Pandas数据清洗

import pandas as pd
print(50*'-', '空值清洗', 50*'-')
df1 = pd.read_csv('F:/python学习文件/实验文件夹/Pandas学习/property-data.csv')
print(30*'-', '源数据表', 30*'-')
print(df1)
missing_values = ["n/a", "na", "--"]  # 指定空值或者要清晰的值，PD默认n/a和NA为空数据，na不是空数据，要指定空数据类型
df2 = pd.read_csv('F:/python学习文件/实验文件夹/Pandas学习/property-data.csv', na_values=missing_values)
print(30*'-', '指定空值后原表中的全部置换为NaN', 30*'-')
print(df2)
print(30*'-', '源数据表删除包含空字符的行', 30*'-')
df3 = df1.dropna()
print(df3)
df1.dropna(subset=['ST_NUM'], inplace=True)
print(30*'-', '源数据表移除"ST_NUM"中为空的行', 30*'-')
print(df1)
df1['PID'].fillna(100005100, inplace=True)
print(30*'-', '源数据表"PID"列空值全部填充为100005100', 30*'-')
print(df1)
df1.fillna(0, inplace=True)
print(30*'-', '源数据表全部空值全部填充为0', 30*'-')
print(df1)
x = df2["ST_NUM"].mean()  # 当然可以用中位数median()或者众数mode()
df2["ST_NUM"].fillna(x, inplace=True)
print(30*'-', 'df2数据表"ST_NUM"列空值替换成其平均值', 30*'-')
print(df2, x)
print(50*'-', '日期数据格式清洗', 50*'-')
data = {
  "Date": ['2020/12/01', '2020/12/02', '20201226', '20201227', '20201228', '20201231'],
  "duration": [50, 40, 45, 25, 65, 89]
}
df4 = pd.DataFrame(data)
df4['Date'] = pd.to_datetime(df4['Date'])
print(df4)
print(50*'-', '异常值修改', 50*'-')
stu = {"name": ['Google', 'Runoob', 'Taobao'], "age": [50, 40, 12345]}
df5 = pd.DataFrame(stu)
for i in df5.index:
    if df5.loc[i, 'age'] >= 100:
        df5.loc[i, 'age'] = 80
print(df5)
print(50*'-', '删除重复值', 50*'-')
stu1 = {"name": ['Google', 'Google', 'Taobao'], "age": [50, 50, 12345]}
df6 = pd.DataFrame(stu1)
df6.drop_duplicates(inplace=True)
print(df6)