python pandas: access the element of the list in DataFrame

Question

i have a problem with accessing list element in DataFrame:

import pandas as pd
from pandas import DataFrame

d=DataFrame({'pos': {0: [0.11,0.14,0.46], 1:[1,2,3]},'n': {0: 2.0,1:1}})

Column 'pos' contain a list. I need to calculate new column with 'n'-th element of the list in 'pos'-column In this case: 0.46, 2.

I think it would be smth like this:

d[u'new column']=d.pos.apply(lambda x: x[0])

but instead x[0] i need x[d.n].

I read manual and search the forum but havn't found anything. I fill it smth obvious, but i stucked. Help me, please.

Your code generates an empty dataframe, can you fix your dataframe construction code, thanks. — EdChum
– EdChum, Commented May 21, 2014 at 15:43
If I understand you correctly you better init you dataframe the following way d=DataFrame([{'pos': {0: [0.11,0.14,0.46]},'n': {0: 2.0}}, {'pos': {1: [0.1,0.4,0.6]},'n': {1: 1.0}}]) — Tom Ron
– Tom Ron, Commented May 21, 2014 at 15:49
Are you using python 2 or 3 as this will determine how to access the first value in the dict, also storin a dict with value as a list is painfully horrible — EdChum
– EdChum, Commented May 21, 2014 at 16:01
If using python 3 then using the dataframe construction code from Tom Ron then this works: def func(x): return next(iter(x.pos.values()))[int(next(iter(x.n.values())))] d['new column']=d.apply(lambda row: func(row), axis=1) — EdChum
– EdChum, Commented May 21, 2014 at 16:02

user3661347 · Accepted Answer · 2014-05-22 05:42:09Z

2

Thanks to all! This works fine:

def func(x): 
    return x.pos[int(x.n)]

d['new column']=d.apply(lambda row: func(row), axis=1)

answered May 22, 2014 at 5:42

user3661347

811 silver badge5 bronze badges

Sign up to request clarification or add additional context in comments.

1 Answer 1