In a not-so-surprising turn of events, calculating the rank of the current month's #pasture production by comparing it with all the previous #sentinel2 monthly estimates runs 80x faster using a dedicated jit Numba function instead of scipy.stats.percentileofscore