arrays into pandas dataframe columns

Question

I have a program that outputs arrays.

For example:

[[0, 1, 0], [0, 0, 0], [1, 3, 3], [2, 4, 4]]

I would like to turn these arrays into a dataframe using pandas. However, when I do the values become row values like this:

As you can see each array within the overall array becomes its own row. I would like each array within the overall array to become its own column with a column name.

Furthermore, in my use case, the number of arrays within the array is variable. There could be 4 arrays or 70 which means there could be 4 columns or 70. This is problematic when it comes to column names and I was wondering if there was anyway to auto increment column names in python.

Check out my attempt below and let me know how I can solve this.

My desired outcome is simply to make each array within the overall array into its own column instead of row and to have titles for the column that increment with each additional array/column.

Thank you so much.

Need help. Please respond!

frame = [[0, 1, 0], [0, 0, 0], [1, 3, 3], [2, 4, 4]]
numpy_data= np.array(frame)

df = pd.DataFrame(data=numpy_data, columns=["column1", "column2", "column3"])
print(frame)
print(df)

Grayrigel · Accepted Answer · 2020-09-26 23:46:33Z

A possible solution could be transposing and renaming the columns after transforming the numpy array into a dataframe. Here is the code:

import numpy as np
import pandas as pd

frame = [[0, 1, 0], [0, 0, 0], [1, 3, 3], [2, 4, 4]]
numpy_data= np.array(frame)

#transposing later
df = pd.DataFrame(data=numpy_data).T 

#creating a list of columns using list comprehension without specifying number of columns
df.columns = [f'mycol{i}' for i in range(0,len(df.T))] 

print(df)

Output:

   mycol0  mycol1  mycol2  mycol3
0       0       0       1       2
1       1       0       3       4
2       0       0       3       4

Same code for 11 columns:

import numpy as np
import pandas as pd

frame = [[0, 1, 0], [0, 0, 0], [1, 3, 3], [2, 4, 4], [5, 2, 2], [6,7,8], [8,9,19] , [10,2,4], [2,6,5], [10,2,5], [11,2,9]]
numpy_data= np.array(frame)

df = pd.DataFrame(data=numpy_data).T
df.columns = [f'mycol{i}' for i in range(0,len(df.T))]

print(df)

   mycol0  mycol1  mycol2  mycol3  mycol4  mycol5  mycol6  mycol7  mycol8  mycol9  mycol10
0       0       0       1       2       5       6       8      10       2      10       11
1       1       0       3       4       2       7       9       2       6       2        2
2       0       0       3       4       2       8      19       4       5       5        9

Michael Szczesny · Accepted Answer · 2020-09-26 20:20:19Z

3

You can transpose the array and add_prefix

frame = [[0, 1, 0], [0, 0, 0], [1, 3, 3], [2, 4, 4]]

pd.DataFrame(np.array(frame).T).add_prefix('column')

Out:

   column0  column1  column2  column3
0        0        0        1        2
1        1        0        3        4
2        0        0        3        4

Works with every number of arrays

frame = [[0, 1, 0], [0, 0, 0], [1, 3, 3], [2, 4, 4], [1,0,1], [2,0,3]]

pd.DataFrame(np.array(frame).T).add_prefix('column')

Out:

   column0  column1  column2  column3  column4  column5
0        0        0        1        2        1        2
1        1        0        3        4        0        0
2        0        0        3        4        1        3

answered Sep 26, 2020 at 20:20

Michael Szczesny

5,0465 gold badges20 silver badges36 bronze badges

Comments

niraj · Accepted Answer · 2020-09-26 20:38:21Z

1

One way may be change it to dictionary with column name by iterating each item in the list as below:

df = pd.DataFrame({'column{}'.format(index):i for index, i in enumerate(frame)})

Alternatively, other way may be to use transpose to what you already have. For column names you can exclude on creating dataframe and add later (not sure if you need numpy ):

df = pd.DataFrame(data=frame)


df = df.T # transposing
df.columns = ['column{}'.format(i+1) for i in df.columns] # adding column names

Result (either way):

    column1 column2 column3 column4
0   0          0      1        2
1   1          0      3        4
2   0          0      3        4

edited Sep 26, 2020 at 20:38

answered Sep 26, 2020 at 20:19

niraj

18.2k4 gold badges36 silver badges50 bronze badges

Comments

BENY · Accepted Answer · 2020-09-26 21:35:54Z

1

Let us try

pd.DataFrame(dict(zip(range(len(frame)), frame)))
   0  1  2  3
0  0  0  1  2
1  1  0  3  4
2  0  0  3  4

answered Sep 26, 2020 at 21:35

BENY

324k22 gold badges176 silver badges250 bronze badges

Collectives™ on Stack Overflow

arrays into pandas dataframe columns

4 Answers 4

Comments

Comments

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

4 Answers 4

Comments

Comments

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related