#!/usr/bin/env python3 # -*- coding: utf-8 -*- """ Created on Mon Dec 27 14:45:00 2021 @author: patricks """ import requests from lxml import etree url = "https://www.bitpush.news/covid19/" headers = { "user-agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/95.0.4638.69 Safari/537.36" } response = requests.get(url, headers=headers) html = response.text #print(html) parse = etree.HTMLParser(encoding='utf-8') doc = etree.HTML(html) # states states = doc.xpath('//div[@class="table_container"]//tbody/tr/td/span/text()') print(states) person = doc.xpath('//div[@class="table_container"]//tbody/tr/td[2]/text()') print(person) person = [x.replace(",", "") for x in person] print(person) death = doc.xpath('//div[@class="table_container"]//tbody/tr/td[3]/text()') print(death) death = [x.replace(",", "") for x in death] message = list(zip(states, person, death)) import csv with open("content.csv", "w") as f: w = csv.writer(f) w.writerows(message) import pandas as pd #读取数据 df = pd.read_csv("content.csv", names=["states", "person", "death"]) df.head() print(df) for i in range(101): df.drop([i],inplace=True) print(df) df.sort_values(by=['person'],ascending=False,ignore_index=True) print(df) df=df.iloc[0:15] print(df) #作图 import matplotlib.pyplot as plt # 设置中文显示 plt.rcParams['font.sans-serif'] = ['SimHei'] plt.rcParams['figure.figsize'] = (10, 5) x = df["states"].values y = df['death'].values plt.bar(x, y) plt.xlabel("states") plt.ylabel("death") plt.show()