Join Dataframe with dictionary-CodePudding

I have a dataframe df as

Code	Designation
01	Example 1
2F	Example 3
40	Example 10

And I have dictionary as

matching = {
    'ABCD': ['01','60' ,'YZ'],
    'EFG': ['04','2F'],
    '007' : ['03','AB' ,'CD', 'DD'],
    ...
    }

I would like to add column on my df base on the matching dictionary as

Code	Designation	Custom 1
01	Example 1	ABCD
2F	Example 3	EFG
40	Example 10	null

The lists contain unique value in whole dictionary The matching can return 0 or 1 value

How should I proceed ?

Thank you

CodePudding user response：

Flip matching then map it over Code.

_matching_lookup = {e: k for k, v in matching.items() for e in v}
df['Custom 1'] = df['Code'].map(_matching_lookup)
df

   Code  Designation  Custom 1
0    01    Example 1      ABCD
1    2F    Example 3       EFG
2    40   Example 10       NaN

CodePudding user response：

As @wjandrea has indicated I would flip the matching dictionary, but if that isn't possible here is a way to accomplish the task leaving matching as it is

import numpy as np
def searchDict(val, dct):
    for ky in dct.keys():
        if val in dct[ky]:
            return ky
    return np.nan  

df['New_col'] = [searchDict(x, matching) for x in df.index]