You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Previously, `w(r)` as printed in `case.psink` was normalized as
`|w²(r)|/#k` leading to a factor `#k`. Fix this to `|w(r)/#k|²`.
Thanks to Ulrich Wedig.
The remaining factor is simply due to the units. If AU units are requested in case.inwplot, then sum(case.psink) * dV ≈ 1 holds, where in the simplest case dV = (L / [Np - 1])³ with axis length L as given in the header of case.psink and Np points in each direction.
If ANG units are selected, L must also be in Å, of course. Otherwise, an extra factor of (Å/Bohr)³ ≈ 6.75 appears.
The RLO-related factor remains to be tested in 2.0.
In
wplot
, the Wannier functions are currently not normalized. Rather,⟨w|w⟩ = Nk X
whereNk
is the number of k-points andX
an unknown factor.Additionally, in tests with SOC, there is a factor (
~2.16
) between a setup with and one without RLOs.X
is and take it into account as well.The text was updated successfully, but these errors were encountered: