How to create and empty matrix of a given size in a way that is compatible with codegen

1 view (last 30 days)
I have a function that is writing to an output matrix (y).
I can create y in the function using:
y = zeros(size(x));
x is a variable size large array and using zeros() is inefficient because it initializes the array with zeros that are then going to be overwritten.
I was wondering if there is a way to do this more efficiently?
I've tried a lot of the options I can find so far but none are compatible with codegen.
Thanks very much, as always,
John Edwards
John Edwards on 11 May 2022
Codegen requires that the size of the array is declared before use so I'd like to declare it and then fill it with the appropriate output values, rather than using the zeros function, which is then overwritten by the output data.

Sign in to comment.

Accepted Answer

Richard McCormack
Richard McCormack on 11 May 2022
Edited: Richard McCormack on 11 May 2022
Hi John,
I think coder.nullcopy is what you are looking for:
Be aware that this function is 'unsafe', in the sense that you can access uninitialized memory after using it.

More Answers (0)




Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!