mpi_array.benchmarks.bench_ufunc.MpiArrayUfuncBench_add

class mpi_array.benchmarks.bench_ufunc.MpiArrayUfuncBench_add

Bases: mpi_array.benchmarks.bench_ufunc.MpiArrayUfuncBench

Benchmark for numpy.add with mpi_array.globale.gndarray array inputs.

Methods

__init__()
free(a)	See free_mpi_array_obj().
free_mpi_array_obj(a)	Free MPI communicators and MPI windows for the given object.
get_globale_shape(locale_shape)	Returns a globale array shape for the given shape of the locale array.
initialise(shape, dtype, module_name)	Sets the module, dtype and ufunc attributes and initialises the a_ary, b_ary, b_scalar and c_ary arrays.
initialise_arrays(shape)	Initialise arrays/scalars passed to ufunc instances.
setup(shape[, dtype])	Import mpi_array module and assign to self.module.
teardown(shape)	Free arrays allocated during setup().
time_array_array_op(shape[, dtype])	Timing for two input ufuncs (two array inputs).
time_array_scalar_op(shape[, dtype])	Timing for two input ufuncs (one array input, one scalar input).
try_import_for_setup(module_name)	Attempt to import module named module_name, return the module if it exists and is importable.

Attributes

a_ary	First input array.
a_ary_range	A (low, high) tuple indicating the uniform random range of scalars for the a_ary array elements.
b_ary	Second input array.
b_ary_range	A (low, high) tuple inficating the uniform random range of scalars for the b_ary array elements.
b_scalar	Second scalar input.
c_ary	Output array.
MpiArrayUfuncBench_add.cart_comm_dims
goal_time
module	The module used to create array instances.
param_names
params
random_state	A numpy.random.RandomState instance, used to generate random values in arrays.
repeat
ufunc	The numpy.ufunc for this benchmark.
warmup_time