Create benchmark for batched time_delay_estimator based on synthetic data about jaybeams HOT 4 CLOSED

Implements a benchmark to compare jb::fftw::time_delay_estimator_many against jb::fftw::time_delay_estimator.
It intends to provide data to help towards answering the following question:
Is it really faster to run jb::fftw::time_delay_estimator_many with 100 arrays as input vs. running jb::fftw::time_delay_estimator 100 times?

This implementation does not worry about the statistics. The benchmark is written modeled after bm_time_delay_estimator.cpp

from jaybeams.

Task Done. I can't close task

from jaybeams.

I think if you say "This fixes #44" in the description of your pull request it automatically closes. We should consider creating an organization to keep the JayBeams code, that way both of us would have the same privileges.

from jaybeams.

from jaybeams.