Create a custom column in a dict format-CodePudding

I currently have a dataframe like this

Person	Analysis	Dexterity	Skills
174	3.76	4.12	1.20
239	4.10	3.78	3.77
557	5.00	2.00	4.40
674	2.23	2.40	2.80
122	3.33	4.80	4.10

I want to add an column to compile all this information like below

Person	Analysis	Dexterity	Skills	new_column
174	3.76	4.12	1.20	{"Analysis":"3.76", "Dexterity":"4.12", "Skills":"1.20"}
239	4.10	3.78	3.77	{"Analysis":"4.10", "Dexterity":"3.78", "Skills":"3.77"}
557	5.00	2.00	4.40	{"Analysis":"5.00", "Dexterity":"2.00", "Skills":"4.40"}
674	2.23	2.40	2.80	{"Analysis":"2.23", "Dexterity":"2.40", "Skills":"2.80"}
122	3.33	4.80	4.10	{"Analysis":"3.33", "Dexterity":"4.80", "Skills":"4.10"}

CodePudding user response：

You can use to to_dict method like so:

import pandas as pd

rows = [
    {'a': 1, 'b': 2},
    {'a': 3, 'b': 4},
]

df = pd.DataFrame(rows)

# define new column as the json format of another
# also convert to str as that is what you have in your output
df['c'] = df[['a', 'b']].astype(str).to_dict(orient='records')

to_dict has a really nice interface for transforming to JSON format or to other object types. In this instance you are looking for orient='records' which is a dictionary of keys.

In your case, you will use:

df['new_column'] = df[
    ['Analysis', 'Dexterity', 'Skills']
].astype(str).to_dict(orient='records')