(a) Show that the phase factors can be computed recursively by.

(b) Perform this computing once using single-precision floating-point arithmetic and once using only four significant digits. Note the deterioration due to the accumulation of round-off errors in the later case.

(c) Show how the results in part (b) can be improve by resetting the results to the correct value ? j. each time ql = N/4.

W = w wl-i) N.