Parallel Python（PP）

测试环境：i5-2300(4核) + Win7

测试任务还是之前用的takeuptime()函数，串行计算实验的结果可以看这里：http://www.redicecn.com/html/Python/20111223/355.html

使用PP的测试代码如下：

import math, sys, time
import pp

def takeuptime(n):  
    chars = 'abcdefghijklmnopqrstuvwxyz0123456789'  
    s = chars * 1000  
    for i in range(10*n):  
        for c in chars:  
            s.count(c)  

print """Usage: test.py [ncpus]
    [ncpus] - the number of workers to run in parallel, 
    if omitted it will be set to the number of processors in the system
"""

# tuple of all parallel python servers to connect with
ppservers = ()
#ppservers = ("10.0.0.1",)

if len(sys.argv) > 1:
    ncpus = int(sys.argv[1])
    # Creates jobserver with ncpus workers
    job_server = pp.Server(ncpus, ppservers=ppservers)
else:
    # Creates jobserver with automatically detected number of workers
    job_server = pp.Server(ppservers=ppservers)

print "Starting pp with", job_server.get_ncpus(), "workers"

start_time = time.time()

# The following submits 4 jobs
inputs = (1000, 1000, 1000, 1000)
jobs = [(input, job_server.submit(takeuptime, (input,), (), ())) for input in inputs]

#wait for jobs in all groups to finish 
job_server.wait()

print "Time elapsed: ", time.time() - start_time, "s"
job_server.print_stats()

程序运行结果如下：

I:\Webscraping\test>test

Usage: test.py [ncpus]

    [ncpus] - the number of workers to run in parallel,

    if omitted it will be set to the number of processors in the system

Starting pp with 4 workers

Time elapsed:  20.9220001698 s

Job execution statistics:

 job count | % of all jobs | job time sum | time per job | job server

         4 |        100.00 |      78.6590 |    19.664750 | local

Time elapsed since server creation 20.9240000248

所需的时间和之前用pprocess模块进行并行运算的结果差不多。

PP与pprocess模块相比优势在哪里？

1）PP不但支持Linux，Windows下也能使用。

2）PP不但支持单机多核（SMP，systems with multiple processors or cores），而且支持多台计算机（clusters，computers connected via network）。

目前只测试了SMP，期待clusters测试。