Creating a Pandas DataFrame with a single vector column from a 2-d numpy array

Question

I have a numpy 2D-array:

arr = [
  [10, 20],
  [30, 40]
]

Converting it into a pandas dataframe with pd.DataFrame(arr) gives me:

    0  1
0  10 20
1  30 40

I am looking for something like this:

          0
0  [10, 20]
1  [30, 40]

I can achieve it using

df.agg(lambda x: np.array(x), axis="columns")

or

df.agg(lambda x: [y for y in x], axis="columns")

But is there a better way to end up with the single column dataframe in the first place?

Don't ask side-questions. Instead, if you really want to know, ask a new SO question. Questions here can actually get closed for "Needs more focus: This question includes multiple questions in one. ..." ;) — user17242583
– user17242583, Commented Dec 20, 2021 at 22:06

score 4 · Accepted Answer · 2021-12-20 22:33:10Z

4

You can convert it to a list first, then a Series first, and finally a DataFrame:

df = pd.DataFrame(pd.Series(arr.tolist()))

df = pd.Series(arr.tolist()).to_frame()

Output:

>>> df
          0
0  [10, 20]
1  [30, 40]

answered Dec 20, 2021 at 22:02

user17242583

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

1 Comment

Maybe pd.Series(arr).to_frame() as well.