[Q-e-developers] Bug in PW/input.f90 ?

[Paolo Giannozzi]

On Wed, 2012-01-11 at 10:40 +0100, Guido Fratesi wrote:

> there might be a wrong definition of the constrains for fixing the 
> magnetization in NC calculations with 'atomic direction'. The following 
> line in PW/input.f90, Revision 8421, line 711:
> 
> 	mcons(3,nt) = cos(theta) * pi / 180.0_dp
> 
> should be:
> 
> 	mcons(3,nt) = cos(theta)
> 
> (and not cos(theta*pi/180), since the conversion to radians is made 
> earlier in that file).

actually what is calculated now is not even cos(theta*pi/180),
but  cos(theta)*pi/180 which seems to me just plain wrong ...

P.
-- 
Paolo Giannozzi, IOM-Democritos and University of Udine, Italy