<div dir="ltr">Dear Matteo,<br>                       perhaps it is just my supposition but in the case of a D4d symmetry the translational symmetry is not enough.<div>The rare earth is inside a square anti prism crystal field, probably It would need to explicit more peculiar symmetry operations as  improper rotation, also called rotoreflection ( they can be implemented with reflections along vertical mirror planes ) in order to describe accurately the system.</div><div><br></div><div>S.</div><div><div><br></div><div><br clear="all"><div><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> Unfortunately the symmetry operations implemented so far in Quantum</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> Espresso can not handle screw rotations and glide reflections as in a</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> standard Quantum Chemistry package</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">? QE should work with all point symmetries that are compatible</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">with translational symmetry</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">P.</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> . So I am trying to impose such occupations as an approximation to the</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> real symmetry of the system. Hereafter I will try also your "dirty</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> trick" suggesting a starting_ns_eigenvalue bigger than one. Using the</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> mixing_fixed_ns > electron_maxstep I obtained big differences in the</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> occupations imposed and obtained, interestingly although I did not end</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> in the predetermined states  the system fell always in the same</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> strange occupation. The convergence was quite nice (tens of</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> iterations) up to 10^-6 Ry. Perhaps, as you pointed out I have a</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> different symmetry than the code finds for my system. Unfortunately,</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> although I would like to try to develop some new parts codes, I do not</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> feel enough trained and comfortable with the timetable of my current</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> project. Thanks a lot for your suggestion !</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> S</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> Dear Simone</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> I try to answer your questions below.</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> On Fri, Nov 7, 2014 at 3:32 PM, simone marocchi <</span><a href="mailto:simone.roz@gmail.com" style="font-family:arial,sans-serif;font-size:12.8000001907349px">simone.roz@gmail.com</a><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> wrote:</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> > Dear all, I am simulating compounds with rare earths within a</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> collinear</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> > calculation. I tried to suggest some occupations of the f orbitals</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> for the</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> > Tb atom, using the starting_ns_eigenvalue(m,</span><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">ispin,I).</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> ></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> why do you need to do that? if you want to force different values of</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> occupation on states that are equivalent by symmetry you will not be</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> successful. If this is the case you have to do something that makes</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> the</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> system loose that symmetry operation connecting the two states.</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> > Unfortunately also with a small value of electronic mixing and big</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> values</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> > of U, the imposed occupations was lost during the iterative cycle.</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> ></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> this can happen. a dirty trick I learned recently is to suggest a</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> starting_ns_eigenvalue(m,</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> ispin,I) bigger than one in input (1.2 or 1.3 maybe). This does not</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> make</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> any physical sense of course. However the Hubbard potential becomes</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> more</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> attractive for the specific eigenvector of the occupation matrix you</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> want</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> to fill completely and the code takes more time to "come back" to a</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> physical value. If that is a state it likes (at least a local minimum</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> of</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> the energy) it might fall into it. Of course you have to check at the</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> end</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> that the occupation has gone back to a value <= 1.</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> > So I used also mixing_fixed_ns > electron_maxstep. Also in this case</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> after</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> > I obtain the total energy convergence the density matrix of the last</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> > iteration is different to the one imposed in the input_file.</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> ></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> this is (possibly) strange. How different is it? How well are you</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> converging?</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> The fact that it is different is not surprising: the routine that</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> prints</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> the ns, always prints the ones that are computed from the KS states</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> just</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> obtained from the new diagonalization. These can be different from the</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> ones</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> e.g. used in contructing the Hubbard potential.</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> However if your calculation is well converged this difference should</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> not be</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> big: both KS wfcs and their occupations should be converged reasonably</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> well.</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> If this does not happen and you still see a significant difference</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> maybe it</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> means that the values you are trying to impose is not consistent with</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> what</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> the system wants to do (e.g., you have less symmetry than the code</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> finds</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> for your crystal)</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> > Can someone of you kindly explain me how the mixing_fixed_ns works ?</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> Do it</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> > uses a sort of Lagrange multipliers to force the occupations or is</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> more</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> > like a penalty function ? Finally, is it possible to work around the</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> > problem, converging to a determined occupation ?</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> ></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> no with the current version of the code. you could implement some</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> (e.g.,</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> quadratic) constraint and try. I think I once tried (long time ago)</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> and</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> seem to remember problems in convergence.</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> Best,</span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">></span><br style="font-family:arial,sans-serif;font-size:12.8000001907349px"><span style="font-family:arial,sans-serif;font-size:12.8000001907349px">> Matteo</span><br></div>-- <br><div class="gmail_signature"><div dir="ltr"><font color="#888888">Simone Marocchi</font><br style="color:rgb(136,136,136)"><br style="color:rgb(136,136,136)"><span style="color:rgb(136,136,136)">S3 Center, Istituto Nanoscienze, CNR</span><br style="color:rgb(136,136,136)"><span style="color:rgb(136,136,136)">via Campi 213/A, 41125, Modena, Italy</span><br style="color:rgb(136,136,136)"><span style="color:rgb(136,136,136)">Tel: </span><a value="+390592055300" style="color:rgb(17,85,204)">+39 </a><u>0592055585</u><span style="color:rgb(136,136,136)">;  Skype: jacobi84</span><br style="color:rgb(136,136,136)"><span style="color:rgb(136,136,136)">URL: </span><a href="http://www.nano.cnr.it/" style="color:rgb(17,85,204)" target="_blank">http://www.nano.cnr.it</a><br style="color:rgb(136,136,136)"></div></div>
</div></div></div>