printing sub-array in numpy as Matlab does

Question

How can you print sub-arrays in numpy the same way Matlab does? I have a 3 by 10000 array and I want to view the first 20 columns. In Matlab you can write

a=zeros(3,10000);
a(:,1:20)
  Columns 1 through 15

 0     0     0     0     0     0     0     0     0     0     0     0     0     0     0
 0     0     0     0     0     0     0     0     0     0     0     0     0     0     0
 0     0     0     0     0     0     0     0     0     0     0     0     0     0     0

  Columns 16 through 20

 0     0     0     0     0
 0     0     0     0     0
 0     0     0     0     0

However in Numpy

import numpy as np
set_printoptions(threshold=nan)
a=np.zeros((3,10000))
print a[:,0:20]
[[  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.]
 [  0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.   0.
    0.   0.   0.   0.   0.   0.   0.   0.]]

As you can see numpy prints the first row, then the second row, then the third row. I would like it to maintain the column structure and not the row structure

Thank you very much

PS: One solution would be for example

print a[:,0:20].T
[[  0.   0.   0.]
 [  0.   0.   0.]
 [  0.   0.   0.]
 [  0.   0.   0.]
 [  0.   0.   0.]
 [  0.   0.   0.]
 [  0.   0.   0.]
 [  0.   0.   0.]
 [  0.   0.   0.]
 [  0.   0.   0.]
 [  0.   0.   0.]
 [  0.   0.   0.]
 [  0.   0.   0.]
 [  0.   0.   0.]
 [  0.   0.   0.]
 [  0.   0.   0.]
 [  0.   0.   0.]
 [  0.   0.   0.]
 [  0.   0.   0.]
 [  0.   0.   0.]]

but, would consume a lot more of space on screen than desired. it would be great if numpy had this option

If your array has few rows but many columns, and you don't mind seeing it printed a bit screwy, you can print the transverse of the array with print a.T. — ari
– ari, Commented Sep 9, 2013 at 15:34
Also it seems that you must import numpy twice, once with from numpy import *; set_printoptions and nan are both from the numpy namespace. Are you using pylab mode in IPython by any chance? — ari
– ari, Commented Sep 9, 2013 at 15:38
Well, there are two more (janky) solutions: you can either use the linewidth option in set_printoptions to increase how much horizontal space each row gets (e.g. np.set_printoptions(linewidth=200) ), or you can try melding CT Zhu's answer with set_string_function to return the array as you want. — ari
– ari, Commented Sep 9, 2013 at 16:02
I am using iPython notebook, I was doing from numpy import stuff but here I tried to make the code minimal and forgot. :) linewidth option works fine and I want to accept the solution, but the problem is that since I am using iPython notebook the notebook limits the width of the block as well as the output. So even though I am happy with the solutions provided, in the future if I want to display long views of the array I would love if there is an easy way to have numpy breaking the line automatically — gota
– gota, Commented Sep 9, 2013 at 18:05

CT Zhu · Accepted Answer · 2013-09-09 15:07:20Z

1

Does this give what you want?

>>> for item in a[:,0:20].T:
    print '\t'.join(map(str,item.tolist()))

Or this?

>>> for item in a[:,0:20]:
    print '\t'.join(map(str,item.tolist()))

answered Sep 9, 2013 at 15:07

CT Zhu

54.6k18 gold badges125 silver badges136 bronze badges

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

1 Comment

gota Over a year ago

Thank you very much for your time in aswering the question, but... it kind of works, I was looking more for a way to specify how I want numpy to print 2-D arrays.

Collectives™ on Stack Overflow

printing sub-array in numpy as Matlab does

1 Answer 1

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related