My calculations (and some of the calculations/results from Higson et al.) suggest that logzerr is generically underestimated by ~sqrt(2). I should consider actually adding this in by hand during runtime, with an associated flag for when users might want to turn this off.
My calculations (and some of the calculations/results from Higson et al.) suggest that
logzerris generically underestimated by ~sqrt(2). I should consider actually adding this in by hand during runtime, with an associated flag for when users might want to turn this off.