OK...except the finite SH harmonic expansion suffers from aliasing, so
(0,0) changes. I did a quick check and the default (14) is about 1E-6
off whereas going up to 31 was about converged to about 1E-7.
On Wed, May 3, 2017 at 8:45 AM, Peter Blaha
wrote:
> Remember, in output0grr we calculate just
Remember, in output0grr we calculate just the average of grad rho / rho
For this average inside the sphere, only the lm=0 term is essential,
non-spherical terms of the fit of grad rho/rho do not enter.
On 05/03/2017 03:39 PM, Laurence Marks wrote:
I just had a look at *.output*grr* for an mBJ
I've never care about that.
On Wednesday 2017-05-03 15:39, Laurence Marks wrote:
Date: Wed, 3 May 2017 15:39:23
From: Laurence Marks
Reply-To: A Mailing list for WIEN2k users
To: A Mailing list for WIEN2k users
Subject: [Wien] XC fit for mBJ
I just had a look at *.output*grr* for a
I just had a look at *.output*grr* for an mBJ case, and noticed that
the XC fit is best described as awful. Has anyone investigated
whether, similar to -eece, a much larger spherical harmonic expansion
should be used. Maybe it does not matter.
--
Professor Laurence Marks
"Research is to see what
4 matches
Mail list logo
