[Pw_forum] from invmat : error #1 singular matrix

[Paolo Giannozzi]

Lorenzo Paulatto wrote:

>>                        ibrav = 0,
>>                    celldm(1) = 1

> if you use ibrav=0 you have to manually specify the unit cell, using the  
> CELL cards list, in units of celldm(1) bohrs(see the input manual).

I think it is correct as it is now. One can
- specify the kind of lattice (ibrav) and the lattice parameters
   (celldm, or else a,b,c ...)
- specify lattice vectors (ibrav=0 + CELL card); the lattice
   parameter is assumed to be the length of lattice vector 1
- specify lattice vectors and lattice parameter (ibrav=0 +
   celldm(1), or else a + CELL card); the lattice vectors are
   in units of the lattice parameter
I think that the third option is a very bad idea, but it shouldn't
be wrong. All these cases are checked for consistency in input anyway.

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