[Wien] About EECE & SO

[Peter Blaha]

Whether you  do initso right at the beginning or after a first   runsp
may depend on the case/purpose. Personally I like to compare results with/without
so   and with/without eece, so I do all "possible calculations" at least once.

I recommend that first you use   runsp -so    and only when this is
(at least partially) converged, switch to   runsp -so -eece

Never do runsp -eece right after init_lapw as this may end up in some arbitrary
metastable state.


Am 26.11.2012 10:16, schrieb Francisco Garcia:
> Dear users,
>
> I have a question about the order in which -so and -eece calculations are run.
>
> (a) First -so. In the userguide, the steps are
>
> runsp_lapw... (without -so flag)----->initso------->runsp_lapw -so......
>
> I would like to know if above set of steps is equivalent to any of the
> two steps below:
>
> (i)  initso ---> runsp_lapw... (without -so flag) ---> runsp_lapw -so.....
> (ii) initso ---> runsp_lapw -so...... (bypass the no-so step and
> directly run so)
>
> (b) Next -eece. This is not clear in the userguide but I came up with
> two possibilities below. I would like to know if they are equivalent.
>
> (i) runsp_lapw (without -eece) -----> runsp_lapw -eece
> (ii) runsp_lapw -eece    (direct calculations)
>
>
> (c) Finally -eece and -so together. It is not clear which one comes
> first. I came up with the following possibilities and I would like to
> know which one is recommended.
>
> (i) runsp_lapw... (without -so and -eece
> flags)----->initso------->runsp_lapw -so (without -eece
> flag)-------------> runsp_lapw -so -ecce.....
> (ii) runsp_lapw... (without -so and -eece flags)----->runsp_lapw -eece
> (without -so flag)-------->initso-------> runsp_lapw -so -ecce.......
> (iii) initso-------> runsp_lapw... (without -so and -eece
> flags)----->runsp_lapw -eece (without -so flag)--------> runsp_lapw
> -so -ecce......
> (iv) initso-------> runsp_lapw... (without -so and -eece
> flags)----->runsp_lapw -so (without -eece flag)--------> runsp_lapw
> -so -ecce......
> (v)  initso--------> runsp_lapw -so -ecce..........
>
> I personally prefer to begin with initso from the onset (so that any
> symmetry changes from nso to so is avoided) and then runsp_lapw -so
> -ecce,,,,,,.However I could be wrong in my reasoning so I will
> appreciate any suggestions from experienced users on the best way to
> proceed. My main issue has to do with accurate energetics. Sometimes,
> the total energies are dependent on the order in which the
> calculations are carried out.
>
> Thank you.
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>

-- 
-----------------------------------------
Peter Blaha
Inst. Materials Chemistry, TU Vienna
Getreidemarkt 9, A-1060 Vienna, Austria
Tel: +43-1-5880115671
Fax: +43-1-5880115698
email: pblaha at theochem.tuwien.ac.at
-----------------------------------------