Parallelization for large datasets

Hi Michael,

Big fan of the repo - thank you very much. I haven't worked up the code yet, but I'd like to submit a PR for loop parallelization across cores for some functions. I'm working with datasets containing ~1e6 data points, creating ~1e8 pairs for each lag distance, and am finding that `varmapv` is particularly cumbersome at this time.

Thanks!