[Q-e-developers] Meshgrid in different calculations

[Paolo Giannozzi]

On Mon, Jun 22, 2015 at 11:29 AM, Jiqiang Li <jqli14 at fudan.edu.cn> wrote:

>
> (1)Do the elements of array for total local potential (variable *vrs*) or
> charge density (variable *rho*) in real space keep the same order in
> different  calculation when the volume and shape of supercell remain
> unchanged?
>

Not sure what you mean, but the logic of real-space arrays is always the
same:


> (2) If do, does the element in array *vrs*(nr1*nr2*(n3-1)+nr1*(n2-1)+n1)
> represent the position [x(n1), y(n2),z(n3)] in real space?
>

in the serial case: index (nr1x*nr2x*(n3-1)+nr1x*(n2-1)+n1) for an array of
dimension (nr1x*nr2x*nr3x) corresponds to index (n1,n2,n3) of the
three-dimensional array of dimensions (nr1x,nr2x,nr3x).
Index (nr1x*nr2x*(n3-1)+nr1x*(n2-1)+n1) correspond to real-space point
n1*at(:,1)+n2*at(:,1)+n3*at(:,3).
IN the parallel case, arrays are split across processors along the third
dimension (nr3) but the logic is the same


> (3) Is it correct that \sum_i {*rho*(i)}*volume/N is not exactly equal to
> the total charge?  N is the size of array *rho*.
>

No, it must be exactly equal to the total charge, and N=nr1*nr2*nr3, which
may differ from the size of the array rho


> (4)What's the difference between variables *nr1* and *nr1x*?
>

see above: one is the true dimension of the FFT transform, one is the
physical dimension of the array. They may differ due to efficiency reasons.

Paolo

Thank you for your consideration!
>
> Best regards!
> Jiqiang Li
> ------------------------------
>
> _______________________________________________
> Q-e-developers mailing list
> Q-e-developers at qe-forge.org
> http://qe-forge.org/mailman/listinfo/q-e-developers
>
>


-- 
Paolo Giannozzi, Dept. Chemistry&Physics&Environment,
Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
Phone +39-0432-558216, fax +39-0432-558222
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/developers/attachments/20150627/69e86520/attachment.html>