lp:~albertog/siesta/trunk-elsi-dm

Branch merges

Propose for merging

1 branch proposed for merging into this one.

No branches dependent on this one.

Related source package recipes

No recipes using this branch. (?)

Create packaging recipe

Related snap packages

Related bugs

Related blueprints

701. By Alberto Garcia on 2019-09-02

Sync to trunk-778 (notably nc/soc wfs support)

700. By Alberto Garcia on 2019-06-09

Add extra ELSI parameters to fdf interface

* New fdf keywords for:
  -- Ill-conditioning checks
  -- PEXSI:
         number of poles
  number of mu-points to be treated in parallel
  inertia-tolerance

* Some interface function calls have been updated. At least ELSI v2.2.0
  is needed.

699. By Alberto Garcia on 2019-04-08

Sync to trunk-764

698. By Alberto Garcia on 2019-02-13

Sync to trunk-760: Fix allocation for Bessel orbitals

697. By Alberto Garcia on 2019-02-13

Sync to trunk-759

696. By Alberto Garcia on 2019-01-25

Include NTPoly documentation in the manual

695. By Alberto Garcia on 2019-01-15

Do not get the EDM during the SCF cycle + NTPoly support

* In compute_dm, the call to the ELSI ('get_dm') interface uses a
flag to disable the generation of the EDM matrix. In this case
only the DM is computed.

After the scf loop, in 'post_scf_work', a further call to the ELSI
interface requests (only) the EDM.

Note that this is achieved through a logical flag 'Get_EDM_Only',
which is not really appropriate. It would be better to have a
more explicit flag, such as:

           output_mode="DM|EDM"

The ELSI shutdown has been moved towards the end of 'siesta_forces',
and the 'Siesta_Worker' logic preserved in that routine (even though
it is not working elsewhere).

* NTpoly support has been added to the interface

(Thanks to Victor M. Garcia-Suarez and Victor Yu)

694. By Alberto Garcia on 2018-10-10

Wrap MPI check on number of procs for ELSI solver

693. By Alberto Garcia on 2018-09-15

Add Si quantum dot test case for ELSI

692. By Alberto Garcia on 2018-09-14

Fix bug in computation of Delta-V for PEXSI mu bracketing

A new array is needed to properly update the value of the
change in potential.

(+ added a declaration for 'date_stamp' in the auxiliary
code for version compatibility)