{"id":909,"date":"2023-03-16T09:48:38","date_gmt":"2023-03-16T13:48:38","guid":{"rendered":"https:\/\/blogs.mathworks.com\/matlab\/?p=909"},"modified":"2023-03-16T09:48:38","modified_gmt":"2023-03-16T13:48:38","slug":"function-handles-are-faster-in-matlab-r2023a","status":"publish","type":"post","link":"https:\/\/blogs.mathworks.com\/matlab\/2023\/03\/16\/function-handles-are-faster-in-matlab-r2023a\/","title":{"rendered":"Function handles are faster in MATLAB R2023a"},"content":{"rendered":"
\r\n
Roughly speaking, there are two ways we at MathWorks can speed up MATLAB. We can dive into individual functions to remove overheads, improve algorithms and so on or we can make performance enhancements to the language itself that have the potential to speed-up a range of computations across many domains. Today, I'm going to take a look at an example of the latter where function handles have been made faster as part of the R2023a release.<\/div>\r\n

What are function handles?<\/h3>\r\n
Straight from the documentation<\/a>: \"A function handle is a MATLAB data type that represents a function. A typical use of function handles is to pass a function to another function. For example, you can use function handles as input arguments to functions that evaluate mathematical expressions over a range of values.\"<\/div>\r\n
For example, say you have the function<\/div>\r\n
\r\n
function <\/span>y = computeSquare(x)<\/span><\/div>\r\n
y = x.^2;<\/span><\/div>\r\n
end<\/span><\/span><\/div>\r\n<\/div>\r\n
We can create a handle<\/div>\r\n
\r\n
\r\n
f = @computeSquare<\/span><\/div>\r\n
\r\n
\r\n
\r\n
f = function_handle with value:<\/span><\/span><\/div>\r\n
@computeSquare<\/div>\r\n<\/div>\r\n<\/div>\r\n<\/div>\r\n<\/div>\r\n<\/div>\r\n
and use it to call the function to compute the square of four as follows:<\/div>\r\n
\r\n
\r\n
b = f(4)<\/span><\/div>\r\n
\r\n
b = 16<\/div>\r\n<\/div>\r\n<\/div>\r\n<\/div>\r\n
This simple looking idea turns out to be extremely useful and you'll find function handles used all over the place in MATLAB code.<\/div>\r\n

Function handle speed test<\/h3>\r\n
Let's use our function handle to compute the square of 100 million numbers without storing the results. A somewhat pointless calculation but it will serve our purposes in demonstrating that the function handle overhead has been significantly reduced.<\/div>\r\n
\r\n
\r\n
n = 1e8;<\/span><\/div>\r\n<\/div>\r\n
\r\n
tic<\/span><\/div>\r\n<\/div>\r\n
\r\n
for <\/span>i = 1:n<\/span><\/div>\r\n<\/div>\r\n
\r\n
out = f(i);<\/span><\/div>\r\n<\/div>\r\n
\r\n
end<\/span><\/span><\/div>\r\n<\/div>\r\n
\r\n
t = toc<\/span><\/div>\r\n
\r\n
t = 0.7665<\/div>\r\n<\/div>\r\n<\/div>\r\n<\/div>\r\n
On my machine, I get the following times<\/div>\r\n