Scatters tasks across processes.
The buckshot library contains multiprocessing functions designed to help developers create distributed applications quickly.
The easiest way to install buckshot is via pip:
$ pip install buckshot
Currently, there are two ways of leveraging buckshot to distribute work across processes: the @distribute decorator and with distributed(...) context manager.
The following example shows how we can distibute the work of calculating many harmonic sum values across processes.
import fractions def harmonic_sum(x): F = fractions.Fraction return sum(F(1, d) for d in xrange(1, x + 1)) for value in range(1, 100): result = harmonic_sum(value) print result
import fractions from buckshot import distribute @distribute(processes=4) def harmonic_sum(x): F = fractions.Fraction return sum(F(1, d) for d in xrange(1, x + 1)) for result in harmonic_sum(range(1, 100)): # Pass in values list. print result
Using with distributed(...)
import fractions from buckshot import distributed def harmonic_sum(x): F = fractions.Fraction return sum(F(1, d) for d in xrange(1, x + 1)) with distributed(harmonic_sum, processes=4) as distributed_harmonic_sum: for result in distributed_harmonic_sum(range(1, 100)): print result
All processes are destroyed when inputs are exhausted and/or the context is exited.
- Function inputs must be picklable.
- Function outputs must be picklable.
- If a child process is killed externally, buckshot will block forever waiting for results.
- This uses os.fork() under the hood, so there is a risk of rapidly exhausting memory.
Read the LICENSE file for details.
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
|Filename, size||File type||Python version||Upload date||Hashes|
|Filename, size buckshot-0.0.10.tar.gz (10.0 kB)||File type Source||Python version None||Upload date||Hashes View hashes|