3 ms·

Sure thing... from multiprocessing import Pool with Pool(8) as p: p.map(function, sequence))

• pddubs 8y ago
  You could just use threads here too - CPython releases the GIL during IO. from concurrent.futures import ThreadPoolExecutor with ThreadPoolExecutor as executor: executor.map(function, sequence)
• myroon5 8y ago
  Neat. I see that you explicitly specify 8 as the number of processes here. In Java, parallelStream() will pick a sane default for you if you haven't previously specified (based on the number of available processors, I believe). Is something like that possible in Python?
  • jstarfish 8y ago
    The ThreadPool class picks a sane default (number of cores), but I believe it uses Python threads instead of processes.