CODEWITHSUNDEEP
pythonpandas
subbu prasant
subbu prasant

Reputation: 11

DataFrame to dict with one column as the key and multiple columns as a value

      id       GIVEN_NAMES FAMILY_NAME     DATE_OF_BIRTH
0     tttyy12  8000004199  8000004199      19660410
1     tttyy13  8000004274  8000004274      19980209

This is a dataframe output

I need to convert this into dict

{'tttyy12': ['8000004199','8000004199','19660410'], 'tttyy13' : ['8000004274' ,'8000004274' ,     '19980209']}

Upvotes: 1

Views: 3770

Answers (3)

DJV
DJV

Reputation: 4863

I know it's an old question, however, another option is to use the pd.DataFrame.groupby().

dict([(key, value) for (key, value) in dt.groupby("id")])

Upvotes: 0

Ynjxsjmh
Ynjxsjmh

Reputation: 30002

Use apply() on rows and convert row value to list with Series.values.tolist().

l = df.set_index('id').apply(lambda row: {row.name: row.values.tolist()}, axis=1).tolist()

print(l)

[{'tttyy12': [8000004199, 8000004199, 19660410]}, {'tttyy13': [8000004274, 8000004274, 19980209]}]

Upvotes: 2

James
James

Reputation: 36598

You can set the index to the id column, then convert to an index-based dictionary. From there you can use a comprehension to turn the value dictionaries to lists.

{
    k: list(v.values()) 
    for k, v 
    in df.set_index('id').to_dict('index').items()
}
# returns:
{'tttyy12': [8000004199, 8000004199, 19660410],
 'tttyy13': [8000004274, 8000004274, 19980209]}

Upvotes: 0

Related Questions