如何在 Octave 中将两个 8 位数字组合为 16 位

Question

I'm working on the data from a GY-87 (acc and gyro sensor module), The data I have received from it is of the form ACCEL_YOUT_HIGH & ACCEL_YOUT_LOW for the Y axis specifying the HIGH and the LOW value respectively. I've been told that "High and Low registers are 8-bit in nature, therefore the combination of both will be a 16-bit signed data". But I really can't figure out how to combine them to 16 bit data. Here is a sample: ACCEL_YOUT_HIGH=254 ACCEL_YOUT_LOW=144

I've already tried the JAVAScript way, but its not working in octave.

var number8Bit1 = firstNumber & 0xff;
var number8Bit2 = ((firstNumber >> 8) & 0xff);

Answer 1

Since your data is 16-bit signed, I think the most elegant solution, is using typecast (as Andy commented).

For typecast solution to work correctly you cad to do the following:

1. Cast each value to uint8
2. Build an array from the two uint8 elements (put the lower element first, because our systems are little-endian )
3. Use typecast for converting the two bytes to single int16 element.
4. Cast the result to double (you can skip this part, but it's a common practice - in case you need to apply computations to result).

Code sample:

high = 254;
low = 144; 
res = double(typecast([uint8(low), uint8(high)], 'int16'))

Result:

res = -368

---

In case you want to use pure mathematical operations (as Sami commented), you can do it as following:

res = high*256 + low;
if (res >= 2^15)
    res = res - 2^16; %Correct the sign.
end

I find it less elegant, because the if statement is unclear...

---

You can also use bitshift , bitor and typecast (as Cris commented).
It looks like the C code res = (short)((high << 8) | low) :

res = typecast(uint16(bitor(bitshift(high, 8), low)), 'int16');

I find it less elegant than the first solution because shift and or operations are not natural in MATLAB / Octave...