ckoerber
commented
4 years ago In the intro of PR #43, I've added the remark:
It is crucial to remove basis dependent scales and utilize zero-temperature continuum observables, as the authors in \Ref{Seki:2005ns} show that lattice artifacts induce terms in the scattering data in the infinite-volume limit. However, instead of utilizing continuum observables, improved discretization schemes are frequently employed in "\textit{ab-initio}" computations to negate the effect of finite lattice spacings because of technical difficulties or extensive numerical costs. For example, [...] to our knowledge, no numerical work leveraging these methods is in the continuum, either. Without a proper analysis of discretization effects, in contrast with the "\textit{ab-initio}" aspect, the accurcy of computed observables is uncertain.
Is this too critical?
NLEFT interactions are fiddled with (dispersion relations and smearing)
NLEFT is putting in a potential to counteract a discretization error... if you understand the error you might not need the 'fix'.