Shuffling arrays in NumPy

Introduction

Shuffling arrays means mixing up the order of elements randomly. It helps when you want to avoid patterns or biases in data.

When you want to randomize the order of data before training a machine learning model.

When you need to create random samples from a dataset without replacement.

When you want to mix up a list of items to ensure fairness, like shuffling a deck of cards.

When testing algorithms that require random input order to check stability.

When splitting data into random groups or batches.

Syntax

NumPy

import numpy as np

# Create an array
array = np.array([1, 2, 3, 4, 5])

# Shuffle the array in-place
np.random.shuffle(array)

np.random.shuffle() changes the original array order directly (in-place).

It works only on the first axis for multi-dimensional arrays.

Examples

Shuffling an empty array keeps it empty without error.

NumPy

import numpy as np

empty_array = np.array([])
np.random.shuffle(empty_array)
print(empty_array)

Shuffling an array with one element keeps it the same.

NumPy

import numpy as np

single_element_array = np.array([42])
np.random.shuffle(single_element_array)
print(single_element_array)

Shuffling a normal array changes the order randomly.

NumPy

import numpy as np

array = np.array([10, 20, 30, 40, 50])
np.random.shuffle(array)
print(array)

For 2D arrays, shuffle only mixes rows, not elements inside rows.

NumPy

import numpy as np

multi_dim_array = np.array([[1, 2], [3, 4], [5, 6]])
np.random.shuffle(multi_dim_array)
print(multi_dim_array)

Sample Program

This program shows the array before and after shuffling to see the change in order.

NumPy

import numpy as np

# Create an array of numbers 1 to 6
numbers = np.array([1, 2, 3, 4, 5, 6])

print("Before shuffling:", numbers)

# Shuffle the array in-place
np.random.shuffle(numbers)

print("After shuffling:", numbers)

Important Notes

Time complexity is O(n), where n is the number of elements.

Space complexity is O(1) because shuffling happens in-place without extra arrays.

Common mistake: expecting np.random.shuffle() to return a new array. It does not; it changes the original array.

Use shuffling when you want to randomize data order. Use sampling if you want random elements without changing the original array.

Summary

Shuffling mixes array elements randomly to remove order bias.

np.random.shuffle() changes the array in-place and works on the first axis for multi-dimensional arrays.

It is useful for preparing data for machine learning and random sampling tasks.

Practice

1. What does np.random.shuffle() do to a NumPy array?

easy

A. Randomly rearranges the elements of the array in-place

B. Sorts the array elements in ascending order

C. Creates a new shuffled copy of the array without changing the original

D. Reverses the order of elements in the array

Solution

1. Step 1: Understand the function purpose

   np.random.shuffle() is designed to mix elements randomly within the array.

2. Step 2: Check how it modifies the array

   This function changes the original array directly (in-place), not creating a new one.

3. Final Answer:

   Randomly rearranges the elements of the array in-place -> Option A

4. Quick Check:

   Shuffle = random in-place rearrangement [OK]

Hint: Shuffle mixes elements inside the same array [OK]

Common Mistakes:

• Thinking shuffle returns a new array
• Confusing shuffle with sorting
• Assuming shuffle reverses elements

2. Which of the following is the correct syntax to shuffle a 1D NumPy array named arr?

easy

A. np.random.shuffle(arr, axis=1)

B. np.shuffle(arr)

C. arr.shuffle()

D. np.random.shuffle(arr)

Solution

1. Step 1: Identify the correct function call

   The shuffle function is inside the np.random module, so it must be called as np.random.shuffle().

2. Step 2: Check the parameters

   It takes the array as the only argument. The axis parameter is not valid for 1D arrays.

3. Final Answer:

   np.random.shuffle(arr) -> Option D

4. Quick Check:

   Correct shuffle syntax = np.random.shuffle(array) [OK]

Hint: Use np.random.shuffle(array) to shuffle in-place [OK]

Common Mistakes:

• Using np.shuffle instead of np.random.shuffle
• Calling shuffle as a method on the array
• Passing axis parameter incorrectly

3. What will be the output of the following code?

import numpy as np
arr = np.array([1, 2, 3, 4, 5])
np.random.shuffle(arr)
print(arr)

medium

A. [1 2 3 4 5]

B. [5 4 3 2 1]

C. A randomly shuffled version of [1 2 3 4 5]

D. Error because shuffle returns a new array

Solution

1. Step 1: Understand shuffle effect on array

   np.random.shuffle(arr) rearranges elements randomly in-place, so arr changes order.

2. Step 2: Predict output of print

   Since shuffle is random, the printed array will be a shuffled version of the original, not sorted or reversed.

3. Final Answer:

   A randomly shuffled version of [1 2 3 4 5] -> Option C

4. Quick Check:

   Shuffle output = random order of original array [OK]

Hint: Shuffle changes order randomly, output varies each run [OK]

Common Mistakes:

• Expecting original order after shuffle
• Thinking shuffle returns a new array
• Assuming shuffle sorts or reverses

4. Identify the error in the following code snippet:

import numpy as np
arr = np.array([[1, 2], [3, 4]])
np.random.shuffle(arr, axis=1)
print(arr)

medium

A. shuffle does not accept an axis argument

B. arr must be 1D to shuffle

C. shuffle returns a new array, so assignment is needed

D. np.random.shuffle cannot shuffle 2D arrays

Solution

1. Step 1: Check shuffle function parameters

   np.random.shuffle() only accepts the array argument; it does not have an axis parameter.

2. Step 2: Understand shuffle behavior on 2D arrays

   Shuffle works in-place on the first axis of multi-dimensional arrays by default, no axis argument needed.

3. Final Answer:

   shuffle does not accept an axis argument -> Option A

4. Quick Check:

   shuffle axis param invalid = error [OK]

Hint: np.random.shuffle has no axis parameter [OK]

Common Mistakes:

• Passing axis argument to shuffle
• Thinking shuffle returns a new array
• Believing shuffle only works on 1D arrays

5. You have a 2D NumPy array representing 5 samples with 3 features each:

data = np.array([[10, 20, 30],
                 [40, 50, 60],
                 [70, 80, 90],
                 [15, 25, 35],
                 [45, 55, 65]])

You want to shuffle the samples (rows) but keep the feature order intact. Which code correctly does this?

hard

A. np.random.shuffle(data.T)

B. np.random.shuffle(data)

C. np.random.permutation(data)

D. np.random.shuffle(data, axis=1)

Solution

1. Step 1: Understand shuffle on 2D arrays

   np.random.shuffle() shuffles along the first axis (rows) in-place, which is what we want.

2. Step 2: Evaluate other options

   np.random.shuffle(data.T) shuffles columns (wrong axis), np.random.permutation(data) returns a new shuffled array (not in-place), np.random.shuffle(data, axis=1) is invalid (axis param not accepted).

3. Final Answer:

   np.random.shuffle(data) -> Option B

4. Quick Check:

   Shuffle rows in-place = np.random.shuffle(data) [OK]

Hint: Shuffle rows by calling np.random.shuffle on 2D array [OK]

Common Mistakes:

• Trying to shuffle columns by transposing
• Expecting shuffle to return a new array
• Passing axis parameter to shuffle