Results

arc_distance (source code)

Plot for arc_distance 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 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 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 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 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