[Wien] How to continue SCF after the process is killed?

[Peter Blaha]

> It has taken me half a month for a large system SCF calculation in wien2k (run_lapw ¨Cec 0.0001 ¨Ci 50 &). All went well until somebody killed my process by accident. My question is how can I restart the SCF right where it stopped? Is this possible? It stopped in the thirtieth iteration when lapw1 is running, maybe I just do the following command can solve the problem: run_lapw ¨Cec 0.0001 ¨Ci 50 ¨Cs lapw1 &. What do you think about, should I do this after save_lapw or deleting all the error files? Please tell me step by step. Thanks in advance.

a) First check if the calculations at least start to converge:
grep :DIS case.scf
Does the distance get smaller ? If yes, you may simply continue:

b) rm *.error;run_lapw -i 50 [-ec 0.0001 this is the default anyway]

c) Are you sure that you do the calculations
   on a powerfull enough machine ?
   with reasonable RKMAX (in relation to your RMTs)
   with reasonable K-mesh (in relation to the cell size, #of atoms,
      insulator/metal ?)

If ONE iteration takes 1 full day, most likely you do something unnecessary!!!
(or you have a very old computer?).
I would do such big systems only, if I'm already familiar with WIEN2k.

Regards
                                      P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-15671             FAX: +43-1-58801-15698
Email: blaha at theochem.tuwien.ac.at    WWW: http://info.tuwien.ac.at/theochem/
--------------------------------------------------------------------------