Difference between interp1() and fixpt_interp1

Hello,

I can't really understand the difference in between these two functions.

Answer 1

[Walter Roberson]

fixpt_interp1 is the Fixed Point Toolbox equivalent of interp1, to be used when the inputs are Fixed Point numbers.

[Laercio Barbosa]

I don't think they are ONLY for fixed points as it accepts float('single')/float('double') as inputs.

Try:

fixpt_interp1(double([10 90]/3.6), double([10,9]), double(50/3.6), float('single'), [], float('single'), [], 'simplest')

[Walter Roberson]

single and double are considered special subsets of fixed point numbers, so that you can do rapid simulation without altering the code all over the place. You would then convert to strict fixed point for deploying to VHDL or FPGA, or to processors which do not offer hardware single or double; also in some cases you would use fixed point for "hard real time" work, because doing so reduces the uncertainty in timing.

Answer 2

[Laercio Barbosa]

From my point of view after read documentation about both is that interp1 only interpolates. fixpt_interp1 offers you the possibility of round the result to fixed point and also scale the inputs.