DDR爱好者之家 Design By 杰米

摘要

本文主要介绍了利用python的 threading和queue库实现多线程编程,并封装为一个类,方便读者嵌入自己的业务逻辑。最后以机器学习的一个超参数选择为例进行演示。

多线程实现逻辑封装

实例化该类后,在.object_func函数中加入自己的业务逻辑,再调用.run方法即可。

# -*- coding: utf-8 -*-
# @Time : 2021/2/4 14:36
# @Author : CyrusMay WJ
# @FileName: run.py
# @Software: PyCharm
# @Blog :https://blog.csdn.net/Cyrus_May
import queue
import threading

class CyrusThread(object):
  def __init__(self,num_thread = 10,logger=None):
    """
    
    :param num_thread: 线程数
    :param logger: 日志对象
    """
    self.num_thread = num_thread
    self.logger = logger

  def object_func(self,args_queue,max_q):
    while 1:
      try:
        arg = args_queue.get_nowait()
        step = args_queue.qsize()
        self.logger.info("progress:{}\{}".format(max_q,step))
      except:
        self.logger.info("no more arg for args_queue!")
        break
        
        
        """
        此处加入自己的业务逻辑代码
        """
        
        
  def run(self,args):
    args_queue = queue.Queue()
    for value in args:
      args_queue.put(value)
    threads = []
    for i in range(self.num_thread):
      threads.append(threading.Thread(target=self.object_func,args = args_queue))
    for t in threads:
      t.start()
    for t in threads:
      t.join()

模型参数选择实例

# -*- coding: utf-8 -*-
# @Time : 2021/2/4 14:36
# @Author : CyrusMay WJ
# @FileName: run.py
# @Software: PyCharm
# @Blog :https://blog.csdn.net/Cyrus_May
import queue
import threading
import numpy as np
from sklearn.datasets import load_boston
from sklearn.svm import SVR
import logging
import sys


class CyrusThread(object):
  def __init__(self,num_thread = 10,logger=None):
    """

    :param num_thread: 线程数
    :param logger: 日志对象
    """
    self.num_thread = num_thread
    self.logger = logger

  def object_func(self,args_queue,max_q):
    while 1:
      try:
        arg = args_queue.get_nowait()
        step = args_queue.qsize()
        self.logger.info("progress:{}\{}".format(max_q,max_q-step))
      except:
        self.logger.info("no more arg for args_queue!")
        break
      # 业务代码
      C, epsilon, gamma = arg[0], arg[1], arg[2]
      svr_model = SVR(C=C, epsilon=epsilon, gamma=gamma)
      x, y = load_boston()["data"], load_boston()["target"]
      svr_model.fit(x, y)
      self.logger.info("score:{}".format(svr_model.score(x,y)))


  def run(self,args):
    args_queue = queue.Queue()
    max_q = 0
    for value in args:
      args_queue.put(value)
      max_q += 1
    threads = []
    for i in range(self.num_thread):
      threads.append(threading.Thread(target=self.object_func,args = (args_queue,max_q)))
    for t in threads:
      t.start()
    for t in threads:
      t.join()

# 创建日志对象
logger = logging.getLogger()
logger.setLevel(logging.INFO)
screen_handler = logging.StreamHandler(sys.stdout)
screen_handler.setLevel(logging.INFO)
formatter = logging.Formatter('%(asctime)s - %(module)s.%(funcName)s:%(lineno)d - %(levelname)s - %(message)s')
screen_handler.setFormatter(formatter)
logger.addHandler(screen_handler)

# 创建需要调整参数的集合
args = []
for C in [i for i in np.arange(0.01,1,0.01)]:
  for epsilon in [i for i in np.arange(0.001,1,0.01)] + [i for i in range(1,10,1)]:
    for gamma in [i for i in np.arange(0.001,1,0.01)] + [i for i in range(1,10,1)]:
      args.append([C,epsilon,gamma])

# 创建多线程工具
threading_tool = CyrusThread(num_thread=20,logger=logger)
threading_tool.run(args)

运行结果

2021-02-04 20:52:22,824 - run.object_func:31 - INFO - progress:1176219\1
2021-02-04 20:52:22,824 - run.object_func:31 - INFO - progress:1176219\2
2021-02-04 20:52:22,826 - run.object_func:31 - INFO - progress:1176219\3
2021-02-04 20:52:22,833 - run.object_func:31 - INFO - progress:1176219\4
2021-02-04 20:52:22,837 - run.object_func:31 - INFO - progress:1176219\5
2021-02-04 20:52:22,838 - run.object_func:31 - INFO - progress:1176219\6
2021-02-04 20:52:22,841 - run.object_func:31 - INFO - progress:1176219\7
2021-02-04 20:52:22,862 - run.object_func:31 - INFO - progress:1176219\8
2021-02-04 20:52:22,873 - run.object_func:31 - INFO - progress:1176219\9
2021-02-04 20:52:22,884 - run.object_func:31 - INFO - progress:1176219\10
2021-02-04 20:52:22,885 - run.object_func:31 - INFO - progress:1176219\11
2021-02-04 20:52:22,897 - run.object_func:31 - INFO - progress:1176219\12
2021-02-04 20:52:22,900 - run.object_func:31 - INFO - progress:1176219\13
2021-02-04 20:52:22,904 - run.object_func:31 - INFO - progress:1176219\14
2021-02-04 20:52:22,912 - run.object_func:31 - INFO - progress:1176219\15
2021-02-04 20:52:22,920 - run.object_func:31 - INFO - progress:1176219\16
2021-02-04 20:52:22,920 - run.object_func:39 - INFO - score:-0.01674283914287855
2021-02-04 20:52:22,929 - run.object_func:31 - INFO - progress:1176219\17
2021-02-04 20:52:22,932 - run.object_func:39 - INFO - score:-0.007992354170952565
2021-02-04 20:52:22,932 - run.object_func:31 - INFO - progress:1176219\18
2021-02-04 20:52:22,945 - run.object_func:31 - INFO - progress:1176219\19
2021-02-04 20:52:22,954 - run.object_func:31 - INFO - progress:1176219\20
2021-02-04 20:52:22,978 - run.object_func:31 - INFO - progress:1176219\21
2021-02-04 20:52:22,984 - run.object_func:39 - INFO - score:-0.018769934807246536
2021-02-04 20:52:22,985 - run.object_func:31 - INFO - progress:1176219\22

DDR爱好者之家 Design By 杰米
广告合作:本站广告合作请联系QQ:858582 申请时备注:广告合作(否则不回)
免责声明:本站资源来自互联网收集,仅供用于学习和交流,请遵循相关法律法规,本站一切资源不代表本站立场,如有侵权、后门、不妥请联系本站删除!
DDR爱好者之家 Design By 杰米

《魔兽世界》大逃杀!60人新游玩模式《强袭风暴》3月21日上线

暴雪近日发布了《魔兽世界》10.2.6 更新内容,新游玩模式《强袭风暴》即将于3月21 日在亚服上线,届时玩家将前往阿拉希高地展开一场 60 人大逃杀对战。

艾泽拉斯的冒险者已经征服了艾泽拉斯的大地及遥远的彼岸。他们在对抗世界上最致命的敌人时展现出过人的手腕,并且成功阻止终结宇宙等级的威胁。当他们在为即将于《魔兽世界》资料片《地心之战》中来袭的萨拉塔斯势力做战斗准备时,他们还需要在熟悉的阿拉希高地面对一个全新的敌人──那就是彼此。在《巨龙崛起》10.2.6 更新的《强袭风暴》中,玩家将会进入一个全新的海盗主题大逃杀式限时活动,其中包含极高的风险和史诗级的奖励。

《强袭风暴》不是普通的战场,作为一个独立于主游戏之外的活动,玩家可以用大逃杀的风格来体验《魔兽世界》,不分职业、不分装备(除了你在赛局中捡到的),光是技巧和战略的强弱之分就能决定出谁才是能坚持到最后的赢家。本次活动将会开放单人和双人模式,玩家在加入海盗主题的预赛大厅区域前,可以从强袭风暴角色画面新增好友。游玩游戏将可以累计名望轨迹,《巨龙崛起》和《魔兽世界:巫妖王之怒 经典版》的玩家都可以获得奖励。