NumPy ravel()
Summary: in this tutorial, you’ll learn how to use the NumPy ravel() to return a contiguous flattened array.
Introduction to the NumPy ravel() function
The ravel() function accepts an array and returns a 1-D array containing the elements of the input array:
numpy.ravel(a, order='C')
In this syntax:
ais a numpy array. It can be any array-like object e.g., a list. An array-like object is an object that can be converted into a numpy array.orderspecifies the order of elements. Check out theflatten()method for detailed information on the order parameter and its values.
NumPy ravel() function example
Let’s take some examples of using the ravel() function.
1) Using NumPy ravel() function to flatten an array
The following example uses the ravel() function to flatten a 2-D array:
import numpy as np
a = np.array([[1, 2], [3, 4]])
b = np.ravel(a)
print(b)
Output:
[1 2 3 4]
How it works.
First, create a 2-D array:
a = np.array([[1, 2], [3, 4]])
Second, flatten the array using the ravel() function:
b = np.ravel(a)
Third, display the array:
print(b)
2) ravel() function vs. flatten() method
The flatten() method creates a copy of an input array while the ravel()ravel()
import numpy as np
a = np.array([[1, 2], [3, 4]])
b = np.ravel(a)
# change element at index 0
b[0] = 0
# show both b & a array
print(b)
print(a)
How it works.
First, use the ravel() function to create a view of the array a:
b = np.ravel(a)
Second, change the element from index 0 of the array b to zero:
b[0] = 0
Third, show both arrays a and b. Since array b is a view of array a, the change in array b is reflected in array a:
print(b)
print(a)
Another important difference between the flatten()ravel()flatten()ndarray while you can call the ravel()
Summary
- Use the numpy
ravel()function to return a contiguous flattened array.