This is a set of math/computer-science problems, not just 1.
Given any function F and an arbitrary vector V, what’s the optimal algorithm to permute the values in V, such that the DFTs of F and V are as similar as possible?
Optionally, either of F or V can be set to a fixed value, thereby allowing specialized algorithms to be designed. If both were fixed values, then the algorithm would be trivial.
Given an audio file A and a binary file B, what’s the fastest program to rearrange non-overlapping N-bit chunks in B, such that playing B as a N-bit THz PCM sounds as similar as possible to A?
N & T can be either (or both):
Just like the general problem, either (but not both) of A or B can be set to arbitrary data.
If I asked you to rearrange the ASCII characters of the entire “Bee Movie” script, such that the bytes that represent those chars can be played as 8-bit ~40Khz PCM that sounds like N.G.G.Y.U. (Rick-Roll), how would you do it?
I bet some mathematician or computer-scientist can come up with an insane conjecture about this