- ~7x speedup on PearsonR. - ~30x speedup on SpearmanR (this will still be tight for hitting a 5 minute compute timeout on Quantopian). - ~25% speedup on LinearRegression.