fix filtering.resample output for even values of num parameter

[mattkinsey]

This just fixes the handling of the nyquest frequency bins when resampling using the FFT method (filtering.resample).

Without enforcing the fourier space be hermitian symmetric like this, the resampled output had a few undesirable properties:

1. It had a different integrated power spectrum than the input
2. It was not reversible i.e. resample(resample(y, <larger size than y>), <original y size>) != y
3. The output disagreed with scipy.signal.resample. The max error was around 1e-2 in my application.

Thanks

[GPUtester]

Can one of the admins verify this patch?

Admins can comment ok to test to allow this one PR to run or add to allowlist to allow all future PRs from the same author to run.

[GPUtester]

Can one of the admins verify this patch?

Admins can comment ok to test to allow this one PR to run or add to allowlist to allow all future PRs from the same author to run.

[awthomp]

ok to test

[awthomp]

Hey @mattkinsey ! Thanks for the PR; looks great! I had to change the base branch you were merging into (branch-22.12 rather than main) and fix some minor merge conflicts, but otherwise looks good. We'll get this merged ASAP. Thanks for your contributions, and Happy Thanksgiving!

[awthomp]

rerun tests

[awthomp]

@gpucibot merge