Two's Complement

Compute the two's complement value of an unsigned integer

Description

This is a very simple utility that computes the two's complement value from an unsigned integer input. The number of bits of the input integer must also be specified

Acknowledgements

The author wishes to acknowledge the following in the creation of this submission:
Two's Complement for MATLAB, from decimal to two's complement

• MATLAB release: MATLAB 7 (R14)

[Loren Shure]

Would be nice if this were vectorized. That would mean not using the if statement but instead operating on the entries where msb was true, but not the others.

[Jeremy]

um...the 2's complement of an unsigned value is just the dec2bin function...what's the point if it's unsigned?

[Anthony Ghannoum]

Hello, I hope you don't mind, I modified it a bit to handle matrices, and overflows. Cheers !

function value = TwosComplement(x, numBits)
if ~isnumeric(x) || ~isnumeric(numBits) || nargin ~= 2
    error (' Function requires two numeric input arguments');
end
if numBits < 2
    error ('numBits should be at least 2 for a Twos Complement representation');
end
% Create MSB matrix (Shift right, and truncate remaining bits if any)
msb = mod(bitshift(x, -(numBits - 1) ),2);
% Calculate all positive values (Truncate the input)
value = mod(x,2^(numBits - 1));
% Calculate all negative values (Invert and add 1)
neg = -1 * ( bitcmp(mod(x, 2^(numBits -1)), numBits -1) + 1 );
% Use MSB matrix as a mask to choose either the positive or negative value
value(msb ~= 0)= neg(msb ~= 0);
end

[Przemyslaw Zelazowski]

Similar to Jeremy, I don't see the functionality of the code if the input must be unsigned integer. Isn't the two's component mainly used to deal with negative numbers?