Question

我有一个包含名称的Pandas DataFrame1。

column1:    column2:
John        some_value
Steve       some_value
Mark        some_value

另一个包含全名的DataFrame2。

column1:        column2:
John Smith      some_value
Steve James     some_value
Mark Taylor     some_value

我需要使等同于SQL的marge：

select
     df1.column1
    ,df2.column2
from DataFrame1 df1
join DataFrame2 df2
  on df1.column1 like '%' + df2.column1 + '%'

任何帮助都会很好。

Answer 1

import pandas as pd
inputdataframe1 = [['John', 4],['Steve', 5],['Mark', 6]]
inputdataframe2= [['John smith', 9],['Steve James', 8],['Mark Taylor', 4]]
dataframe1 = pd.DataFrame(inputdataframe1)
dataframe2= pd.DataFrame(inputdataframe2)
merged_dataframe = pd.merge(dataframe1, dataframe2, left_on=[0],right_on=[0],how='outer')

输出将是这样，因为我们无法直接合并

        0   1_x     1_y
0   John    4.0     NaN
1   Steve   5.0     NaN
2   Mark    6.0     NaN
3   John smith  NaN 9.0
4   Steve James NaN 8.0
5   Mark Taylor NaN 4.0

如果您需要使用下面的代码 like 连接两个数据框，则将为您提供帮助

import pandas as pd
inputdataframe1 = [['John', 4],['Steve', 5],['Mark', 6]]
inputdataframe2= [['John smith', 9],['Steve James', 8],['Mark Taylor', 4]]
dataframe1 = pd.DataFrame(inputdataframe1)
dataframe2= pd.DataFrame(inputdataframe2)
dataframe1_names=[key for key,value in inputdataframe1]
dataframe2_names=[key for key,value in inputdataframe2]
d=dict(inputdataframe1)
list_like_values=[[dataframe2_names[j],d[dataframe1_names[i]]] for i in range(len(dataframe2_names)) for j  in range(len(dataframe1_names)) if(dataframe1_names[i] in dataframe2_names[j])]
dataframe1= pd.DataFrame(list_like_values)
merged_dataframe = pd.merge(dataframe1, dataframe2, left_on=[0],right_on=[0],how='inner')

输出将采用以下格式

    0   1_x     1_y
0   John smith  4   9
1   Steve James 5   8
2   Mark Taylor 6   4

Python合并等效于使用like运算符的SQL连接

1 个答案: