Playing With Random Numbers in MATLAB and its Commands

 

MATLABit

MATLAB stands for MATrix LABoratory. It’s a powerful programming language and software tool created by MathWorks. Its extensive application across engineering, scientific research, academic instruction, and algorithmic design stems from its strengths in numerical computation, data analysis, graphical visualization, and simulation. MATLAB effectively handles big datasets and intricate mathematical models thanks to its foundation in matrix algebra. So, let's commence to know how to generate random numbers in MATLAB.

Table of Contents

• Introduction
• Generation of Random Numbers in MATLAB
• Applications
• Conclusion
• Tips in MATLAB for Playing with Random Numbers

Introduction

In scientific computing, engineering analysis, and physical simulations, random numbers are often required to model uncertainty, represent noise, or execute probabilistic algorithms. MATLAB provides several built-in functions to generate random numbers for various distributions. The most common among them are rand, randi, and randn. Each command serves a specific purpose — generating uniformly distributed real numbers, uniformly distributed integers, and normally distributed real numbers, respectively. Understanding their usage, syntax, and transformation methods enables users to simulate realistic data and perform stochastic modeling efficiently.

Generation of Random Numbers in MATLAB

1. The rand Command

> v = 30 * rand(1,8) - 10
v = 12.4387 7.2165 1.2458 17.9023 -8.4631 19.1152 -2.5847 10.7653

2. The randi Command

The randi function generates uniformly distributed random integers. It allows specifying both the upper and lower limits of the range. This command is particularly useful in generating random indices, simulation of discrete events, and randomized testing.

Command	Description	Example
randi(imax)	Generates a random integer between 1 and imax.	>> a = randi(20) → a = 14
randi(imax, m, n)	Generates an m×n matrix of random integers between 1 and imax.	>> b = randi(20, 3, 2) → b = 11 3; 8 17; 15 12
randi([imin, imax], m, n)	Generates an m×n matrix of random integers between imin and imax.	>> d = randi([100 150], 3, 3) → d = 142 121 109; 118 145 136; 130 127 101

3. The randn Command

The randn command generates normally distributed random numbers with a mean of 0 and a standard deviation of 1. These numbers can be scaled and shifted to achieve different mean and standard deviation values. This function is highly useful in modeling noise and other natural random variations.

Command	Description	Example
randn	Utilizes the conventional normal distribution to produce a single random number.	>> randn → ans = -0.8123
randn(m, n)	Generates an m×n matrix of normally distributed numbers.	>> d = randn(3, 4) → d = -0.8123 0.2257 -1.5142 0.8791; 0.4725 -0.3489 1.2314 -0.5821; 1.0198 0.6543 -0.1278 0.3126

To change the mean (μ) and standard deviation (σ) of these numbers:

v = σ * randn + μ

Example: generating six random numbers with mean 40 and standard deviation 8.

> v = 8 * randn(1,6) + 40
v = 43.7125 35.1982 47.5264 41.9310 30.5862 38.1448

If integer values are needed, they can be obtained using the round function:

> w = round(8 * randn(1,6) + 40)
w = 37 44 41 39 42 33

Applications

1. Monte Carlo Simulations: Random numbers are used to approximate complex mathematical models and evaluate integrals through repeated random sampling.
2. Noise Generation in Signal Processing: The randn function is used to add Gaussian noise to clean signals for testing filters or algorithms.
3. Randomized Algorithm Initialization: Machine learning and optimization techniques often use random numbers to initialize parameters or weight vectors.
4. Data Shuffling and Sampling: Random numbers generated through randperm or randi help in splitting datasets into training and testing portions.
5. Game Development and Simulation: In gaming, random numbers determine unpredictable outcomes such as dice rolls or random events.
6. Statistical Modeling: Random numbers form the basis for creating synthetic datasets, sampling distributions, and hypothesis testing simulations.

Conclusion

The ability to generate random numbers is central to computational science and engineering. MATLAB's rand, randi, and randn functions provide an efficient and versatile way to produce random numbers for different purposes — from uniform and normal distributions to integer-based random events. With the right scaling, rounding, and shifting operations, these functions can model almost any random variable needed in simulations or analysis. By combining them with proper seed control using rng, one can ensure reproducibility and consistency across experiments. Overall, MATLAB offers a robust platform for all random number generation requirements in academic, industrial, and research-based applications.

Tips in MATLAB for Playing with Random Numbers

The following tips will help you effectively use random number generation functions in MATLAB such as rand, randn, and randi.

1. Set the Seed for Reproducibility

Random numbers differ every time you run the program. Use a seed to get the same results repeatedly:

rng(0);      % Sets the seed for reproducibility
a = rand(1,5)

Use rng('default') to reset MATLAB’s random number generator to its default settings.

2. Check or Save the Generator Settings

Check or save the current generator configuration for reproducibility:

s = rng;     % Save current random number generator settings
rng(s);      % Restore settings later

3. Generate Random Numbers in a Specific Range

To create uniform random numbers between two limits a and b:

a = -5; b = 10;
r = (b - a) * rand(1,10) + a;

4. Generate Random Integers in a Range

To generate integer values within a given range:

r = randi([50 90], 3, 4);

This produces a 3×4 matrix of random integers between 50 and 90.

5. Normal Distribution with Custom Mean and Standard Deviation

Adjust the mean and standard deviation of normally distributed data:

mu = 50; sigma = 6;
v = sigma * randn(1,6) + mu;

6. Integers from Normally Distributed Numbers

Use rounding to convert continuous random numbers into integers:

w = round(4*randn(1,6) + 50);

7. Random Permutations

Generate random arrangements of integers:

p = randperm(8);

Useful for random sampling, random order testing, or shuffling data.

8. Visualizing Random Distributions

Visualize the distribution of generated numbers:

x = randn(1,1000);
histogram(x, 30);    % Normal distribution

y = rand(1,1000);
histogram(y, 20);    % Uniform distribution

9. Generate Random Logical Arrays

Create random true/false arrays for binary simulations:

logicalArray = rand(1,10) > 0.5;

10. Use Different Random Number Streams (Advanced)

When performing parallel computations, assign different random seeds:

parfor i = 1:4
    rng(i);        % Unique seed for each worker
    A{i} = rand(3);
end

---

Summary: By using these tips—especially setting the seed, customizing distributions, and visualizing results—you can ensure reproducibility and accuracy in MATLAB simulations that rely on random number generation.

Sidebar

Useful links, recent posts, or ads.

© 2025 MATLABit. All rights reserved.