# 一样，数据处理就先给导入pandas先 import pandas as pd # df1==df2 df1 = pd.DataFrame({'一班':[90,80,66,75,99,55,76,78,98,None,90], '二班':[75,98,100,None,77,45,None,66,56,80,57], '三班':[45,89,77,67,65,100,None,75,64,88,99]}) df2 = pd.DataFrame({'一班':[90,80,66,75,99,55,76,78,98,None,90], '二班':[75,98,100,None,77,45,None,66,56,80,57], '三班':[45,89,77,67,65,100,None,75,64,88,99]})
1数据堆叠
数据堆叠分为以下两种：
-
- 行堆叠
- 列堆叠
pd.concat(objs, axis=0)
- objs：参与合并的多个DataFrame。无默认
- axis：表示轴向，axis=0表示行合并，axis=1表示列合并
pd.concat([df1, df2, df3], axis=1)
| 一班 | 三班 | 二班 | 一班 | 三班 | 二班 | 一班 | 三班 | 二班 | |
|---|---|---|---|---|---|---|---|---|---|
| 0 | 90.0 | 45.0 | 75.0 | 90.0 | 45.0 | 75.0 | 90.0 | 45.0 | 75.0 |
| 1 | 80.0 | 89.0 | 98.0 | 80.0 | 89.0 | 98.0 | 80.0 | 89.0 | 98.0 |
| 2 | 66.0 | 77.0 | 100.0 | 66.0 | 77.0 | 100.0 | 66.0 | 77.0 | 100.0 |
| 3 | 75.0 | 67.0 | NaN | 75.0 | 67.0 | NaN | 75.0 | 67.0 | NaN |
| 4 | 99.0 | 65.0 | 77.0 | 99.0 | 65.0 | 77.0 | 99.0 | 65.0 | 77.0 |
| 5 | 55.0 | 100.0 | 45.0 | 55.0 | 100.0 | 45.0 | 55.0 | 100.0 | 45.0 |
| 6 | 76.0 | NaN | NaN | 76.0 | NaN | NaN | 76.0 | NaN | NaN |
| 7 | 78.0 | 75.0 | 66.0 | 78.0 | 75.0 | 66.0 | 78.0 | 75.0 | 66.0 |
| 8 | 98.0 | 64.0 | 56.0 | 98.0 | 64.0 | 56.0 | 98.0 | 64.0 | 56.0 |
| 9 | NaN | 88.0 | 80.0 | NaN | 88.0 | 80.0 | NaN | 88.0 | 80.0 |
| 10 | 90.0 | 99.0 | 57.0 | 90.0 | 99.0 | 57.0 | 90.0 | 99.0 | 57.0 |
当然，如果axis=0（行堆叠）时，也可以使用append函数
# append 直接在末尾追加，注意特征数目相同，并且数据类型相同 df1.append(df2)
| 一班 | 三班 | 二班 | |
|---|---|---|---|
| 0 | 90.0 | 45.0 | 75.0 |
| 1 | 80.0 | 89.0 | 98.0 |
| 2 | 66.0 | 77.0 | 100.0 |
| 3 | 75.0 | 67.0 | NaN |
| 4 | 99.0 | 65.0 | 77.0 |
| 5 | 55.0 | 100.0 | 45.0 |
| 6 | 76.0 | NaN | NaN |
| 7 | 78.0 | 75.0 | 66.0 |
| 8 | 98.0 | 64.0 | 56.0 |
| 9 | NaN | 88.0 | 80.0 |
| 10 | 90.0 | 99.0 | 57.0 |
| 0 | 90.0 | 45.0 | 75.0 |
| 1 | 80.0 | 89.0 | 98.0 |
| 2 | 66.0 | 77.0 | 100.0 |
| 3 | 75.0 | 67.0 | NaN |
| 4 | 99.0 | 65.0 | 77.0 |
| 5 | 55.0 | 100.0 | 45.0 |
| 6 | 76.0 | NaN | NaN |
| 7 | 78.0 | 75.0 | 66.0 |
| 8 | 98.0 | 64.0 | 56.0 |
| 9 | NaN | 88.0 | 80.0 |
| 10 | 90.0 | 99.0 | 57.0 |
2主键合并
主键合并大概是应用最关的合并方式了，也是我最喜欢的方式。
pd.merge(left, right, how='inner', on=None, left_on=None, right_on=None, suffixes=('_x', '_y'))
- left：表示进行合并的左边的DataFrame。无默认。
- right：表示进行合并的右边的DataFrame。无默认。
- how：表示合并的方法。默认为'inner'。可取'left'(左连接),'right'（右连接）,'inner'（内连接）,'outer'（外连接）。
- on：表示合并的主键。默认为空。
- left_on：表示左边的合并主键。默认为空。
- right_on：表示右边的合并主键。默认为空。
- suffixes：表示列名相同的时候的后缀。默认为('_x', '_y')
# 合并数据 pd.merge(df1, df2, on='一班')
| 一班 | 三班_x | 二班_x | 三班_y | 二班_y | |
|---|---|---|---|---|---|
| 0 | 90.0 | 45.0 | 75.0 | 45.0 | 75.0 |
| 1 | 90.0 | 45.0 | 75.0 | 99.0 | 57.0 |
| 2 | 90.0 | 99.0 | 57.0 | 45.0 | 75.0 |
| 3 | 90.0 | 99.0 | 57.0 | 99.0 | 57.0 |
| 4 | 80.0 | 89.0 | 98.0 | 89.0 | 98.0 |
| 5 | 66.0 | 77.0 | 100.0 | 77.0 | 100.0 |
| 6 | 75.0 | 67.0 | NaN | 67.0 | NaN |
| 7 | 99.0 | 65.0 | 77.0 | 65.0 | 77.0 |
| 8 | 55.0 | 100.0 | 45.0 | 100.0 | 45.0 |
| 9 | 76.0 | NaN | NaN | NaN | NaN |
| 10 | 78.0 | 75.0 | 66.0 | 75.0 | 66.0 |
| 11 | 98.0 | 64.0 | 56.0 | 64.0 | 56.0 |
| 12 | NaN | 88.0 | 80.0 | 88.0 | 80.0 |
pd.merge(df1, df2, left_on='一班', right_on='二班', suffixes=('_1', '_2))
| 一班_1 | 三班_1 | 二班_1 | 一班_2 | 三班_2 | 二班_2 | |
|---|---|---|---|---|---|---|
| 0 | 80.0 | 89.0 | 98.0 | NaN | 88.0 | 80.0 |
| 1 | 66.0 | 77.0 | 100.0 | 78.0 | 75.0 | 66.0 |
| 2 | 75.0 | 67.0 | NaN | 90.0 | 45.0 | 75.0 |
| 3 | 98.0 | 64.0 | 56.0 | 80.0 | 89.0 | 98.0 |
| 4 | NaN | 88.0 | 80.0 | 75.0 | 67.0 | NaN |
| 5 | NaN | 88.0 | 80.0 | 76.0 | NaN | NaN |
3重叠合并
不是特别建议，毕竟重叠合并没什么依据，而且浪费数据资源。
DataFrame.combine_first(other) 重叠合并，当两者皆有以前者为准，为空时，则使用后者的补上。
df1['一班'].combine_first(df1['二班'])
0 90.0 1 80.0 2 66.0 3 75.0 4 99.0 5 55.0 6 76.0 7 78.0 8 98.0 9 80.0 10 90.0 Name: 一班, dtype: float64