python生成器和yield关键字怎么用

这篇文章主要介绍了python生成器和yield关键字怎么用，具有一定借鉴价值，感兴趣的朋友可以参考下，希望大家阅读完这篇文章之后大有收获，下面让小编带着大家一起了解一下。

下列代码用于先体验普通列表推导式和生成器的差别:

# def add():#     temp = ["姓名", "学号", "班级", "电话"]#     dic = {}#     lst = []#     for item in temp:#         inp = input("请输入{}:".fORMat(item))#         if inp == "exit":#             print("成功退出输入")#             return False#         else:#             dic[item] = inp#     lst.append(dic)#     print("添加成功")#     return lst## def show(lst):#     print("-"*30)#     print("姓名\t\t学号\t\t班级\t\t电话")#     print("=" * 30)#     for i in range(len(lst)):#         for val in lst[i].values():#             print(val, "\t", end="")#         print()#     print("-" * 30)## def search(total_lst):#     name = input("请输入您要查询的学生姓名：")#     flag = False#     tmp = []#     for i in range(len(total_lst)):#         if total_lst[i]["姓名"] == name:#             tmp.append(total_lst[i])#             show(tmp)#             flag = True#     if not flag:#         print("抱歉，没有找到该学生")## if __name__ == '__main__':#     total_lst = []#     while True:#         flag = add()#         if flag:#             total_lst = total_lst + flag#         else:#             break#     show(total_lst)#     search(total_lst)## def show(lst):#     print("="*30)#     print("{:^25s}".format("输出F1赛事车手积分榜"))#     print("=" * 30)#     print("{:<10s}".format("排名"), "{:<10s}".format("车手"), "{:<10s}".format("积分"))#     for i in range(len(lst)):#         print("{:0>2d}{:<9s}".format(i+1, ""), "{:<10s}".format(lst[i][0]), "{:<10d}".format(lst[i][1]))## if __name__ == '__main__':#     data = 'lisi 380,jack 256,bob 385,rose 204,alex 212'#     data = data.split(",")#     dic = {}#     da = []#     for i in range(len(data)):#         da.append(data[i].split())#     for i in range(len(da)):#         dic[da[i][0]] = int(da[i][1])#     data2 = sorted(dic.items(), key=lambda kv: (kv[1], kv[0]), reverse=True)#     show(data2)# class Fun:#     def __init__(self):#         print("Fun:__init__()")#     def test(self):#         print("Fun")## class InheritFun(Fun):#     def __init__(self):#         print("InheritedFun.__init__()")#         super().__init__()#     def test(self):#         super().test()#         print("InheritedFun")# a = InheritFun()# a.test()# from math import *# class Circle:#     def __init__(self, radius=1):#         self.radius = radius#     def getPerimeter(self):#         return 2 * self.radius * pi#     def getArea(self):#         return self.radius * self.radius * pi#     def setRadius(self, radius):#         self.radius = radius## a=Circle(10)# print("{:.1f},{:.2f}".format(a.getPerimeter(), a.getArea()))# from math import *# class Root:#     def __init__(self, a, b, c):#         self.a = a#         self.b = b#         self.c = C#     def getDiscriminant(self):#         return pow(self.b, 2)-4*self.a*self.c#     def getRoot1(self):#         return (-self.b+pow(pow(self.b, 2)-4*self.a*self.c, 0.5))/(2*self.a)#     def getRoot2(self):#         return (-self.b - pow(pow(self.b, 2) - 4 * self.a * self.c, 0.5)) / (2 * self.a)# inp = input("请输入a，b，c: ").split(" ")# inp = list(map(int, inp))# Root = Root(inp[0], inp[1], inp[2])# print("判别式为：{:.1f}；  x1：{:.1f};  x2：{:.1f}".format(Root.getDiscriminant(), Root.getRoot1(), Root.getRoot2()))# class Stock:#     def __init__(self, num, name, pre_price, now_price):#         self.num = num#         self.name = name#         self.pre_price = pre_price#         self.now_price = now_price#     def getCode(self):#         return self.num#     def getName(self):#         return self.name#     def getPriceYesterday(self):#         return self.pre_price#     def getPriceToday(self):#         return self.now_price#     def getChangePercent(self):#         return (self.now_price-self.pre_price)/self.pre_price## sCode = input() #输入代码# sName = input() #输入名称# priceYesterday = float(input()) #输入昨日价格# priceToday = float(input()) #输入今日价格# s = Stock(sCode,sName,priceYesterday,priceToday)# print("代码:",s.getCode())# print("名称:",s.getName())# print("昨日价格:%.2f\n今天价格:%.2f" % (s.getPriceYesterday(),s.getPriceToday()))# print("价格变化百分比:%.2f%%" % (s.getChangePercent()*100))# from math import pi## class Shape:#     def __init__(self, name='None', area=None, perimeter=None):#         self.name = name#         self.area = area#         self.perimeter = perimeter#     def calArea(self):#         return self.area#     def calPerimeter(self):#         return self.perimeter#     def display(self):#         print("名称：%s 面积：%.2f 周长：%.2f" % (self.name, self.area, self.perimeter))## class Rectangle(Shape):#     def __init__(self, width, height):#         super().__init__()#         self.width = width#         self.height = height#     def calArea(self):#         self.area = self.height*self.width#         return self.area#     def calPerimeter(self):#         self.perimeter = (self.height+self.width)*2#         return self.perimeter#     def display(self):#         self.name = "Rectangle"#         Rectangle.calArea(self)#         Rectangle.calPerimeter(self)#         super(Rectangle, self).display()## class Triangle(Shape):#     def __init__(self, bottom, height, edge1, edge2):#         super().__init__()#         self.bottom = bottom#         self.height = height#         self.edge1 = edge1#         self.edge2 = edge2#     def calArea(self):#         self.area = (self.bottom*self.height) / 2#         return self.area#     def calPerimeter(self):#         self.perimeter = self.bottom+self.edge2+self.edge1#         return self.perimeter#     def display(self):#         self.name = "Triangle"#         Triangle.calArea(self)#         Triangle.calPerimeter(self)#         super(Triangle, self).display()## class Circle(Shape):#     def __init__(self, radius):#         super(Circle, self).__init__()#         self.radius = radius#     def calArea(self):#         self.area = pi*pow(self.radius, 2)#         return self.area#     def calPerimeter(self):#         self.perimeter = 2*pi*self.radius#         return self.perimeter#     def display(self):#         self.name = "Circle"#         Circle.calArea(self)#         Circle.calPerimeter(self)#         super(Circle, self).display()## rectangle = Rectangle(2, 3)# rectangle.display()## triangle = Triangle(3,4,4,5)# triangle.display()## circle = Circle(radius=1)# circle.display()## lst = list(map(lambda x: int(x), ['1', '2', '3']))# print(lst)## class Listnode(object):#     def __init__(self):#         self.val = None#         self.next = None## #尾插法# def creatlist_tail(lst):#     L = ListNode() #头节点#     first_node = L#     for item in lst:#         p = ListNode()#         p.val = item#         L.next = p#         L = p#     return first_node# #头插法# def creatlist_head(lst):#     L = ListNode() #头节点#     for item in lst:#         p = ListNode()#         p.val = item#         p.next = L#         L = p#     return L# #打印linklist# def print_ll(ll):#     while True:#         if ll.val:#             print(ll.val)#             if ll.next==None: #尾插法停止点#                 break#         elif not ll.next: #头插法停止点#             break#         ll = ll.next# #题解# class Solution:#     def printListFromTailToHead(self, listNode):#         # write code here#         res = []#         while(listNode):#             res.append(listNode.val)#             listNode=listNode.next#         return res[3:0:-1]## if __name__ == "__main__":#     lst = [1, 2, 3]#     linklist = creatlist_tail(lst)#     solution = Solution()#     res = solution.printListFromTailToHead(linklist)#     print(res)# -*- coding:utf-8 -*-# class Solution:#     def __init__(self):#         self.stack1 = []#         self.stack2 = []#     def push(self, node):#         # write code here#         self.stack1.append(node)#     def pop(self):#         # return xx#         if self.stack2:#             return self.stack2.pop()#         else:#             for i in range(len(self.stack1)):#                 self.stack2.append(self.stack1.pop())#             return self.stack2.pop()## if __name__ == '__main__':#     solution = Solution()#     solution.push(1)#     solution.push(2)#     print(solution.pop())#     print(solution.pop())# # binary search# def binary_search(lst, x):#     lst.sort()#     if len(lst) > 0:#         pivot = len(lst) // 2#         if lst[pivot] == x:#             return True#         elif lst[pivot] > x:#             return binary_search(lst[:pivot], x)#         elif lst[pivot] < x:#             return binary_search(lst[pivot+1:], x)#     return False## def binary_search3(lst, x):#     lst.sort()#     head = 0#     tail = len(lst)#     pivot = len(lst) // 2#     while head <= tail:#         if lst[pivot]>x:#             tail = pivot#             pivot = (head+tail) // 2#         elif lst[pivot]<x:#             head = pivot#             pivot = (head+tail) // 2#         elif lst[pivot] == x:#             return True#     return False# if __name__ == '__main__':#     lst = [5, 3, 1, 8, 9]#     print(binary_search(lst, 3))#     print(binary_search(lst, 100))##     print(binary_search(lst, 8))#     print(binary_search(lst, 100))# 括号匹配# def bracket_matching(ans):#     stack = []#     flag = True#     left = ['(', '{', '[']#     right = [')', '}', ']']#     for i in range(len(ans)):#         if ans[i] in left:#             stack.append(ans[i])#         else:#             tmp = stack.pop()#             if left.index(tmp) != right.index(ans[i]):#                 flag = False#     if stack:#         flag = False#     return flag## print(bracket_matching('({})()[[][]'))# print(bracket_matching('({})()[[]]'))# def longestValidParentheses(s):#     maxlen = 0#     stack = []#     for i in range(len(s)):#         if s[i] == '(':#             stack.append(s[i])#         if s[i] == ')' and len(stack) != 0:#             stack.pop()#             maxlen += 2#     return maxlen# print(longestValidParentheses('()(()'))# def GetLongestParentheses(s):#     maxlen = 0#     start = -1#     stack = []#     for i in range(len(s)):#         if s[i]=='(':#             stack.append(i)#         else:#             if not stack:#                 start = i#             else:#                 stack.pop()#                 if not stack:#                     maxlen = max(maxlen, i-start)#                 else:#                     maxlen = max(maxlen, i-stack[-1])#     return maxlen# print(GetLongestParentheses('()(()'))# print(GetLongestParentheses('()(()))'))# print(GetLongestParentheses(')()())'))# import torch# a = torch.tensor([[[1,0,3],#                   [4,6,5]]])# print(a.size())# b = torch.squeeze(a)# print(b, b.size())# b = torch.squeeze(a,-1)# print(b, b.size())# b = torch.unsqueeze(a,2)# print(b, b.size())## print('-----------------')# x = torch.zeros(2, 1, 2, 1, 2)# print(x.size())# y = torch.squeeze(x)# print(y.size())# y = torch.squeeze(x, 0)# print(y.size())# y = torch.squeeze(x, 1)# print(y.size())# from typing import List# class Solution:#     def duplicate(self, numbers: List[int]) -> int:#         # write code here#         dic = dict()#         for i in range(len(numbers)):#             if numbers[i] not in dic.keys():#                 dic[numbers[i]] = 1#             else:#                 dic[numbers[i]] += 1#         for key, value in dic.items():#             if value > 1:#                 return key#         return -1# if __name__ == '__main__':#     solution = Solution()#     print(solution.duplicate([2,3,1,0,2,5,3]))# class TreeNode:#     def __init__(self, data=0):#         self.val = data#         self.left = None#         self.right = None### class Solution:#     def TreeDepth(self , pRoot: TreeNode) -> int:#         # write code here#         if pRoot is None:#             return 0#         count = 0#         now_layer =[pRoot]#         next_layer = []#         while now_layer:#             for i in now_layer:#                 if i.left:#                     next_layer.append(i.left)#                 if i.right:#                     next_layer.append(i.right)#             count +=1#             now_layer, next_layer = next_layer,[]#         return count## if __name__ == '__main__':#     inp = [1,2,3,4,5,'#',6,'#','#',7]#     bt = TreeNode(1)##     bt.left = TreeNode(2)#     bt.right = TreeNode(3)##     bt.left.left = TreeNode(4)#     bt.left.right = TreeNode(5)#     bt.right.left = None#     bt.right.right = TreeNode(6)##     bt.left.left.left = None#     bt.left.left.right = None#     bt.left.right.left = TreeNode(7)##     solution = Solution()#     print('深度：', solution.TreeDepth(bt))# class ListNode:#     def __init__(self):#         self.val = None#         self.next = None## def creatlist_tail(lst):#     L = ListNode()#     first_node = L#     for item in lst:#         p = ListNode()#         p.val = item#         L.next = p#         L = p#     return first_node## def show(node:ListNode):#     print(node.val,end=' ')#     if node.next is not None:#         node = show(node.next)## class Solution:#     def ReverseList(self, head: ListNode) -> ListNode:#         # write code here#         res = None#         while head:#             nextnode = head.next#             head.next = res#             res = head#             head = nextnode#         return res## if __name__ == '__main__':#     lst = [1,2,3]#     linklist = creatlist_tail(lst)#     show(linklist)#     print()#     solution = Solution()#     show(solution.ReverseList(linklist))# 字典推导式# a = ['a', 'b', 'c']# b = [4, 5, 6]# dic = {k:v for k,v in zip(a,b)}# print(dic)#列表推导式# l = [i for i in range(10)]# print(l)#### # 生成器推导式# l1 = (i for i in range(10))# print(type(l1))  # 输出结果：<class 'generator'># for i in l1:#     print(i)# print('{pi:0>10.1f}'.format(pi=3.14159855))# print("'","center".center(40),"'")# print("center".center(40,'-'))# print("center".zfill(40))# print("center".ljust(40,'-'))# print("center".rjust(40,'-'))# s = "Python is easy to learn, easy to use."# print(s.find('to',0,len(s)))# print(s.find('es'))# num = [1,2,3]# print("+".join(str(i) for i in num),"=",sum(num))# print(''.center(40,'-'))## import torch# from torch import nn# import numpy as np## # 一维BN# d1 = torch.rand([2,3,4]) #BCW# bn1 = nn.BatchNorm1d(3, momentum=1)# res = bn1(d1)# print(res.shape)## #二维BN（常用）# d2 = torch.rand([2,3,4,5])  #BCHW# bn2 = nn.BatchNorm2d(3, momentum=1)# res = bn2(d2)# print(res.shape)# print(bn2.running_mean) #3个chanel均值# print(bn2.running_var) #3个chanel方差### a = np.array(d2.tolist())# mean = np.mean(a,axis=(0,2,3))# print(mean)### def batchnorm_forward(x, gamma, beta, bn_param):#     """#     Forward pass for batch normalization##     Input:#     - x: Data of shape (N, D)#     - gamma: Scale parameter of shape (D,)#     - beta: Shift parameter of shape (D,)#     - bn_param: Dictionary with the following keys:#       - mode: 'train' or 'test'#       - eps: Constant for numeric stability#       - momentum: Constant for running mean / variance#       - running_mean: Array of shape(D,) giving running mean of features#       - running_var Array of shape(D,) giving running variance of features#     Returns a tuple of:#     - out: of shape (N, D)#     - cache: A tuple of values needed in the backward pass#     """#     mode = bn_param['mode']#     eps = bn_param.get('eps', 1e-5)#     momentum = bn_param.get('momentum', 0.9)##     N, D = x.shape#     running_mean = bn_param.get('running_mean', np.zeros(D, dtype=x.dtype))#     running_var = bn_param.get('running_var', np.zeros(D, dtype=x.dtype))##     out, cache = None, None##     if mode == 'train':#         sample_mean = np.mean(x, axis=0)  # np.mean([[1,2],[3,4]])->[2,3]#         sample_var = np.var(x, axis=0)#         out_ = (x - sample_mean) / np.sqrt(sample_var + eps)##         running_mean = momentum * running_mean + (1 - momentum) * sample_mean#         running_var = momentum * running_var + (1 - momentum) * sample_var##         out = gamma * out_ + beta#         cache = (out_, x, sample_var, sample_mean, eps, gamma, beta)#     elif mode == 'test':#         # scale = gamma / np.sqrt(running_var + eps)#         # out = x * scale + (beta - running_mean * scale)#         x_hat = (x - running_mean) / (np.sqrt(running_var + eps))#         out = gamma * x_hat + beta#     else:#         raise ValueError('Invalid forward batchnorm mode "%s"' % mode)##     # Store the updated running means back into bn_param#     bn_param['running_mean'] = running_mean#     bn_param['running_var'] = running_var##     return out, cache## import numpy as np# import matplotlib.pyplot as plt### def py_cpu_nms(dets, thresh):##    x1 = dets[:, 0]#    y1 = dets[:, 1]#    x2 = dets[:, 2]#    y2 = dets[:, 3]#    scores = dets[:, 4]#    areas = (x2-x1+1)*(y2-y1+1)#    res = []#    index = scores.argsort()[::-1]#    while index.size>0:#        i = index[0]#        res.append(i)#        x11 = np.maximum(x1[i],x1[index[1:]])#        y11 = np.maximum(y1[i], y1[index[1:]])#        x22 = np.minimum(x2[i],x2[index[1:]])#        y22 = np.minimum(y2[i],y2[index[1:]])##        w = np.maximum(0,x22-x11+1)#        h = np.maximum(0,y22-y11+1)##        overlaps = w * h#        iou = overlaps/(areas[i]+areas[index[1:]]-overlaps)##        idx = np.where(iou<=thresh)[0]#        index = index[idx+1]#    print(res)#    return res## def plot_boxs(box,c):#     x1 = box[:, 0]#     y1 = box[:, 1]#     x2 = box[:, 2]#     y2 = box[:, 3]##     plt.plot([x1,x2],[y1,y1],c)#     plt.plot([x1,x2],[y2,y2],c)#     plt.plot([x1,x1],[y1,y2],c)#     plt.plot([x2,x2],[y1,y2],c)## if __name__ == '__main__':#     boxes = np.array([[100, 100, 210, 210, 0.72],#                       [250, 250, 420, 420, 0.8],#                       [220, 220, 320, 330, 0.92],#                       [230, 240, 325, 330, 0.81],#                       [220, 230, 315, 340, 0.9]])#     plt.figure()#     ax1 = plt.subplot(121)#     ax2 = plt.subplot(122)#     plt.sca(ax1)#     plot_boxs(boxes,'k')##     res = py_cpu_nms(boxes,0.7)#     plt.sca(ax2)#     plot_boxs(boxes[res],'r')#     plt.show()# 2 3 3 4# 1 2 3# 4 5 6# 1 2 3 4# 5 6 7 8# 9 10 11 12# lst1, lst2 = [], []# n1,m1,n2,m2 = map(int,input().split())# for i in range(n1):#     nums = list(map(int,input().split())) #输入一行数据#     lst1.append(nums)# for i in range(n2):#     nums = list(map(int,input().split()))#     lst2.append(nums)# res = []# for i in range(n1):#     res.append([])#     for j in range(m2):#         lst4 = []#         lst3 = lst1[i]#         for k in range(n2):#             lst4.append(lst2[k][j])#         res_num = sum(map(lambda x,y:x*y,lst3,lst4))#         res[i].append(res_num)# print(res)## import numpy as np# print('numpy:',np.dot(lst1,lst2))#定义残差块# import torch# import torch.nn as nn# import torch.nn.functional as F## class ResBlock(nn.Module):#     def __init__(self,inchanel,outchanel,stride=1):#         super(ResBlock,self).__init__()#         self.left = nn.Sequential(#             nn.Conv2d(inchanel,outchanel,kernel_size=3,stride=stride,padding=1,bias=False),#             nn.BatchNorm2d(outchanel),#             nn.ReLU(inplace=True),#             nn.Conv2d(outchanel,outchanel,kernel_size=3,stride=1,padding=1,bias=False),#             nn.BatchNorm2d(outchanel)#         )#         self.shortcut = nn.Sequential()#         if stride!=1 or inchanel!=outchanel:#             self.shortcut = nn.Sequential(#                 nn.Conv2d(inchanel,outchanel,kernel_size=1,stride=stride,padding=1,bias=False),#                 nn.BatchNorm2d(outchanel)#             )#     def forward(self,x):#         out = self.left(x)#         out = out + self.shortcut(x)#         out = F.relu(out)##         return out## class ResNet(nn.Module):#     def __init__(self,Resblock,num_classes=10):#         super(ResNet,self).__init__()#         self.inchanel = 64#         self.conv1 = nn.Sequential(#             nn.Conv2d(3,64,kernel_size=3,stride=1,padding=1,bias=False),#             nn.BatchNorm2d(64),#             nn.ReLU()#         )#         self.layer1 = self.make_layer(ResBlock,64,2,1)#         self.layer2 = self.make_layer(ResBlock, 128, 2, 2)#         self.layer3 = self.make_layer(ResBlock, 256, 2, 2)#         self.layer4 = self.make_layer(ResBlock, 512, 2, 2)#         self.fc = nn.Linear(512,num_classes)##     def make_layer(self,ResBlock,channels,num_blocks,stride):#         strides = [stride] + [1] * (num_blocks-1)#         layers = []#         for stride in strides:#             layers.append(ResBlock(self.inchanel,channels,stride))#             self.inchanel=channels#         return nn.Sequential(*layers)#     def forward(self,x):#         out = self.conv1(x)#         out = self.layer1(out)#         out = self.layer2(out)#         out = self.layer3(out)#         out = self.layer4(out)#         out = F.avg_pool2d(out,4)#         out = out.view(out.size(0),-1)#         out = self.fc(out)#         return out# import torch# import torch.nn as nn# import torch.nn.functional as F## class ASPP(nn.Module):#     def __init__(self,in_channel=512,depth=256):#         super(ASPP,self).__init__()#         self.mean = nn.AdaptiveAvgPool2d((1,1))#         self.conv = nn.Conv2d(in_channel,depth,1,1)#         self.atrous_block1 = nn.Conv2d(in_channel,depth,1,1)#         self.atrous_block6 = nn.Conv2d(in_channel,depth,3,1,padding=6,dilation=6)#         self.atrous_block12 = nn.Conv2d(in_channel,depth,3,1,padding=12,dilation=12)#         self.atrous_block18 = nn.Conv2d(in_channel,depth,3,1,padding=18,dilation=18)#         self.conv1x1_output = nn.Conv2d(depth*5,depth,1,1)#     def forward(self,x):#         size = x[2:]#         pool_feat = self.mean(x)#         pool_feat = self.conv(pool_feat)#         pool_feat = F.upsample(pool_feat,size=size,mode='bilinear')##         atrous_block1 = self.atrous_block1(x)#         atrous_block6 = self.atrous_block6(x)#         atrous_block12 = self.atrous_block12(x)#         atrous_block18 = self.atrous_block18(x)##         out = self.conv1x1_output(torch.cat([pool_feat,atrous_block1,atrous_block6,#                                              atrous_block12,atrous_block18],dim=1))#         return out#牛顿法求三次根# def sqrt(n):#     k = n#     while abs(k*k-n)>1e-6:#         k = (k + n/k)/2#     print(k)## def cube_root(n):#     k = n#     while abs(k*k*k-n)>1e-6:#         k = k + (k*k*k-n)/3*k*k#     print(k)# sqrt(2)# cube_root(8)# -*- coding:utf-8 -*-# import random## import numpy as np# from matplotlib import pyplot### class K_Means(object):#     # k是分组数；tolerance‘中心点误差'；max_iter是迭代次数#     def __init__(self, k=2, tolerance=0.0001, max_iter=300):#         self.k_ = k#         self.tolerance_ = tolerance#         self.max_iter_ = max_iter##     def fit(self, data):#         self.centers_ = {}#         for i in range(self.k_):#             self.centers_[i] = data[random.randint(0,len(data))]#         # print('center', self.centers_)#         for i in range(self.max_iter_):#             self.clf_ = {} #用于装归属到每个类中的点[k,len(data)]#             for i in range(self.k_):#                 self.clf_[i] = []#             # print("质点:",self.centers_)#             for feature in data:#                 distances = [] #装中心点到每个点的距离[k]#                 for center in self.centers_:#                     # 欧拉距离#                     distances.append(np.linalg.norm(feature - self.centers_[center]))#                 classification = distances.index(min(distances))#                 self.clf_[classification].append(feature)##             # print("分组情况:",self.clf_)#             prev_centers = dict(self.centers_)##             for c in self.clf_:#                 self.centers_[c] = np.average(self.clf_[c], axis=0)##             # '中心点'是否在误差范围#             optimized = True#             for center in self.centers_:#                 org_centers = prev_centers[center]#                 cur_centers = self.centers_[center]#                 if np.sum((cur_centers - org_centers) / org_centers * 100.0) > self.tolerance_:#                     optimized = False#             if optimized:#                 break##     def predict(self, p_data):#         distances = [np.linalg.norm(p_data - self.centers_[center]) for center in self.centers_]#         index = distances.index(min(distances))#         return index### if __name__ == '__main__':#     x = np.array([[1, 2], [1.5, 1.8], [5, 8], [8, 8], [1, 0.6], [9, 11]])#     k_means = K_Means(k=2)#     k_means.fit(x)#     for center in k_means.centers_:#         pyplot.scatter(k_means.centers_[center][0], k_means.centers_[center][1], marker='*', s=150)##     for cat in k_means.clf_:#         for point in k_means.clf_[cat]:#             pyplot.scatter(point[0], point[1], c=('r' if cat == 0 else 'b'))##     predict = [[2, 1], [6, 9]]#     for feature in predict:#         cat = k_means.predict(feature)#         pyplot.scatter(feature[0], feature[1], c=('r' if cat == 0 else 'b'), marker='x')##     pyplot.show()# def pred(key, value):#     if key == 'math':#         return value>=40#     else:#         return value>=60# def func(dic,pred):#     # temp = []#     # for item in dic:#     #     if not pred(item,dic[item]):#     #         temp.append(item)#     # for item in temp:#     #     del dic[item]#     # return dic##     for k in list(dic.keys()):#         if dic[k]<60:#             del dic[k]#     return dic## if __name__ == '__main__':#     dic={'math':66,'c':78,'c++':59,'python':55}#     dic = func(dic,pred)#     print(dic)## class TreeNode:#     def __init__(self):#         self.left = None#         self.right = None#         self.data = None## def insert(tree,x):#     temp = TreeNode()#     temp.data = x#     if tree.data>x:#         if tree.left == None:#             tree.left = temp#         else:#             insert(tree.left,x)#     else:#         if tree.right == None:#             tree.right = temp#         else:#             insert(tree.right,x)## def print_tree(node):#     if node is None:#         return 0#     print_tree(node.left)#     print(node.data)#     print_tree(node.right)### def sort(lst):#     tree = TreeNode()#     tree.data = lst[0]#     for i in range(1, len(lst)):#         insert(tree,lst[i])#     print_tree(tree)## sort([5,2,4])# from collections import Iterable, Iterator### class Person(object):#     """定义一个人类"""##     def __init__(self):#         self.name = list()#         self.name_num = 0##     def add(self, name):#         self.name.append(name)##     def __iter__(self):#         return self#     def __next__(self):#         # 记忆性返回数据#         if self.name_num < len(self.name):#             ret = self.name[self.name_num]#             self.name_num += 1#             return ret#         else:#             raise StopIteration## person1 = Person()# person1.add("张三")# person1.add("李四")# person1.add("王五")## print("判断是否是可迭代的对象：", isinstance(person1, Iterable))# print("判断是否是迭代器：", isinstance(person1,Iterator))# for name in person1:#     print(name)# nums = []# a = 0# b = 1# i = 0# while i < 10:#     nums.append(a)#     a,b = b,a+b#     i += 1# for i in nums:#     print(i)## class Fb():#     def __init__(self):#         self.a = 0#         self.b = 1#         self.i = 0#     def __iter__(self):#         return self#     def __next__(self):#         res = self.a#         if self.i<10:#             self.a,self.b = self.b,self.a+self.b#             self.i += 1#             return res#         else:#             raise StopIteration## fb = Fb()# for i in fb:#     print(i)import timedef get_time(func):    def wraper(*args, **kwargs):        start_time = time.time()        result = func(*args, **kwargs)        end_time = time.time()        print("Spend:", end_time - start_time)        return result    return wraper@get_timedef _list(n):    l = [i*i*i for i in range(n)]@get_timedef _generator(n):    ge = (i*i*i for i in range(n))@get_timedef _list_print(l1):    for i in l1:        print(end='')@get_timedef _ge_print(ge):    for i in ge:        print(end='')n = 100000print('list 生成耗时：')_list(n)print('生成器 生成耗时：')_generator(n)l1 = [i*i*i for i in range(n)]ge = (i*i*i for i in range(n))# print(l1)# print(ge)print('list遍历耗时：')_list_print(l1)print('生成器遍历耗时：')_ge_print(ge)

结论：

生成速度：生成器>列表
for_in_循环遍历：1、速度方面：列表>生成器；2、内存占用方面：列表<生成器
总的来说，生成器就是用于降低内存消耗的。

感谢你能够认真阅读完这篇文章，希望小编分享的“python生成器和yield关键字怎么用”这篇文章对大家有帮助，同时也希望大家多多支持编程网，关注编程网Python频道，更多相关知识等着你来学习!