<div dir="ltr"><div><div><div><div>In my opinion it is neither simple not very useful. What might be useful in my opinion is the possibility to<br>1. restart from lower-cutoff charge density AND wavefunctions, and<br></div></div>2. interpolate wavefunctions to a denser grid of k-points, adding the missing bands if needed.<br></div>The charge density alone often typically doesn't make a big difference. Moreover the charge density, now written in real space, should be written in G-space, like wavefunctions, so that one can use a similar logic (or illogic).<br><br></div>Paolo<br></div><div class="gmail_extra"><br><div class="gmail_quote">On Thu, Sep 29, 2016 at 4:42 PM, Nicola Marzari <span dir="ltr"><<a href="mailto:nicola.marzari@epfl.ch" target="_blank">nicola.marzari@epfl.ch</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class="">On 29/09/2016 16:36, Lorenzo Paulatto wrote:<br>
> If you want to have a try at implementing this rouself you have to start<br>
> from./Modules/xml_io_base.f90, subroutine read_rho_xml (line 668 in the<br>
> current development version), in the subroutine you'll find a line<br>
> IF ( nr1 /= nr(1) .OR. nr2 /= nr(2) .OR. nr3 /= nr(3) ) &<br>
> CALL errore( 'read_rho_xml', 'dimensions do not match', 1 )<br>
><br>
> You should change the code in order to have it read even in the case<br>
> where the dimensions do not match, as long as nr(1)>nr1, ..., but read<br>
> in a temporary buffer, take it to G space (with the old FFT grid, this<br>
> is the hard part), pad it with zeros to the new grid size and take back<br>
> to real space with the fine grid.<br>
><br>
> Not impossible, but does require some effort and some testing.<br>
><br>
<br>
<br>
</span>Thanks Lorenzo, much appreciated.<br>
<br>
Maybe this could be a topic for the hackaton in January - it is,<br>
to be honest, very easy compared to anything else done in QE, and<br>
would be super-super useful - a lot of calculations could start with<br>
lower cutoffs and especially low FFT grids, and ramp up only at<br>
the very end - if used well it could boost computation time significantly.<br>
<span class="HOEnZb"><font color="#888888"><br>
<br>
nicola<br>
<br>
<br>
<br>
--<br>
------------------------------<wbr>------------------------------<wbr>----------<br>
Prof Nicola Marzari, Chair of Theory and Simulation of Materials, EPFL<br>
Director, National Centre for Competence in Research NCCR MARVEL, EPFL<br>
<a href="http://theossrv1.epfl.ch/Main/Contact" rel="noreferrer" target="_blank">http://theossrv1.epfl.ch/Main/<wbr>Contact</a> <a href="http://nccr-marvel.ch/en/project" rel="noreferrer" target="_blank">http://nccr-marvel.ch/en/<wbr>project</a><br>
______________________________<wbr>_________________<br>
Q-e-developers mailing list<br>
<a href="mailto:Q-e-developers@qe-forge.org">Q-e-developers@qe-forge.org</a><br>
<a href="http://qe-forge.org/mailman/listinfo/q-e-developers" rel="noreferrer" target="_blank">http://qe-forge.org/mailman/<wbr>listinfo/q-e-developers</a><br>
</font></span></blockquote></div><br><br clear="all"><br>-- <br><div class="gmail_signature" data-smartmail="gmail_signature"><div dir="ltr"><div><div dir="ltr"><div>Paolo Giannozzi, <span>Dip. Scienze Matematiche Informatiche e Fisiche</span>,<br>Univ. Udine, via delle Scienze 208, 33100 Udine, Italy<br>Phone +39-0432-558216, fax +39-0432-558222<br><br></div></div></div></div></div>
</div>