比较具有不同格式列值的两个数据框

Question

I have two dataframes

df1:

AccountNo	name	a/ctype
11.22.21	Henry	checking
11.22.22	Sam	Saving.
11.22.23	John	Checking

df2:

AccountNo	name	a/ctype
11-22-21	Henry	checking
11-22-23	John	Checking
11-22-24	Rita	Checking

output: df3:

A/cNO_df1	A/cNO_df2	result.	Name_df1	Name_df2	result	a/ctype_df1	a/ctype_df2	result.
11.22.21	11-22-21	Match	Henry	Henry	Match.	checking	checking	Match.
11.22.22		Notindf2	Sam.		Notindf2	checking		Notindf2
11.22.23	11-22-23	Match	John.	john	Match.	checking	checking	Match.
.	11-22-24	Notindf1	.	Rita	Notindf1		checking	Notindf2

I tried removing the non numeric character for the accounts to compare both data set using: df1['AccountNo'] = df1.AccountNo.replace(regex=[r'\D+', value='') df2['AccountNo'] = df2.AccountNo.replace(regex=[r'\D+', value='')

And then concat two dataframes. But, When I remove the character I cannot print it in the same format and for ac not in df1 and ac not in df2 I am not able to concat that. I tried using numpy where to compare and concat.

Is there a way it can be done?

Answer 1

You can merge with an external Series as key:

df1.merge(df2, left_on='AccountNo', right_on=df2['AccountNo'].str.replace('-', '.'),
          suffixes=('_df1', '_df2'), how='outer')

output:

  AccountNo AccountNo_df1 name_df1 a/ctype_df1 AccountNo_df2 name_df2 a/ctype_df2
0  11.22.21      11.22.21    Henry    checking      11-22-21    Henry    checking
1  11.22.22      11.22.22      Sam      Saving           NaN      NaN         NaN
2  11.22.23      11.22.23     John    Checking      11-22-23     John    Checking
3  11.22.24           NaN      NaN         NaN      11-22-24     Rita    Checking