How to quickly and simply convert floating-point to fixed-point in matlab script?

I want to compare the performance and speed of running the same script as floating-point and fixed-point. How do I simple convert the script or parts of the script into fixed-point from double?

I have tried fi(). However there seems to be a problem using it within function such as fft's. What is the best way to do this?

Answer 1

[Gargi Patil]

MATLAB provides several tools for floating point to fixed point. There is a toolbox for Fixed-Point Conversion that enables conversion to fixed point:

https://in.mathworks.com/help/fixedpoint/index.html

You can refer to the Fixed-point converter app given in the toolbox:

https://in.mathworks.com/help/fixedpoint/ref/fixedpointconverter-app.html

Furthermore, to effectively convert FFT to fixed point, you can refer to the following documentation as a starting point:

https://in.mathworks.com/help/fixedpoint/ug/convert-fast-fourier-transform-fft-to-fixed-point.html