Numpy nd array to pandas column without []

Question

I'm trying to transform a numpy nd array into a pandas column, but the data is coming with brackets.

This is my np array:

array([[[  7.10105920e+07],
        [  9.18736320e+07],
        [  8.35562800e+07],
        [  7.16590640e+07],
        [  8.28060960e+07],
        [  6.77042000e+07],
        [  7.07195360e+07],
        [  1.04754616e+08],
        [  7.27420400e+07],
        [  7.33461760e+07],
        [  6.34156040e+07],
        [  8.00440800e+07],

This is how I'm sending to the dataframe:

predictions = pd.DataFrame()
predictions['y_test'] = Y_test[0].tolist()

This is what I'm getting:

           y_test
0    [71010592.0]
1    [91873632.0]
2    [83556280.0]
3    [71659064.0]
4    [82806096.0]
5    [67704200.0]
6    [70719536.0]
7   [104754616.0]
8    [72742040.0]
9    [73346176.0]

How can I remove the brackets ([])?

score 4 · Accepted Answer · 2017-08-14 16:19:08Z

4

It looks like a 3D array. You can pass its first element to the DataFrame constructor:

pd.DataFrame(Y_test[0], columns=['y_test'])
Out: 
         y_test
0    71010592.0
1    91873632.0
2    83556280.0
3    71659064.0
4    82806096.0
5    67704200.0
6    70719536.0
7   104754616.0
8    72742040.0
9    73346176.0
10   63415604.0
11   80044080.0

A better alternative from Divakar is to use squeeze:

pd.DataFrame(arr.squeeze(), columns=['y_test'])
Out: 
         y_test
0    71010592.0
1    91873632.0
2    83556280.0
3    71659064.0
4    82806096.0
5    67704200.0
6    70719536.0
7   104754616.0
8    72742040.0
9    73346176.0
10   63415604.0
11   80044080.0

edited Aug 14, 2017 at 16:19

answered Aug 14, 2017 at 16:12

user2285236

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

2 Comments

Divakar Over a year ago

Rather than being bothered to find out the non-singleton dim, I would just use squeeze.

user2285236 Over a year ago

@Divakar yeah much better. Thanks.

akuiper · Accepted Answer · 2017-08-14 16:12:17Z

2

It seems you have a 3d array, you could try:

predictions['y_test'] = Y_test[0,:,0]

predictions

#       y_test
#0  71010592.0
#1  91873632.0
#2  83556280.0
#3  71659064.0
#4  82806096.0
#5  67704200.0
#6  70719536.0
#7  104754616.0
#8  72742040.0
#9  73346176.0
#10 63415604.0
#11 80044080.0

answered Aug 14, 2017 at 16:12

akuiper

216k33 gold badges362 silver badges379 bronze badges

Comments

Zero · Accepted Answer · 2017-08-14 16:15:14Z

1

Couple of ways you could do

Option 1. Numpy indexing

predictions['y_test'] = Y_test[0,:,0]

Option 2. Flatten using list comprehension

predictions['y_test'] = [x[0] for x in Y_test[0]]

Option 3. Numpy flatten function

predictions['y_test'] = Y_test.flatten()

answered Aug 14, 2017 at 16:15

Zero

77.4k22 gold badges153 silver badges153 bronze badges

Comments

Trideep Rath · Accepted Answer · 2017-08-14 16:19:17Z

0

prediction = pd.DataFrame(Y_test.flatten(), columns=['y_test'])
prediction.head()

    y_test
    0     1.0
    1     1.0
    2     1.0
    3     1.0
    4     1.0

answered Aug 14, 2017 at 16:19

Trideep Rath

3,7431 gold badge27 silver badges14 bronze badges

Collectives™ on Stack Overflow

Numpy nd array to pandas column without []

4 Answers 4

2 Comments

Comments

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

4 Answers 4

2 Comments

Comments

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related