Question

以下代码的目标是使用三个pandas数据帧实现FULL OUTER JOIN。应打印所有数据框的所有记录，如果两个或三个记录之间存在关系，则应将它们打印在同一行中。

用于关联数据帧的字段是第一个数据帧中的type_1和id_1，第二个数据帧中的type_2和id_2以及{{1第三个数据框中的{}和type_3。

问题是第二个和第三个数据帧之间的关系不起作用。看一下第11行和第13行的情况，它应该是一行，因为id_3 = type_2和type_3 = id_2。预期输出位于第11行id_3，不应打印第13行。如何解决这个问题？

11 NaN NaN NaN 7.0 8 KoKo 7.0 8 Kuku

结果：

import pandas as pd
raw_data = {
        'type_1': [0, 1, 1, 2, 2],
        'id_1': ['3', '4', '5', '3', '3'],
        'name_1': ['Alex', 'Amy', 'Allen', 'Peter', 'Liz']}
df_a = pd.DataFrame(raw_data, columns = ['type_1', 'id_1', 'name_1' ])

raw_datab = {
        'type_2': [1, 1, 1, 0,7],
        'id_2': ['4', '5', '5', '7', '8'],
        'name_2': ['Billy', 'Brian', 'Joe', 'Bryce', 'KoKo']}
df_b = pd.DataFrame(raw_datab, columns = ['type_2', 'id_2', 'name_2'])

raw_datac = {
        'type_3': [1, 1, 1, 1, 2, 2, 7],
        'id_3': ['4', '6', '5', '5', '3', '3','8'],
        'name_3': ['School', 'White', 'Jane', 'Homer', 'Paul', 'Lorel', 'Kuku']}
df_c = pd.DataFrame(raw_datac, columns = ['type_3', 'id_3', 'name_3'])

merged = df_a
merged = merged.merge(df_b, how='outer', left_on=['type_1', 'id_1'],
                      right_on=['type_2', 'id_2'])
merged = merged.merge(df_c, how='outer', left_on=['type_1', 'id_1'], 
                      right_on=['type_3', 'id_3'])

print(merged)

Answer 1

您需要在merge

之前创建合并密钥

df_a[['key1','key2']]=df_a[['type_1', 'id_1']]
df_b[['key1','key2']]=df_b[['type_2', 'id_2']]
df_c[['key1','key2']]=df_c[['type_3', 'id_3']]


merged = df_a
merged = merged.merge(df_b, how='outer')
merged = merged.merge(df_c, how='outer')
merged.drop(['key1','key2'],1)
Out[81]: 
    type_1 id_1 name_1  type_2 id_2 name_2  type_3 id_3  name_3
0      0.0    3   Alex     NaN  NaN    NaN     NaN  NaN     NaN
1      1.0    4    Amy     1.0    4  Billy     1.0    4  School
2      1.0    5  Allen     1.0    5  Brian     1.0    5    Jane
3      1.0    5  Allen     1.0    5  Brian     1.0    5   Homer
4      1.0    5  Allen     1.0    5    Joe     1.0    5    Jane
5      1.0    5  Allen     1.0    5    Joe     1.0    5   Homer
6      2.0    3  Peter     NaN  NaN    NaN     2.0    3    Paul
7      2.0    3  Peter     NaN  NaN    NaN     2.0    3   Lorel
8      2.0    3    Liz     NaN  NaN    NaN     2.0    3    Paul
9      2.0    3    Liz     NaN  NaN    NaN     2.0    3   Lorel
10     NaN  NaN    NaN     0.0    7  Bryce     NaN  NaN     NaN
11     NaN  NaN    NaN     7.0    8   KoKo     7.0    8    Kuku
12     NaN  NaN    NaN     NaN  NaN    NaN     1.0    6   White

外部连接Pandas中的三个数据帧不起作用

1 个答案: