How to Use the NumPy flatten() Method in Python

09/09/2021

Contents

In this article, you will learn how to use the numpy flatten() method in Python.

NumPy flatten() Method

The flatten() method in NumPy is used to collapse a multi-dimensional NumPy array into a one-dimensional array.

Here’s how you can use it in Python:

import numpy as np

# Create a multi-dimensional array
arr = np.array([[1, 2, 3], [4, 5, 6]])

# Use the flatten method to collapse the array
flattened_arr = arr.flatten()

# Print the result
print(flattened_arr)

# Output
# [1 2 3 4 5 6]

Note that the flatten() method returns a flattened array as a copy. If you want to modify the original array in-place, you can use the ravel() method instead:

import numpy as np

# Create a multi-dimensional array
arr = np.array([[1, 2, 3], [4, 5, 6]])

# Use the ravel method to collapse the array
arr.ravel()

# Print the result
print(arr)

# Output
# [1 2 3 4 5 6]
 

The flatten() method has several optional parameters that you can use to control the behavior of the method:

order

Specifies the memory layout of the flattened array. The default is ‘C’ (row-major), but you can use ‘F’ for column-major order.

import numpy as np

# Create a multi-dimensional array
arr = np.array([[1, 2, 3], [4, 5, 6]])

# Use the flatten method to collapse the array with column-major order
flattened_arr = arr.flatten(order='F')

# Print the result
print(flattened_arr)

# Output
# [1 4 2 5 3 6]
copy

Specifies whether the flattened array should be a copy of the original data. The default is True, which means that the flattened array will be a copy of the original data. If you set copy to False, the flattened array will be a view of the original data and any changes made to the flattened array will be reflected in the original data.

import numpy as np

# Create a multi-dimensional array
arr = np.array([[1, 2, 3], [4, 5, 6]])

# Use the flatten method to collapse the array without creating a copy
flattened_arr = arr.flatten(copy=False)

# Print the result
print(flattened_arr)

# Output
# [1 2 3 4 5 6]

Note that in this case, the flattened_arr is a view of the original data, so any changes made to the flattened_arr will also be reflected in the original data.

#Python