Why is my .apply() function also outputting a series of 'None'?

Question

This may not be a big problem, I just haven't noticed this output of None before when doing .apply()

Toy example:

mydf = pd.DataFrame({'col1':['test1',np.nan,'test3','test4'],
                   'col2':['test5','test6','test7','test8']})

mydf


    col1   col2
0  test1  test5
1    NaN  test6
2  test3  test7
3  test4  test8

Function to just add the values together in a string:

def myfunc(row):


    thing = str(row['col1']) + str(row['col2'])

    print(thing)

Applying it:

mydf.apply(myfunc,axis=1)

My output:

test1test5
nantest6
test3test7
test4test8

0    None
1    None
2    None
3    None
dtype: object

Is it something to be worried about? I will be applying something like this to some real data shortly. I am doing this in Jupyter Notebook if it makes a difference.

you can replace `print(thing)` with `return thing` – anky Aug 15 '19 at 16:04 — anky, Aug 15 '19 at 16:04

score 3 · Accepted Answer · answered Aug 15 '19 at 16:05

3

You should return the string instead of printing it.

answered Aug 15 '19 at 16:05

Chris K

467
5
11

OK, I eventually plan to return something from the function. Was just testing it. Thanks – SCool Aug 15 '19 at 16:06
This happens :-) but sometime pointers are helpful when looking for it . – Karn Kumar Aug 15 '19 at 16:08

Karn Kumar · Answer 2 · 2019-08-15T16:13:19.067

user return not print before applying to df..

>>> mydf
    col1   col2
0  test1  test5
1    NaN  test6
2  test3  test7
3  test4  test8


>>> def myfunc(row):
...   thing = str(row['col1']) + str(row['col2'])
...   return thing
...

>>> mydf.apply(myfunc,axis=1)
0    test1test5
1      nantest6
2    test3test7
3    test4test8
dtype: object

Just an idea of doing it other way around ..

>>> cols = ['col1', 'col2']
>>> mydf[cols].astype(str).sum(axis=1)
0    test1test5
1      nantest6
2    test3test7
3    test4test8
dtype: object

Why is my .apply() function also outputting a series of 'None'?

2 Answers2

Linked

Related