pandas: how to store a list in a dataframe?

Question

I want to set a cell value as a list. Say for example:

df.loc['a']['b'] = ['one', 'two', 'three']

However, I'm unable to do so as I get the following error:

ValueError: Must have equal len keys and value when setting with an iterable

My dataframe currently is just all zeros and is nxn. Is there any way to be able to set the cell value so that when I execute df.loc['a']['b'], I get back ['one', 'two', 'three'].

does `df.loc['a']['b'] = [['one', 'two', 'three']]` work? – EdChum Jun 30 '16 at 22:03 — EdChum, Jun 30 '16 at 22:03

score 28 · Accepted Answer · answered Jun 30 '16 at 22:28

28

The problem is that you likely have a dataframe where all the columns are series of type float or int. The solution is to change them type 'object.'

In [3]: df = pd.DataFrame(np.zeros((4,4)))

In [4]: df
Out[4]: 
     0    1    2    3
0  0.0  0.0  0.0  0.0
1  0.0  0.0  0.0  0.0
2  0.0  0.0  0.0  0.0
3  0.0  0.0  0.0  0.0

In [5]: df.dtypes
Out[5]: 
0    float64
1    float64
2    float64
3    float64
dtype: object

In [6]: df = df.astype('object')

In [7]: df[1][2] = [1,2,3]

In [8]: df
Out[8]: 
   0          1  2  3
0  0          0  0  0
1  0          0  0  0
2  0  [1, 2, 3]  0  0
3  0          0  0  0

answered Jun 30 '16 at 22:28

juanpa.arrivillaga

88,713
10
131
172

Couldnt you set the dtype on one column? – Merlin Jul 01 '16 at 01:12
@juanpa.arrivillaga Yes, I just need to do one colum as Merlin mentions – TheRealFakeNews Jul 03 '16 at 18:43

score 6 · Answer 2 · answered Jun 30 '16 at 22:10

6

This is a tortured way of doing it. I really hope someone has a better answer.

df.loc[['a'], ['b']] = df.loc[['a'], ['b']].applymap(lambda x: ['one', 'two', 'three'])

answered Jun 30 '16 at 22:10

piRSquared

285,575
57
475
624

Good solution. Keeping it on bookmark :) – c1377554 Nov 13 '19 at 14:37

pandas: how to store a list in a dataframe?

2 Answers2

Linked