MATLAB: Matlab2017b why uint32(255)/128 = 2

MATLABuint32

Matlab2017b why uint32(255)/128 = 2?

Best Answer

Operations on integer data types are carried out as if converted to double with the results rounded when converted back.

(Operations are not necessarily actually carried out in double: uint64 does not drop down to double accuracy for example)

Related Solutions

MATLAB: How to perform bitwise operations on other data types besides unsigned integers in MATLAB

This enhancement has been incorporated in Release 2012b (R2012b). For previous product releases, read below for any possible workarounds:

The ability to perform bitwise operations on data types other than unsigned integers is not available in base MATLAB.

As a workaround, use the TYPECAST function to reinterpret the bits of a number of one datatype as an unsigned integer with the same number of bits, and then perform the bitwise operation. Finally, use TYPECAST to convert the result back to the original type. For example:

i1 = int8(-7)
u1 = typecast(i1, 'uint8')
u2 = bitshift(u1, 1)
i2 = typecast(u2, 'int8')

MATLAB: Matlab integer format for answer

Try a function, like int32(), fix(), ceil(), floor(), uint8(), int32(round()), etc.

a = (2+24)/(4*3) % Floating point.
aInteger = int32(a)
whos aInteger; % int32
aInteger = round(a)
whos aInteger; % Double


aInteger = floor(a)
whos aInteger; % Double
aInteger = ceil(a)
whos aInteger; % Double
aInteger = fix(a)
whos aInteger; % double
aInteger = uint16(a)
whos aInteger; % uint16

Note that with round(), ceil(), floor(), and fix() you will end up with a double but with no fractional part. If you want an actual integer, you have to cast to an integer data type with int32(), int16(), etc.

Best Answer

Related Solutions

MATLAB: How to perform bitwise operations on other data types besides unsigned integers in MATLAB

MATLAB: Matlab integer format for answer

Related Question