Math Forum / Math Software / MATLAB / July 2008

Hi Michael,

One way to do it is to leave the product-mode in full-precision, then
assign into a left-hand-side argument to cast to a specific type.

For example, leave the ProductMode in the FullPrecision default, define
the types of the left-hand-side variables, and assign into them:

  A = fi(2, true,  8, 0);
  B = fi(3, true,  8, 0);
  C = fi(5, true, 16, 0);

  D = fi(0, true,  8, 0);  % To hold the first product
  E = fi(0, true, 16, 0);  % To hold the second product

  % The full-precision product is s16,0, then assign into s8,0
  D(:) = A*B  % Assign into D
  % D =  6   s8,0

  % The full-precision product is s24,0, then assign into s16,0
  E(:) = A*C  % Assign into E
  % E =  10  s16,0

If this doesn't answer the question, I would be happy to follow up with
you if you could send a specific example.

Best wishes,
Tom Bryan
tbryan@mathworks.com

p.s.

Here is a section from the demos, and there is more in the
documentation.  From the MATLAB command line, type demos.  Then select
Fixed-Point Toolbox, Fixed-Point Basics, and the section

A(:) = B vs. A = B

There is a difference between
 A = B

and
 A(:) = B

In the first case, A = B replaces A with B, and A assumes B's numeric type.

In the second case, A(:) = B assigns the value of B into A, while
keeping A's numeric type. This is very handy for casting one numeric
type into another.