Convert numpy array to a string using numpy.array_str()

Question

I have a 2 dimensional numpy array. I am trying to make pairs of each row with every other row.

For example:

1 2 3 4 5
6 7 8 9 10
11 12 13 14 15

It should have the following pairs(comma separated):

Pair(1,2) 1 2 3 4 5, 6 7 8 9 10
Pair(1,3) 1 2 3 4 5, 11 12 13 14 15
Pair(2,1) 6 7 8 9 10, 1 2 3 4 5
Pair(2,3) 6 7 8 9 10, 11 12 13 14 15
Pair(3,1) 11 12 13 14 15, 1 2 3 4 5
Pair(3,2) 11 12 13 14 15, 6 7 8 9 10

To achieve this task, I am converting each row to string then removing multiple white spaces, end line chars and [ ] chars.

But when I use numpy.array_str() for converting a large array (1500*100) it converts the row to:

1 2 ...., 4 5, 6 7 ....,8 9

Why is this happening and how can I fix this problem such that it doesn't slows down my program? If there is no possible way to do this without compromising speed then how can I achieve this task?

Hadrián · Accepted Answer · 2015-11-09 22:08:31Z

0

Maybe then a posible answer is:

import numpy as np
def pair(matrix,i,j):
    str1=''
    for value in matrix[i]:
        str1+=str(value)+' '
    # Delete one extra space
    str1=str1[:-1]

    str2=''
    for value in matrix[j]:
        str2+=str(value)+' '
    str2=str2[:-1]
    return str1+', '+str2

matrix=np.array([[1,2,3,4,5],[6,7,8,9,10],[11,12,13,14,15]])

print pair(matrix,0,1)
print pair(matrix,0,2)

edited Nov 9, 2015 at 22:08

answered Nov 9, 2015 at 21:47

Hadrián

3541 gold badge2 silver badges10 bronze badges

pair(matrix, 0,1) will give: [ 1 2 3 4 5 6 7 8 9 10]. I want [ 1 2 3 4 5, 6 7 8 9 10] with or without brackets.

Black Dragon
– Black Dragon

11/09/2015 21:51:48
Commented Nov 9, 2015 at 21:51
@BlackDragon So you need it to be a string? if not you cannot put a comma there

Hadrián
– Hadrián

11/09/2015 21:53:49
Commented Nov 9, 2015 at 21:53
Yes. I need it to be a string as that's the only possible way I found in doing that but the problem is that numpy.array_str() for a large row/col will give result like this: [1 2 ..., 4 5, 6 7 ..., 9 10] . The ellipses(...) are appearing in the result.

Black Dragon
– Black Dragon

11/09/2015 22:02:47
Commented Nov 9, 2015 at 22:02

Add a comment |

hpaulj · Accepted Answer · 2015-11-09 23:00:52Z

When the array gets big enough the default print starts using ellipsis

In [351]: np.array_str(np.ones((15,100)))
Out[351]: '[[ 1.  1.  1. ...,  1.  1.  1.]\n [ 1.  1.  1. ...,  1.  1.  1.]\n [ 1.  1.  1. ...,  1.  1.  1.]\n ..., \n [ 1.  1.  1. ...,  1.  1.  1.]\n [ 1.  1.  1. ...,  1.  1.  1.]\n [ 1.  1.  1. ...,  1.  1.  1.]]'

In [354]: np.ones((15,100))
Out[354]: 
array([[ 1.,  1.,  1., ...,  1.,  1.,  1.],
       [ 1.,  1.,  1., ...,  1.,  1.,  1.],
       [ 1.,  1.,  1., ...,  1.,  1.,  1.],
       ..., 
       [ 1.,  1.,  1., ...,  1.,  1.,  1.],
       [ 1.,  1.,  1., ...,  1.,  1.,  1.],
       [ 1.,  1.,  1., ...,  1.,  1.,  1.]])

In [355]: np.get_printoptions()
Out[355]: 
{'threshold': 1000,
 'infstr': 'inf',
 'suppress': False,
 'linewidth': 75,
 'formatter': None,
 'nanstr': 'nan',
 'edgeitems': 3,
 'precision': 8}

printoptions lets you tweak the formatting. I think it's the threshold that triggers the ....

np.ones((11,100)) gets the ..., np.ones((10,100)) does not.

np.set_printoptions(threshold=3000)
np.ones((21,100))   # no ...

Collectives™ on Stack Overflow

Convert numpy array to a string using numpy.array_str()

2 Answers 2

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Your Answer

Sign up or log in

Post as a guest

Related