numpy.random.shuffle() function can help us to permute a sequence randomly along the first axis , in this tutorial we will introduce how to use this function correctly.
Syntax
numpy.random.shuffle(x)
Modify the elements randomly along the first axis.
Parameter
x : array_like
Return
This function return None.
Here are some examples on how to use.
import numpy as np
x1 = np.array(range(9)).reshape(3, 3)
x1_seq = np.random.permutation(x1)
print("new x1 = ")
print(x1_seq)
print("x1 = ")
print(x1)The nex x1 and x1 is:
new x1 = [[3 4 5] [0 1 2] [6 7 8]] x1 = [[0 1 2] [3 4 5] [6 7 8]]
numpy.random.shuffle(x) is also can permute elements in x randomly, you can read this tutorial to knw how to do.
Understand numpy.random.permutation(): Randomly permute a sequence – Numpy Tutorial
The difference between numpy.random.permutation(x) and numpy.random.shuffle(x)
When x is an array like, both numpy.random.permutation(x) and numpy.random.shuffle(x) can permute the elements in x randomly along the first axis. However, numpy.random.permutation(x) will return a new varialbe and x is not change, numpy.random.shuffle(x) will change x and does not return a new variable.
Here is an example.
import numpy as np
#create a 4 * 4 matrix
x1 = np.array(range(16)).reshape(4,4)
x2 = np.array(range(16)).reshape(4,4)
print("x1 = ")
print(x1)
print("x2 = ")
print(x2)
Print x1 and x2, the result is:
x1 = [[ 0 1 2 3] [ 4 5 6 7] [ 8 9 10 11] [12 13 14 15]] x2 = [[ 0 1 2 3] [ 4 5 6 7] [ 8 9 10 11] [12 13 14 15]]
You will find x1 and x2 are the same.
Permute elements randomly along the first axis of x1 and x2.
x1_seq = np.random.permutation(x1) np.random.shuffle(x2)
Print new x1 and x1
print("new x1 = ")
print(x1_seq)
print("x1 = ")
print(x1)
The result is:
new x1 = [[ 4 5 6 7] [ 0 1 2 3] [12 13 14 15] [ 8 9 10 11]] x1 = [[ 0 1 2 3] [ 4 5 6 7] [ 8 9 10 11] [12 13 14 15]]
You will find x1 is not changed.
Print x2
print("x2 = ")
print(x2)
The result is:
x2 = [[ 4 5 6 7] [12 13 14 15] [ 0 1 2 3] [ 8 9 10 11]]
You will find x2 is changed.