, typesystem
File "/usr/local/lib/python2.7/site-packages/numba/utils.py", line 102, in 
context = get_minivect_context()
File "/usr/local/lib/python2.7/site-packages/numba/utils.py", line 100, in get_minivect_context
return NumbaContext()
File "/usr/local/lib/python2.7/site-packages/numba/minivect/miniast.py", line 140, in __init__
self.llvm_module = llvm.core.Module.new('default_module')
AttributeError: 'NoneType' object has no attribute 'core'

growcut ¶

Benchmark loading errors ¶

growcut_numba ¶

AttributeError: 'NoneType' object has no attribute 'core'
Traceback (most recent call last):
File "run_benchmarks.py", line 98, in find_benchmarks
module = __import__(abs_module_name, fromlist="dummy")
File "/Users/ogrisel/code/python-benchmarks/growcut/growcut_numba.py", line 2, in 
from numba import autojit
File "/usr/local/lib/python2.7/site-packages/numba/__init__.py", line 18, in 
from numba import utils, typesystem
File "/usr/local/lib/python2.7/site-packages/numba/utils.py", line 102, in 
context = get_minivect_context()
File "/usr/local/lib/python2.7/site-packages/numba/utils.py", line 100, in get_minivect_context
return NumbaContext()
File "/usr/local/lib/python2.7/site-packages/numba/minivect/miniast.py", line 140, in __init__
self.llvm_module = llvm.core.Module.new('default_module')
AttributeError: 'NoneType' object has no attribute 'core'

julia ¶

Benchmark loading errors ¶

julia_numba ¶

AttributeError: 'NoneType' object has no attribute 'core'
Traceback (most recent call last):
File "run_benchmarks.py", line 98, in find_benchmarks
module = __import__(abs_module_name, fromlist="dummy")
File "/Users/ogrisel/code/python-benchmarks/julia/julia_numba.py", line 2, in 
from numba import autojit
File "/usr/local/lib/python2.7/site-packages/numba/__init__.py", line 18, in 
from numba import utils, typesystem
File "/usr/local/lib/python2.7/site-packages/numba/utils.py", line 102, in 
context = get_minivect_context()
File "/usr/local/lib/python2.7/site-packages/numba/utils.py", line 100, in get_minivect_context
return NumbaContext()
File "/usr/local/lib/python2.7/site-packages/numba/minivect/miniast.py", line 140, in __init__
self.llvm_module = llvm.core.Module.new('default_module')
AttributeError: 'NoneType' object has no attribute 'core'

pairwise ¶

Benchmark loading errors ¶

pairwise_numba ¶

AttributeError: 'NoneType' object has no attribute 'core'
Traceback (most recent call last):
File "run_benchmarks.py", line 98, in find_benchmarks
module = __import__(abs_module_name, fromlist="dummy")
File "/Users/ogrisel/code/python-benchmarks/pairwise/pairwise_numba.py", line 5, in 
from numba import autojit
File "/usr/local/lib/python2.7/site-packages/numba/__init__.py", line 18, in 
from numba import utils, typesystem
File "/usr/local/lib/python2.7/site-packages/numba/utils.py", line 102, in 
context = get_minivect_context()
File "/usr/local/lib/python2.7/site-packages/numba/utils.py", line 100, in get_minivect_context
return NumbaContext()
File "/usr/local/lib/python2.7/site-packages/numba/minivect/miniast.py", line 140, in __init__
self.llvm_module = llvm.core.Module.new('default_module')
AttributeError: 'NoneType' object has no attribute 'core'

rosen_der ¶

Benchmark loading errors ¶

rosen_der_numba ¶

AttributeError: 'NoneType' object has no attribute 'core'
Traceback (most recent call last):
File "run_benchmarks.py", line 98, in find_benchmarks
module = __import__(abs_module_name, fromlist="dummy")
File "/Users/ogrisel/code/python-benchmarks/rosen_der/rosen_der_numba.py", line 5, in 
from numba import autojit
File "/usr/local/lib/python2.7/site-packages/numba/__init__.py", line 18, in 
from numba import utils, typesystem
File "/usr/local/lib/python2.7/site-packages/numba/utils.py", line 102, in 
context = get_minivect_context()
File "/usr/local/lib/python2.7/site-packages/numba/utils.py", line 100, in get_minivect_context
return NumbaContext()
File "/usr/local/lib/python2.7/site-packages/numba/minivect/miniast.py", line 140, in __init__
self.llvm_module = llvm.core.Module.new('default_module')
AttributeError: 'NoneType' object has no attribute 'core'

Runtime Environment ¶

Hardware

Platform: Darwin-13.0.0-x86_64-i386-64bit
CPU type: i386
CPU count: 4

Software

Python version: 2.7.5