The ABINIT Users Mailing List ( CLOSED )

["Allan, Douglas C Dr" <AllanDC@Corning.com>]

Unless the convention has changed, the complex wavefunction coefficients CG 
should be normalized to 1.0 separately for each band.  There should not be 
any factor of unit cell volume or spin degeneracy in the wavefunction 
coefficients themselves -- such factors are applied later where needed.  For 
example, to produce an electron density you would need spin degeneracy and 
unit cell volume. I can only guess that the simple sum was not performed 
correctly or was not reading the CG values correctly.
Doug Allan

-----Original Message-----
From: Kristopher Andersen [mailto:andersen@yclept.ucdavis.edu] 
Sent: Friday, June 20, 2003 7:26 PM
To: ABINIT Forum
Subject: [abinit-forum] wavefunction normalization


I was doing some simple tests of the wavefunction, as read from _WFK, and 
found that the wavefunction doesn't appear to be normalized? For example, a 
simple sum of abs(cg)**2 gives 0.5 instead of 1.0? (Where cg are the 
coefficients in the planewave expansion of the
wavefunction.) Also, the values seem to fluctuate quite a bit per band. I saw 
numbers between 0.4 and 0.8. Shouldn't the normalization be essentially 
identical between bands? Does anybody know what's happening here? From what I 
can tell, the wavefunction coefficients, as read from _WFK, need to be 
multiplied by a factor of sqrt(2.0 / V), where V is the volume of the unit 
cell. Is this correct?

-- 
Kristopher Andersen               http://yclept.ucdavis.edu/~andersen
Department of Physics                    andersen@physics.ucdavis.edu
University of California at Davis                   +1 (530) 752-0446