EPW Forum

Posted: **Fri Oct 14, 2016 10:29 pm**

Dear EPW Users,

I have installed qe-6.0/EPW-4.1.0. When restarting with an arbitrary number of cpus from the epmatwp1 file with the following settings:

Code: Select all

epbwrite    = .false.
epbread     = .false.
kmaps       = .true.
epwwrite    =  .false.
epwread     =  .true.
wannierize  = .false.

The run continues but some pieces of information are missing. For example, for diamond, nsym=48 but in the restart run, it is reset to 1. wkf(ikk) also returns inconsistent values with the filkf file. What would be the best way to restore that information in the restart?

One way would be to reread the input file and do the symmetry analysis in the SUBROUTINE "epw_setup_restart" but this may be the longer way.

I would appreciate any feedback on this matter.

Thank you,

Vahid

Vahid Askarpour
Department of Physics and Atmospheric Science
Dalhousie University,
Halifax, NS, Canada

Posted: **Sat Oct 15, 2016 2:05 pm**

Dear Vahid,

As written in the code when you do a restart

Code: Select all

     ------------------------------------------------------------------------
                   RESTART - RESTART - RESTART - RESTART
     Restart is done without reading PWSCF save file.
     Be aware that some consistency checks are therefore not done.
     ------------------------------------------------------------------------

all the crystals and symmetry informations are not computed/read from file.

However, some informations required for the fine grid interpolations are stored on file.
The new file is call "crystal.fmt". Symmetries are not stored in that file since symmetry is not used at the moment during the fine grid interpolation.

wkf should however be correct. What do you mean by "inconsistent"?

Best,

Samuel

Posted: **Sat Oct 15, 2016 3:52 pm**

Dear Samuel,

When I add the following lines in the ephwann_shuffle.f90, I get wkf values different from the one in the filkf file:

Code: Select all

DO ik = 1, nkqtotf/2
   ikk = 2 * ik - 1
   IF (mpime .eq. ionode_id) THEN
       write(*,*) wkf(ikk)
   ENDIF
ENDDO

For symmetry inclusion in the restart, I guess I need to reread the epw.in file and redo the symmetry analysis.

Thank you for your input.

Vahid

Posted: **Sat Oct 15, 2016 5:29 pm**

you got me scared ....

I made a quick test and it works for me.

Using the following filkf file:

Code: Select all

5 crystal
0.0 0.0 0.0 0.1
0.1 0.2 0.2 0.2
0.2 0.2 0.3 0.3
0.5 0.5 0.5 0.1
0.0 0.2 0.2 0.3

I get with the full calculation:

Code: Select all

 iq            1
 ik            1
 wkf(ikk)   0.20000000000000001
 iq            1
 ik            2
 wkf(ikk)   0.40000000000000002
 iq            1
 ik            3
 wkf(ikk)   0.59999999999999998
 iq            1
 ik            4
 wkf(ikk)   0.20000000000000001
 iq            1
 ik            5
 wkf(ikk)   0.59999999999999998

There is off course a factor 2 (see the EPW paper for more info).

Now if I run a restart calculation I get:

Code: Select all

 iq            1
 ik            1
 wkf(ikk)   0.20000000000000001
 iq            1
 ik            2
 wkf(ikk)   0.40000000000000002
 iq            1
 ik            3
 wkf(ikk)   0.59999999999999998
 iq            1
 ik            4
 wkf(ikk)   0.20000000000000001
 iq            1
 ik            5
 wkf(ikk)   0.59999999999999998

Best,

Samuel

Posted: **Sat Oct 15, 2016 5:51 pm**

Dear Samuel,

My apologies. I did not consider the factor of 2.
I am sorry about the scare.

Vahid

EPW Forum

Missing information after restart

Missing information after restart

Re: Missing information after restart

Re: Missing information after restart

Re: Missing information after restart

Re: Missing information after restart