python学习的第三天

1.三国TOP10人物分析

1.读取小说内容
2.分词
3.词语过滤，删除无关词、重复分词
4.排序
5.得出结论

import jieba
# 1. 读取小说内容
with open('./novel/threekingdom.txt', 'r', encoding='utf-8') as f:
    words = f.read()
    counts = {} #{'曹操': 234, '回寨': 56}
# 2.分词
    words_list = jieba.lcut(words)
    for word in words_list:
        if len(word) <= 1:
            continue
        else:
            #向字典中更新字典中的值
            #counts[word] = 取出字典中原来键对应的值 + 1
            # counts[word] = counts[word] + 1  counts[word]没有就会报错
            #字典.get(k) 如果字典中没有这个键 返回 none
            counts[word] = counts.get(word, 0) + 1
    print(counts)
# 3.词语过滤，删除无关词、重复分词
    # 4.排序 [(), ()]
    items = list(counts.items())
    print('排序前的列表', items)
    def sort_by_count(x):
        return x[1]
    items.sort(key=sort_by_count, reverse=True)
    for i in range(20):
        #序列解包
        role, count = items[i]
        print(role, count)

排除不是人名的分词，合并人名，然后排出top10

exclude = {"将军", "却说", "丞相", "二人", "不可", "荆州", "不能", "如此", "商议",
               "如何", "主公", "军士", "军马", "左右", "次日", "引兵", "大喜", "天下",
               "东吴", "于是", "今日", "不敢", "魏兵", "陛下", "都督", "人马", "不知", 
               "孔明曰", "玄德曰", "刘备", "云长"}

   counts['孔明'] = counts['孔明'] + counts['孔明曰']
   counts['玄德'] = counts['玄德'] + counts['玄德曰'] + counts['刘备']
   counts['关公'] = counts['关公'] + counts['云长']
   for word in exclude:
       del counts[word]

最终代码：（其中collocations=False ：取消相邻两个重复词之间的匹配)

import jieba
from wordcloud import WordCloud
import imageio
# 1. 读取小说内容
with open('./novel/threekingdom.txt', 'r', encoding='utf-8') as f:
    words = f.read()
    counts = {} #{'曹操': 234, '回寨': 56}
    exclude = {"将军", "却说", "丞相", "二人", "不可", "荆州", "不能", "如此", "商议",
               "如何", "主公", "军士", "军马", "左右", "次日", "引兵", "大喜", "天下",
               "东吴", "于是", "今日", "不敢", "魏兵", "陛下", "都督", "人马", "不知",
               "孔明曰", "玄德曰", "刘备", "云长"}

    # 2.分词
    words_list = jieba.lcut(words)
    for word in words_list:
        if len(word) <= 1:
            continue
        else:
            #向字典中更新字典中的值
            #counts[word] = 取出字典中原来键对应的值 + 1
            # counts[word] = counts[word] + 1  counts[word]没有就会报错
            #字典.get(k) 如果字典中没有这个键 返回 none
            counts[word] = counts.get(word, 0) + 1
    print(counts)
    # 3.词语过滤，删除无关词、重复分词
    counts['孔明'] = counts['孔明'] + counts['孔明曰']
    counts['玄德'] = counts['玄德'] + counts['玄德曰'] + counts['刘备']
    counts['关公'] = counts['关公'] + counts['云长']
    for word in exclude:
        del counts[word]
    # 4.排序 [(), ()]
    items = list(counts.items())
    print('排序前的列表', items)
    def sort_by_count(x):
        return x[1]
    items.sort(key=sort_by_count, reverse=True)

    li = []  # ['孔明',孔明,孔明,'曹操'。。。。。]
    for i in range(10):
        #序列解包
        role, count = items[i]
        print(role, count)
        # _是告诉看代码的人，循环里面不需要使用临时变量
        for _ in range(count):
            li.append(role)
    # 5.得出结论
    mask = imageio.imread('./china.jpg')
    text = ' '.join(li)
    WordCloud(
        font_path='msyh.ttc',
        background_color='white',
        width=800,
        height=600,
        mask=mask,
        # 相邻两个重复词之间的匹配
        collocations=False
    ).generate(text).to_file('top10.png')

2.匿名函数

# 匿名函数
# 结构
# lambda x1, x2....xn: 表达式
sum_num = lambda x1, x2: x1+x2
print(sum_num(2, 3))
# # 参数可以是无限多个，但是表达式只有一个
name_info_list = [
    ('张三',4500),
    ('李四',9900),
    ('王五',2000),
    ('赵六',5500),
]
name_info_list.sort(key=lambda x:x[1], reverse=True)
print(name_info_list)
stu_info = [
    {"name":'zhangsan', "age":18},
    {"name":'lisi', "age":30},
    {"name":'wangwu', "age":99},
    {"name":'tiaqi', "age":3},
]
stu_info.sort(key=lambda i:i['age'])
print(stu_info)
# 列表推导式，列表解析个字典解析
# 之前我们使用普通for 创建列表
li = []
for i in range(10):
    li.append(i)
print(li)
# # 使用列表推导式
# # [表达式 for 临时变量 in 可迭代对象 可以追加条件]
print([i for i in range(10)])
# 列表解析
# # 筛选出列表中所有的偶数
li = []
for i in range(10):
    if i%2 == 0:
        li.append(i)
print(li)
# # 使用列表解析
print([i for i in range(10) if i%2 == 0])
# 筛选出列表中 大于0 的数
from random import randint
num_list = [randint(-10, 10) for _ in range(10)]
print(num_list)
print([i for i in num_list if i>0])

# 字典解析

# 生成100个学生的成绩
stu_grades = {'student{}'.format(i):randint(50, 100) for i in range(1, 101)}
print(stu_grades)

# 筛选大于 60分的所有学生
print({k: v for k, v in stu_grades.items() if v >60})

3. Matplotlib

Matplotlib 是一个Python的2D绘图库，它以各种硬拷贝格式和跨平台的交互式环境生成出版质量级别的图形 。
通过 Matplotlib，开发者可以仅需要几行代码，便可以生成绘图，直方图，功率谱，条形图，错误图，散点图等。

绘制图形

# matplotlib
#  导入
from matplotlib import pyplot as plt
plt.rcParams["font.sans-serif"] = ['SimHei']
plt.rcParams['axes.unicode_minus'] = False
import numpy as np

# #  使用100个点 绘制 [0 , 2π]正弦曲线图
# #.linspace 左闭右闭区间的等差数列
x = np.linspace(0, 2*np.pi, num=100)
print(x)
y = np.sin(x)
# #  正弦和余弦在同一坐标系下
cosy = np.cos(x)
plt.plot(x, y, color='g', linestyle='--',label='sin(x)')
plt.plot(x, cosy, color='r',label='cos(x)')
plt.xlabel('时间（s）')
plt.ylabel('电压（V）')
plt.title('欢迎来到python世界')
# # 图例
plt.legend()
plt.show()

# 绘制柱状图
import string
from random import randint
# print(string.ascii_uppercase[0:6])
# ['A', 'B', 'C'...]
x = ['口红{}'.format(x) for x in string.ascii_uppercase[:5] ]
y = [randint(200, 500) for _ in range(5)]
print(x)
print(y)
plt.xlabel('口红品牌')
plt.ylabel('价格（元）')
plt.bar(x, y)
plt.show()

#绘制饼图
from random import randint
import string
counts = [randint(3500, 9000) for _ in range(6)]
labels = ['员工{}'.format(x) for x in string.ascii_lowercase[:6] ]
# # 距离圆心点距离
explode = [0.1,0,0, 0, 0,0]
colors = ['red', 'purple','blue', 'yellow','gray','green']
plt.pie(counts,explode = explode,shadow=True, labels=labels, autopct = '%1.1f%%',colors=colors)
plt.legend(loc=2)
plt.axis('equal')
plt.show()

# 绘制散点图
# 均值为 0 标准差为1 的正太分布数据
x = np.random.normal(0, 1, 100)
y = np.random.normal(0, 1, 100)
plt.scatter(x, y)
plt.show()
x = np.random.normal(0, 1, 1000000)
y = np.random.normal(0, 1, 1000000)
# alpha透明度
plt.scatter(x, y, alpha=0.1)
plt.show()

4.练习

4.1 红楼梦TOP10人物分析

import jieba
from wordcloud import WordCloud
# 1.读取小说内容
with open('./all.txt', 'r', encoding='utf-8') as f:
    words = f.read()

    counts = {}
    excludes = {"什么", "一个", "我们", "你们", "如今", "说道", "知道", "起来", "这里",
               "出来", "众人", "那里", "自己", "一面", "只见", "太太", "两个", "没有",
               "怎么", "不是", "不知", "这个", "听见", "这样", "进来", "咱们", "就是",
               "老太太", "东西", "告诉", "回来", "只是", "大家", "姑娘", "奶奶", "凤姐儿"}
    # 2. 分词
    words_list = jieba.lcut(words)
    # print(words_list)
    for word in words_list:
        if len(word) <= 1:
            continue
        else:
            # 更新字典中的值
            # counts[word] = 取出字典中原来键对应的值 + 1
            # counts[word] = counts[word] + 1  # counts[word]如果没有就要报错
            # 字典。get（k） 如果字典中没有这个键 返回 NONE
            counts[word] = counts.get(word, 0) + 1

    print(len(counts))
    # 3. 词语过滤,删除无关词，重复词
    counts['贾母'] = counts['老太太'] + counts['贾母']
    counts['林黛玉'] = counts['林妹妹'] + counts['黛玉']
    counts['贾宝玉'] = counts['宝玉'] +counts['贾宝玉']
    for word in excludes:
        del counts[word]

    # 4.排序 [(), ()]
    items = list(counts.items())
    print(items)

    def sort_by_count(x):
        return x[1]
    items.sort(key=sort_by_count, reverse=True)

    li = []  
    for i in range(10):
        # 序列解包
        role, count = items[i]
        print(role, count)
        # _ 是告诉看代码的人，循环里面不需要使用临时变量
        for _ in range(count):
            li.append(role)

    # 5得出结论

    text = ' '.join(li)
    WordCloud(
        font_path='msyh.ttc',
        background_color='black',
        width=800,
        height=600,
        # 相邻两个重复词之间的匹配
        collocations=False
    ).generate(text).to_file('top10.png')

红楼梦人物分析

4.2 绘制三国top10人物饼图

#绘制三国人物TOP10饼图
import jieba
from matplotlib import pyplot as plt
plt.rcParams["font.sans-serif"] = ['SimHei']
plt.rcParams['axes.unicode_minus'] = False
# 1.读取小说内容
with open('./novel/threekingdom.txt', 'r', encoding='utf-8') as f:
    words = f.read()
    counts = {}  # {‘曹操’:234，‘回寨’:56}
    excludes = {"将军", "却说", "丞相", "二人", "不可", "荆州", "不能", "如此", "商议",
                "如何", "主公", "军士", "军马", "左右", "次日", "引兵", "大喜", "天下",
                "东吴", "于是", "今日", "不敢", "魏兵", "陛下", "都督", "人马", "不知",
                "孔明曰","玄德曰","刘备","云长"}
    # 2. 分词
    words_list = jieba.lcut(words)
    # print(words_list)
    for word in words_list:
        if len(word) <= 1:
            continue
        else:
            # 更新字典中的值
            # counts[word] = 取出字典中原来键对应的值 + 1
            # counts[word] = counts[word] + 1  # counts[word]如果没有就要报错
            # 字典。get（k） 如果字典中没有这个键 返回 NONE
            counts[word] = counts.get(word, 0) + 1

    print(len(counts))
    # 3. 词语过滤,删除无关词，重复词
    counts['孔明'] =  counts['孔明'] +  counts['孔明曰']
    counts['玄德'] = counts['玄德'] + counts['玄德曰'] +counts['刘备']
    counts['关公'] = counts['关公'] +counts['云长']
    for word in excludes:
        del counts[word]

    # 4.排序 [(), ()]
    items = list(counts.items())
    print(items)

    def sort_by_count(x):
        return x[1]
    items.sort(key=sort_by_count, reverse=True)
    counthtml=[]
    sanguo=[]
    li = []  # ['孔明'， 孔明， 孔明，孔明...., '曹操'。。。。。]
    for i in range(10):
        # 序列解包
        role, count = items[i]
        print(role, count)
        counthtml.append(count)
        sanguo.append(role)
   #5.绘图
  plt.pie(counthtml,shadow=True, labels=sanguo, autopct = '%1.1f%%')
  plt.legend(loc=2)
  plt.axis('equal')
  plt.show()

三国演义人物饼图