Coverage for local/lib/python2.7/site-packages/sage/parallel/reference.py : 57%

""" 

Reference Parallel Primitives 

These are reference implementations of basic parallel 

primitives. These are not actually parallel, but work the same way. 

They are good for testing. 

""" 

from __future__ import print_function, absolute_import 

from sage.misc.prandom import shuffle 

def parallel_iter(f, inputs): 

""" 

Reference parallel iterator implementation. 

INPUT: 

- ``f`` -- a Python function that can be pickled using 

the pickle_function command. 

- ``inputs`` -- a list of pickleable pairs (args, kwds), where 

args is a tuple and kwds is a dictionary. 

OUTPUT: 

- iterator over 2-tuples ``(inputs[i], f(inputs[i]))``, where the 

order may be completely random 

EXAMPLES:: 

sage: def f(N,M=10): return N*M 

sage: inputs = [((2,3),{}), (tuple([]), {'N':3,'M':5}), ((2,),{})] 

sage: set_random_seed(0) 

sage: for a, val in sage.parallel.reference.parallel_iter(f, inputs): 

....: print((a, val)) 

(((2,), {}), 20) 

(((), {'M': 5, 'N': 3}), 15) 

(((2, 3), {}), 6) 

sage: for a, val in sage.parallel.reference.parallel_iter(f, inputs): 

....: print((a, val)) 

(((), {'M': 5, 'N': 3}), 15) 

(((2,), {}), 20) 

(((2, 3), {}), 6) 

""" 

v = list(inputs) 

shuffle(v) 

for args, kwds in v: 

yield ((args, kwds), f(*args, **kwds))