The bitshift function will change in a future release. This function will no longer accept an integer argument for N. Instead, you must specify a character vector containing the assumed integer type A.

Replace the number of bits N with an appropriate character vector that indicates the integer type of A: 'int8', 'uint8', 'int16', 'uint16', 'int32', 'uint32', 'int64', or 'uint64'.

You still can use the bitshift function on integers of arbitrary bit length. For example, bitshift(x,k,N) and bitand(bitshift(x,k),2^N-1) yield the same result:

x = 1000;
k = 3;
N = 10;
oldsyntax = bitshift(x,k,N)


oldsyntax =

   832
newsyntax = bitand(bitshift(x,k),2^N-1) newsyntax =

   832

For further information, see the bitshift function reference page.