Convert 2-D matrix into Dataframe in python

Question

a = np.array([[5, 6, 7, 8],[5, 6, 7, 8]])

df = pd.DataFrame(a, columns=['a'])

ValueError: Shape of passed values is (2, 4), indices imply (2, 1)

I hope that the final result is:

a
-----------
[5, 6, 7, 8]
[5, 6, 7, 8]

Edit:

    df = pd.DataFrame({"a": [a]})

    a
----------------------------------
0   [[5, 6, 7, 8], [5, 6, 7, 8]]

why?

If you need to collect all data under the column 'a' because they have a correlation, just use a tuple then — crissal, Nov 07 '19 at 04:09
a is document embedding, used as features for machine learning — Ivan Lee, Nov 07 '19 at 04:15

score 2 · Accepted Answer · answered Nov 07 '19 at 04:40

According to https://stackoverflow.com/a/18646275/5405298, you have to turn the array into a list. In your case, you can use

import pandas as pd
import numpy as np

a = np.array([[5, 6, 7, 8], [5, 6, 7, 8]])
df = pd.DataFrame({"a": a.tolist()})
print(df)

this returns:

              a
---------------
0  [5, 6, 7, 8]
1  [5, 6, 7, 8]

score 0 · Answer 2 · answered Nov 07 '19 at 04:00

0

Convert each sub-array to string, you will have what you want.

>>>a = np.array([[5, 6, 7, 8],[5, 6, 7, 8]])
>>>df = pd.DataFrame([",".join([str(i) for i in j]) for j in a], columns=['a'])
>>>df
         a
0  5,6,7,8
1  5,6,7,8

answered Nov 07 '19 at 04:00

Lê Tư Thành

1,063
2
10
19

if a column with list, not string? – Ivan Lee Nov 07 '19 at 04:03
@IvanLee, seem it doesn't work. `DataFrame` use `pandas_dtype()` to validating, and I don't see `list` in that. And in that case, `DataFrame` will detect the length of a list to require the number of columns also. – Lê Tư Thành Nov 07 '19 at 04:15

Convert 2-D matrix into Dataframe in python

2 Answers2