中国大学排名爬取
import requests
from bs4 import BeautifulSoup as bs
import pandas as pd
from matplotlib import pyplot as plt
def get_rank(url):
count = 0
rank = []
headers = {
"user-agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/101.0.4951.54 Safari/537.36 Edg/101.0.1210.3"
}
resp = requests.get(url, headers=headers).content.decode()
soup = bs(resp, "lxml")
univname = soup.find_all('a', class_="name-cn")
for i in univname:
if count != 10:
university = i.text.replace(" ", "")
score = soup.select("#content-box > div.rk-table-box > table > tbody > tr:nth-child({}) > td:nth-child(5)"
.format(count + 1))[0].text.strip()
rank.append([university, score])
else:
break
count += 1
return rank
total = []
u_year = 2015
for i in range(15, 20):
url = "https://www.shanghairanking.cn/rankings/bcur/20{}11".format(i)
print(url)
title = ['学校名称', '总分']
df = pd.DataFrame(get_rank(url), columns=title)
total.append(df)
for i in total:
plt.rcParams['font.sans-serif'] = ['SimHei'] # 用来正常显示中文标签
x = list(i["学校名称"])[::-1]
y = list(i["总分"])[::-1]
# 1.创建画布
plt.figure(figsize=(20, 8), dpi=100)
# 2.绘制图像
plt.plot(x, y, label="大学排名")
# 2.2 添加网格显示
plt.grid(True, linestyle="--", alpha=0.5)
# 2.3 添加描述信息
plt.xlabel("大学名称")
plt.ylabel("总分")
plt.title(str(u_year) + "年软科中国最好大学排名Top10", fontsize=20)
# 2.5 添加图例
plt.legend(loc="best")
# 3.图像显示
plt.savefig(str(u_year)+".png")
plt.show()
u_year += 1
while True:
info = input("请输入要查询的大学名称和年份：")
count = 0
university, year = info.split()
year = int(year)
judge = 2019 - year
tmp = total[::-1]
if 4 >= judge >= 0:
name = list(total[judge - 1]["学校名称"])
for j in name:
if university == j:
print(university + "在{0}年排名第{1}".format(year, count + 1))
break
count += 1
if count ==10:
print("很抱歉，没有该学校的排名记录！！！")
print("请选择以下选项:")
print(" 1.继续查询")
print(" 2.结束查询")
select = int(input(""))
if select == 1:
continue
elif select == 2:
break
else:
break
else:
print("很抱歉，没有该年份的排名记录！！！")
print("请选择以下选项:")
print(" 1.继续查询")
print(" 2.结束查询")
select = int(input(""))
if select == 1:
continue
elif select == 2:
break
标签：count,plt,4.17,break,year,print,select From： https://www.cnblogs.com/Christmas77/p/18236169