数据清洗-地区

导包

import pandas as pd
import re
from matplotlib import pyplot as plt
font = {
   
	"family":"Microsoft Yahei"
}
plt.rc("font",**font)

数据读取

data01 = pd.read_csv("data01.csv")
data02 = pd.read_csv("data02.csv")

表拼接

data02 = data02.iloc[:,1:]
df = pd.concat([data01,data02],axis=0)

去重

df.drop_duplicates(inplace=True)

获取省

def get_province(x):
	if "省" not in x:
		return "未识别"
	else:
		return x.split("省")[0]+"省"

df["省份"] = df["工作地"].apply(get_province)

获取市

def get_city(x):
	# 市京市六合区(特殊情况增加进函数里)
	if "市京市" in x:
		return "市京市"
	elif "市" in x:
		if "省" in x:
			return x.split("省")[1].split("市")[0]+"市"
		else:
			return x.split("市")[