pandas中concatenate和combine_first的用法详解

import numpy as np
import pandas as pd
from pandas import Series, DataFrame

# 设置一个随机种子，方便调试
np.random.seed(666)

# Series
arr1 = np.arange(9).reshape(3, 3)
arr2 = np.arange(9).reshape(3, 3)

# numpy的 concatenate 用法
print(np.concatenate([arr1, arr2]))
'''
[[0 1 2]
 [3 4 5]
 [6 7 8]
 [0 1 2]
 [3 4 5]
 [6 7 8]]
'''

print(np.concatenate([arr1, arr2], axis=1))
'''
[[0 1 2 0 1 2]
 [3 4 5 3 4 5]
 [6 7 8 6 7 8]]
'''


s1 = Series([1, 2, 3], index=['A', 'B', 'C'])
s2 = Series([4, 5], index=['E', 'F'])
# 可以看出和numpy的效果一样
print(pd.concat([s1, s2]))
'''
A    1
B    2
C    3
E    4
F    5
dtype: int64
'''
# 用法和 np 一样 axis = 1, 等于增加了一列
print(pd.concat([s1, s2], axis=1))
# 但是，返回的是一个 <class 'pandas.core.frame.DataFrame'>
print(type(pd.concat([s1, s2], axis=1)))
'''
     0    1
A  1.0  NaN
B  2.0  NaN
C  3.0  NaN
E  NaN  4.0
F  NaN  5.0
'''


df1 = DataFrame(np.random.randn(4, 3), columns=['X', 'Y', 'Z'])
print(df1)
'''
          X         Y         Z
0  0.824188  0.479966  1.173468
1  0.909048 -0.571721 -0.109497
2  0.019028 -0.943761  0.640573
3 -0.786443  0.608870 -0.931012
'''

df2 = DataFrame(np.random.randn(3, 3), columns=['X', 'Y', 'A'])
print(df2)
'''
          X         Y         A
0  0.978222 -0.736918 -0.298733
1 -0.460587 -1.088793 -0.575771
2 -1.682901  0.229185 -1.756625
'''

print(pd.concat([df1, df2]))
'''
          A         X         Y         Z
0       NaN  0.824188  0.479966  1.173468
1       NaN  0.909048 -0.571721 -0.109497
2       NaN  0.019028 -0.943761  0.640573
3       NaN -0.786443  0.608870 -0.931012
0 -0.298733  0.978222 -0.736918       NaN
1 -0.575771 -0.460587 -1.088793       NaN
2 -1.756625 -1.682901  0.229185       NaN
'''

# combine

s1 = Series([2, np.nan, 4, np.nan], index=['A', 'B', 'C', 'D'])
s2 = Series([1, 2, 3, 4], index=['A', 'B', 'C', 'D'])

# 用 s2 中的数值来填充 s1
print(s1.combine_first(s2))

'''
A    2.0
B    2.0
C    4.0
D    4.0
dtype: float64
'''

df1 = DataFrame({
    'X':[1, np.nan, 3, np.nan],
    'Y':[5, np.nan, 7, np.nan],
    'Z':[9, np.nan, 11, np.nan]
})

df2 = DataFrame({
    'Z':[np.nan, 10, np.nan, 12],
    'A':[1, 2, 3, 4]
})

# 功能同样是填充
print(df1.combine_first(df2))
'''
     A    X    Y     Z
0  1.0  1.0  5.0   9.0
1  2.0  NaN  NaN  10.0
2  3.0  3.0  7.0  11.0
3  4.0  NaN  NaN  12.0
'''