How do you subtract/add with roll under/over

When you subtract two uint numbers together example:

uint8(1) - uint8(2)

I would like the answer to be 255, instead no matter what the value MATLAB returns 0, likewise if you try adding:

uint8(255) + uint8(1)

MATLAB returns 255.

I have written my own add and subtract functions for this purpose which work fine but I am wondering whether there is actually something in built that I am missing. If there is could you please point me in the right direction.

Thanks.

[Matthew Doveton]

The way I posted my question appears to have caused some confusion. I do not have a problem implementing the desired effect myself but would like to know if this is an option within MATLAB that I'm unaware of (fairly new to MATLAB)

for uint8(1) - uint8(2) = 255, is what I want.

I basically want it to mimic my microcontroller code. If I were to write:

a = 1 - 2;

a would == 255, if a was a byte. In MATLAB a == 0;

As stated above, I have written a function to replicate this behaviour but thought that MATLAB would have this as an option. I personally would like MATLAB to mimic what would happen within a microcontroller as I use MATLAB primarily to test code that I then would convert to C for use within a microcontroller. I had some code that had an issue with rolling under at a certain condition. I could not replicate this in MATLAB due to this issue.

Answer 1

[Azzi Abdelmalek]

Edit

a=uint8(-1);
out=uint8(255)*uint8(not(a))+a

[Matthew Doveton]

for uint8(1) - uint8(2) = 255, is what I want.

I basically want it to mimic my microcontroller code. If I were to write:

a = 1 - 2;

a would == 255, if a was a byte. In MATLAB a == 0;

As stated above, I have written a function to replicate this behaviour but thought that MATLAB would have this as an option. I personally would like MATLAB to mimic what would happen within a microcontroller as I use MATLAB primarily to test code that I then would convert to C for use within a microcontroller. I had some code that had an issue with rolling under at a certain condition. I could not replicate this in MATLAB due to this issue.

[Azzi Abdelmalek]

Look at edited answer

Answer 2

[Sean de Wolski]

x = uint8(1);
y = uint8(2);

z = max(double(x)-double(y),0)+uint8(x < y).*255