How to Print Full NumPy Arrays Without Truncation?

Printing Full NumPy Arrays

When working with NumPy arrays, you may encounter truncated representations when printing them. This can be frustrating if you need to view the complete array for debugging or analysis purposes. To address this issue, you can utilize numpy.set_printoptions.

numpy.set_printoptions allows you to configure various printing options for NumPy arrays. By setting threshold to the maximum value of sys.maxsize, you can increase the threshold at which NumPy prints an abbreviated representation of the array.

Here's how to use it:

import sys
import numpy

numpy.set_printoptions(threshold=sys.maxsize)

This will set the threshold to the maximum possible value, ensuring that the complete NumPy array is printed, regardless of its size.

For example, if you have a large array of shape (250, 40) as shown below:

>>> numpy.arange(10000).reshape(250, 40)

The default printing will truncate the array:

array([[   0,    1,    2, ...,   37,   38,   39],
       [  40,   41,   42, ...,   77,   78,   79],
       [  80,   81,   82, ...,  117,  118,  119],
       ..., 
       [9880, 9881, 9882, ..., 9917, 9918, 9919],
       [9920, 9921, 9922, ..., 9957, 9958, 9959],
       [9960, 9961, 9962, ..., 9997, 9998, 9999]])

However, using numpy.set_printoptions, you can print the entire array:

>>> numpy.set_printoptions(threshold=sys.maxsize)
>>> numpy.arange(10000).reshape(250, 40)

[[  0   1   2 ... 37 38 39]
 [ 40 41 42 ... 77 78 79]
 [ 80 81 82 ...117 118 119]
 ...
 [9880 9881 9882 ...9917 9918 9919]
 [9920 9921 9922 ...9957 9958 9959]
 [9960 9961 9962 ...9997 9998 9999]]

By adjusting the threshold parameter, you can control how NumPy prints arrays. This allows you to balance readability with the ability to view small or large arrays in their entirety.

The above is the detailed content of How to Print Full NumPy Arrays Without Truncation?. For more information, please follow other related articles on the PHP Chinese website!