Python数据预处理简单笔记持续补充

import pandas as pd
import numpy as np

df1 = pd.DataFrame({
    "id": ["a1", "a2", np.nan, "a4"],
    "age": [20, 23, np.nan, 10],
    "sex": ["男", np.nan, np.nan, np.nan],
    "time": ["0101", "0202", np.nan, np.nan]
})
print(df1)
print(df1.dropna(how="all"))  # 删除所有为空的一行
print(df1.fillna(0))  # 空值填充0
print(df1.fillna({"id": "A", "age": 0, "sex": "-", "time": 0000}))  # 比较智能 用字典方式替换需要替换的

df2 = pd.DataFrame({
    "id": ["a1", "a2", "a3", "a4", "a4", "a4"],
    "age": [20, 23, 26, 10, 10, 10],
    "name": ["男", "女", "女", "男", "男", "男"],
    "time": ["0101", "0202", "0305", "0506", "0506", "0506"]
})
print(df2)
print(df2.drop_duplicates())  # 删除相同数据 但保留第一个
print(df2["time"].dtype)  # 检测数据类型
df2["time"] = df2["time"].astype("string")  # 更改标题下的数据类型
print(df2["time"].dtype)