From the "coders" mailing list

This is not the first time such a request has come up.  As we
presumably wish to encourage users to perform convergence
testing, I have a suggestion.  How about two new tasks?

task : cutoff_test

   perform a series of singlepoint runs using FINITE_BASIS_NPOINTS
   and FINITE_BASIS_SPACING, printing the force and stress at each
   cutoff energy.  Better still, have some new variables to define the
   range.

task : kpoint_test

   perform a series of singlepoint runs for a range of MP grid
   sizes/spacings.  Re-using the density could make this reasonably
   efficient if elec_method=DM is used. (I assume that wave_reassign
   would not help in this case??)

Keith
-------------------------------------------------------------------------
I don't think this is quite what we want as this will fix the grids, which 
isn't the same as if the user performed a series of single-point energy 
runs. However I agree with the spirit of the suggestions.


> task : kpoint_test
[snip]
>   (I assume that wave_reassign would not help in this case??)

Not yet, but wave_copy would.

Phil

-------------------------------------------------------------------------
Sounds like a sensible suggestion for v4.5  - put it on the "feature 
request" list. It is too late in the v4.4 development cycle to implement 
this properly now - there are issues with grids, etc. but these can all 
be fixed. In the meantime, I've uncommented the 'hack' I mentioned to 
Charles and made it devel_code dependent so that MS users can (sort of) 
get access to this functionality in v4.4. Hence adding
devel_code = FBSC_FORCES
to the .param file in v4.4 onwards will turn on forces & stresses during 
the FBSC calculation.

          Matt