Currently something like DQMC(model, beta=2.2, delta_tau=0.1, safe_mult=10) fails, because we require that N * safe_mult * delta_tau == beta. We can simplify this requirement to N * delta_tau == beta if we either
perform an irregular stabilization at the first (last?) time propagation or
Currently something like
DQMC(model, beta=2.2, delta_tau=0.1, safe_mult=10)
fails, because we require thatN * safe_mult * delta_tau == beta
. We can simplify this requirement toN * delta_tau == beta
if we either