How to convert byte number into float in Simulink

Hello,

In my model, I have input as byte number and need to convert into float number, how can I do this in MATLAB simulink?

like, input is 2 byte and need to convert to float variable so I can compare this number with another floating number in simulink.. without this my generated code is not working correctly. I need to convert byte into float.

Any feedback is appreciated.

Thank you in advance!

Answer 1

[Walter Roberson]

https://www.mathworks.com/help/simulink/slref/datatypeconversion.html Data Type Conversion block

[Tejas]

Actually I used it, but it wan't get me right output... do this block directly convert byte into float? or need do any other action?

Like when I removed my byte input and feed float as input... then comparision work and my generated code also work as expected..

But when I feedback byte as input then the comparision is not happening between byte and float in if action block and output is not as expected..

[Walter Roberson]

I have no reason to expect that for the fundamental types that the generated code is doing anything other than cast or just directly calling single() or double()

If you have reason to double, then create a MATLAB Function Block along the lines of

out = function fcn(u1)
  out = single(u1);      %or double() as appropriate
end

Question: is your input possibly an array of uint8? If so then

out = function fcn(u1)
  temp16 = typecast(u1, 'uint16');   %or int16 as appropriate
  out = single(temp16);      %or double() as appropriate
end

If it is an array of uint8 then you might potentially need to swapbytes(temp16)

Mind you you might prefer the simplicity of

out = function fcn(u1)
   out = single(u1(1)) * single(256) + single(u1(2));
end

which allows you to easily exchange the (1) and (2) if the data is in the other byte order. The expression is a bit longer if you are using signed integers.

[Tejas]

Thanks for the explanation.. I will try this way.. hopefully work out..