CODEWITHSUNDEEP
pythondictionarypandas
user2872701
user2872701

Reputation: 171

python pandas convert dataframe to dictionary with multiple values

I have a dataframe with 2 columns Address and ID. I want to merge IDs with the same addresses in a dictionary

import pandas as pd, numpy as np

df = pd.DataFrame({'Address' : ['12 A', '66 C', '10 B', '10 B', '12 A', '12 A'],
                'ID' : ['Aa', 'Bb', 'Cc', 'Dd', 'Ee', 'Ff']})
AS=df.set_index('Address')['ID'].to_dict()

print df

  Address  ID
0    12 A  Aa
1    66 C  Bb
2    10 B  Cc
3    10 B  Dd
4    12 A  Ee
5    12 A  Ff

print AS

{'66 C': 'Bb', '12 A': 'Ff', '10 B': 'Dd'}

What I want is for the duplicates to store multiple values like:

{'66 C': ['Bb'], '12 A': ['Aa','Ee','Ff'], '10 B': ['Cc','Dd']}

Upvotes: 11

Views: 8105

Answers (2)

lababidi
lababidi

Reputation: 2752

In response to the comment about multiple columns:

>>> df
  Address  ID  Name
0    12 A  Aa  Alpha
1    66 C  Bb  Bravo
2    10 B  Cc  Charlie
3    10 B  Dd  Delta
4    12 A  Ee  Edgar
5    12 A  Ff  Frank
>>> {k: v.to_dict() for k,v in df.groupby("Address")}

Upvotes: 1

DSM
DSM

Reputation: 353019

I think you can use groupby and a dictionary comprehension here:

>>> df
  Address  ID
0    12 A  Aa
1    66 C  Bb
2    10 B  Cc
3    10 B  Dd
4    12 A  Ee
5    12 A  Ff
>>> {k: list(v) for k,v in df.groupby("Address")["ID"]}
{'66 C': ['Bb'], '12 A': ['Aa', 'Ee', 'Ff'], '10 B': ['Cc', 'Dd']}

Upvotes: 19

Related Questions