What are anonymous functions in MATLAB, and how can they be used to simplify code? Provide an example where an anonymous function is used effectively.

What are anonymous functions in MATLAB, and how can they be used to simplify code? Provide an example where an anonymous function is used effectively.

[Torsten]

You seem to have internet. Thus I suggest to type

anonymous function & MATLAB

into a search engine.

Answer 1

[recent works]

Anonymous functions, also known as function handles, are nameless functions that can be defined using the @(arguments) expression syntax.

Program:

% Define the anonymous function
squareFunc = @(x) x.^2;

% Sample array
arr = 1:5;

% Apply the anonymous function
squaredArr = squareFunc(arr);

disp("Original array: " + arr);
disp("Squared array: " + squaredArr);

[Walter Roberson]

Anonymous functions are one kind of function handles. There are other function handles that are not anonymous functions, such as

f = @sin

Answer 2

[chicken vector]

Anonymous functions are used to store a function into a variable:

doubleNumber = @(x) 2*x;
doubleNumber(10)

ans = 20

The syntax is expressed as:

variableName = @(variable) function(variable)

I can give you two examples for which anonymous functions can be useful but there are many more.

Example #1:

It may happen that you have some descrete values and you want to make them continuous via interpolation.

If you are going to use this values many times, it can make your code more readable to do:

% Base data:
population = [1000 1100 1150 1200 1300 1375];
years = [2000 2004 2008 2012 2016 2020];

% Anonymous function:
populationHandle = @(year) interp1(years,population,year);

% Example:
populationHandle(2018)

ans = 1.3375e+03

Example #2:

When developing a program, you may want to give your code some flexibility based on user input.

For example, if your code needs an integration with ode45, but the user can select, i.g. between two integrating functions, you can create these two function and use a function handle to decide the input of ode45.

Let's say that you can integrate your dynamics in 2D and 3D, using the functions dynamics2D.m and dynamics3D.m, and the user can decides which one to use. Then, you can setup your code as follows:

% User input:
userInput = '2';

% Your code:
functionHandle = str2func(['dynamics' userInput 'D'])

functionHandle = function_handle with value:
    @dynamics2D

You can use the function handle in the integrator such that your code can adapt to the user input:

propagation = ode45(@(t,y) functionHandle(t,y),tspan,y0,odeOptions);

Beware, this is not the only way to do this as a simple switch or if statement could also be used.