Gavin Wiggins

About  /  Notes  /  Photos

Save and load arrays

Written on November 7, 2022

NumPy arrays can be saved and loaded using different file formats. Examples of using the NumPy file formats .npy and .npz as well as a plain text format are given below.

Using the npy format

Use the NumPy save() function to save an array to a .npy file. Use the load() function to load the array from the file.

import numpy as np

# Save an array named `z` to an `.npy` file

z = np.array([[3, 4, 8.91], [1.05, 2, 7], [5.4, 3, 1]])

with open('zarray.npy', 'wb') as file:
    np.save(file, z)

# Load the array data from the `.npy` file into `zz`

with open('zarray.npy', 'rb') as file:
    zz = np.load(file)

# Print results

print('z is\n', z)
print('zz is\n', zz)

Output from the above example is shown below.

z is
 [[3.   4.   8.91]
  [1.05 2.   7.  ]
  [5.4  3.   1.  ]]

zz is
 [[3.   4.   8.91]
  [1.05 2.   7.  ]
  [5.4  3.   1.  ]]

Using the npz format

Use the savez() function to save several arrays to a single .npz file. Use the load() function to load the saved arrays from the file.

import numpy as np

# Save arrays `a`, `b`, `c` to an `.npz` file

a = np.array([[4, 5, 81], [10, 2, 7], [1, 21, 5]])
b = np.array([[90, 51, 81], [10, 21, 74], [19, 1, 15]])
c = np.array([[0.1, 5.8, 0.71], [3.9, 2, 7.9], [1.05, 21, 5]])

with open('zdata.npz', 'wb') as file:
    np.savez(file, a=a, b=b, c=c)

# Load arrays `aa`, `bb`, `cc` from the `.npz` file

with np.load('zdata.npz') as data:
    aa = data['a']
    bb = data['b']
    cc = data['c']

# Print results

print('a is\n', a)
print('b is\n', b)
print('c is\n', c)

print('aa is\n', aa)
print('bb is\n', bb)
print('cc is\n', cc)

Output from this example is given below.

a is
 [[ 4  5 81]
  [10  2  7]
  [ 1 21  5]]
b is
 [[90 51 81]
  [10 21 74]
  [19  1 15]]
c is
 [[ 0.1   5.8   0.71]
  [ 3.9   2.    7.9 ]
  [ 1.05 21.    5.  ]]

aa is
 [[ 4  5 81]
  [10  2  7]
  [ 1 21  5]]
bb is
 [[90 51 81]
  [10 21 74]
  [19  1 15]]
cc is
 [[ 0.1   5.8   0.71]
  [ 3.9   2.    7.9 ]
  [ 1.05 21.    5.  ]]

Using plain text format

A plain text format can also be used to save and load a NumPy array with the savetxt() and loadtxt() functions.

import numpy as np

# Save array to a text file

a = np.array([[1, 2, 3], [4, 5, 6], [7, 8, 9]])

np.savetxt('array.txt', a)

# Load array from a text file named `array.txt`

b = np.loadtxt('array.txt')

# Print comparison of `a` and `b`

print('a is\n', a)
print('b is\n', b)

Gavin Wiggins © 2024.
Made on a Mac with Genja. Hosted on GitHub Pages.