多处理问题，使用池和地图

Question

我想通过使用python中的pool和map函数来获得输出，具体如下。

**按随机顺序输出。

PoolWorker-10输出： [0, 4, 8, 12, 16, 20]

PoolWorker-11输出。[0, 4, 8, 12, 16, 20]

PoolWorker-12输出。[0, 4, 8, 12, 16, 20]

PoolWorker-1输出。[0, 4, 8, 12, 16, 20]

.....**

我多次尝试将输出结果和current_process名称写在一起，但都失败了。我不知道应该改变哪一项才能得到上面那样的打印输出。

from multiprocessing import Pool, current_process

def multiprocessing_func():
   input_list = [0, 1, 2, 3, 4, 5]
   print("input list: {}".format(input_list))
   print("Output in random order: ")

   pool = Pool(10, current_process_func)
   result = pool.map(calculate, input_list)
   print(result)
   pool.close()
   pool.join()

def calculate(y_list):
   return y_list * 4

def current_process_func():
   print("{} output: ".format(current_process().name))

if __name__ == '__main__':
   multiprocessing_func()

---

当我运行它时，我得到的结果是这样的。

按随机顺序输出：

PoolWorker -1输出：

PoolWorker-2输出：

PoolWorker-3输出：

PoolWorker-4输出：PoolWorker-5输出：PoolWorker-4输出

PoolWorker-5输出：PoolWorker-7输出：PoolWorker-6输出

PoolWorker-7输出：PoolWorker-7输出

PoolWorker-6输出。

[0, 4, 8, 12, 16, 20]

PoolWorker-8输出。

PoolWorker -9输出:

PoolWorker-10输出。

---

我应该怎么改才能得到我想要的结果？

Answer 1

修改了你的代码段，如下图所示。总体考虑如下。

• pool.map(func, data) data应该被理解为分布在cpu池中的数据（也是这个进程的专属数据）。进一步说至少应该有和进程一样多的项目，以充分利用池子。我只是把10个相同的列表放在那里）。

• 中跳过了初始化参数。Pool() 构，只需在计算完成后输出想要的数据即可。

  from multiprocessing import Pool, current_process
  
  def multiprocessing_func():
      input_list = [0, 1, 2, 3, 4, 5]
      print("input list: {}".format(input_list))
      print("Output in random order: ")
  
      pool = Pool(10) 
      result = pool.map(calculate, [input_list]*10)
  
      print()
      print("collected:" , result)
      pool.close()
      pool.join()
  
  def calculate(y_list):
      out = [x*4 for x in y_list] 
      current_process_func(out)
      return out   
  
  def current_process_func(data):
      print("{} output: {}".format(current_process().name, data))
  
  if __name__ == '__main__':
      multiprocessing_func()
  

导致。

input list: [0, 1, 2, 3, 4, 5]
Output in random order: 
ForkPoolWorker-52 output: [0, 4, 8, 12, 16, 20]
ForkPoolWorker-55 output: [0, 4, 8, 12, 16, 20]
ForkPoolWorker-57 output: [0, 4, 8, 12, 16, 20]
ForkPoolWorker-51 output: [0, 4, 8, 12, 16, 20]
ForkPoolWorker-53 output: [0, 4, 8, 12, 16, 20]
ForkPoolWorker-54 output: [0, 4, 8, 12, 16, 20]
ForkPoolWorker-56 output: [0, 4, 8, 12, 16, 20]
ForkPoolWorker-58 output: [0, 4, 8, 12, 16, 20]
ForkPoolWorker-60 output: [0, 4, 8, 12, 16, 20]
ForkPoolWorker-59 output: [0, 4, 8, 12, 16, 20]

collected: [[0, 4, 8, 12, 16, 20], [0, 4, 8, 12, 16, 20], [0, 4,    8, 12, 16, 20], [0, 4, 8, 12, 16, 20], [0, 4, 8, 12, 16, 20], [0, 4, 8, 12, 16, 20], [0, 4, 8, 12, 16, 20], [0, 4, 8, 12, 16, 20], [0, 4, 8, 12, 16, 20], [0, 4, 8, 12, 16, 20]]

这是你所期望的吗？