I am writing to report a small bug in the computation of amn in pw2wan2epw.f90.
The problem is that the guiding functions gf is normalized while it should not be. The relevant lines are https://gitlab.com/QEF/q-e/-/blob/devel ... L2999-3001
Code: Select all
anorm = REAL(ZDOTC(npw, gf(1, iw), 1, gf(1, iw), 1))
CALL mp_sum(anorm, intra_pool_comm)
gf(:, iw) = gf(:, iw) / DSQRT(anorm)
(In the MR, I only fixed PP/src/pw2wannier90.f90, not the EPW one.)
Regards,
Jae-Mo Lihm
Seoul National University