When even a tiny box is added to a semicore potential,
the ghost testing fails.
Logged In: YES
I fixed this today 10/6/02. This required many changes:
1) adding a separate routine to apply the aug. op.
2) calling ghostnl AT THE END of the run
3) passing the fact that an orbital is a semicore state to
ghostnl.f (so the 1st and 2nd excited local states are
I will have to test this thouroughly. -EJW
Log in to post a comment.