.ravel()

arisdelaCruz1413618857's avatar
arisdelaCruz1413618857's avatar
arisdelaCruz1413618857
Published May 30, 2025
Contribute to Docs

The .ravel() method returns a contiguous flattened array. It is used to convert a multi-dimensional array into a one-dimensional array. The returned array is a view of the original array whenever possible, meaning that it shares the same data buffer.

Syntax

ndarray.ravel(a, order='C')

Parameters:

  • a: The input array to be flattened.
  • order: This parameter specifies the order in which the array elements are read. It can take the following values:
    • 'C': C-style row-major order (default).
    • 'F': Fortran-style column-major order.
    • 'A': ‘F’ if input is Fortran contiguous in memory, ‘C’ otherwise.
    • 'K': As close to the memory layout as possible.

Return value:

A flattened 1D array. Returns a view whenever possible, otherwise a copy.

Example 1

In this example, a 2D array is flattened into a 1D array using .ravel(). The elements are listed in row-major order by default:

import numpy as np


arr = np.array([[1, 2, 3], [4, 5, 6]])
flattened_arr = np.ravel(arr)
print(flattened_arr)

This produces the following output:

[1 2 3 4 5 6]

Example 2

In this example, a 3D array is flattened into a 1D array using .ravel(). The elements are listed in row-major order by default:

import numpy as np


three_d_array = np.array([[[1, 2], [3, 4]], [[5, 6], [7, 8]]])
flattened_array = np.ravel(three_d_array)
print(flattened_array)

This produces the following output:

[1 2 3 4 5 6 7 8]

Codebyte Example

Run the following codebyte example to understand the usage of the ravel() method:

Code
Output
Loading...


All contributors
Contribute to Docs
Learn Python:NumPy on Codecademy