how define input data type

when defining a function, for example

function y = func(x)

how can I define that x is of a specific type, like uint8?

Answer 1

[David Young]

See

 doc validateattributes

Answer 2

[Jan]

I do not see the problem.

function y = func(x)
disp(x + x)

And then call it with an UINT8:

func(uint8(1:10))

Or do you want to reject inputs with a deviating type? Then:

function y = func(x)
if ~isa(x, 'uint8')
   error('Bad type!');
end
disp(x + x)

Or do you want to convert the input?

function y = func(x)
x = uint8(x);
disp(x + x)

Answer 3

[Vasiliki]

I want to generate c code using the embedded toolbox. The default type of the input is double, but I need integers or fixed point in order to use the c code in a microprocesor. So, I want to see if there is a way that matlab can "understand" that the input will be integer so that when it generates the c code i will have:

uint16 func (uint16 u)

Of course after the generation of the code I can do it maually, but I want to see if there is an automatic way.

[Vasiliki]

it can be done at the compilation with emlc and -eg emlcoder.egs, defining that the input is of variable size and its type and class is the desirable one.