Extract element from Pandas dataframe column and create new column

Question

I have data like:

df =
id1, id2, string
1,   [2], 'foo'
2,   [3], 'bar'
3,   [4], 'baz'

I'd like to replace id2 with the value from the array.

Here is what I've tried:

x = df['id2'].map(lambda x : x[0])

With the hopes of x being a series with the values I want that I can column bind to my DF. What actually happens is that it errors out with an IndexError. In that case I made a function to apply, in place of the lambda function, so that it could except the error, but this returned all nulls.

Seems like this should be straight forward, but I haven't been able to figure it out even after sleeping on it.

Answer 1

This works for if 'id2' is stored as array. You may just need to call tolist() to get the output you desire.

x = df['id2'].map(lambda x: x[0]).tolist()

Alternatively if 'id2' is stored as a string, you can use the ast package to call literal_eval to consume as an array:

import pandas as pd
from io import StringIO

data = StringIO(u'''id1,id2,string
1,[2],foo
2,[3],bar
3,[4],baz''')

df = pd.read_csv(data)

import ast
x = df['id2'].map(lambda x: ast.literal_eval(x)[0]).tolist()