Need of Sponsoring for the blog

Friday, 9 September 2016

The full version of the convergence tests package

We present you the full version of the convergence tests package.

This package allows us to do the optimization of the following parameters:

- K-point,
- RKmax,
- Gmax,
- Rmts

NB: The package is still under test and improvement and any one find a bug he contacts us.

To download the package click on the link below:


Description:

The package include 4 parts for the optimization of the 4 parametrs and every part include 3 scripts.

  • Script for setup                           parameter_setup_lapw
  • Script for initialization and SCF     parameter_opt_lapw
  • Script for analyzing and plotting   parameter_ana_lapw
 Parameter=kpt, rkmax, gmax or rmt

Installation: You have to uninstall the previous scripts

Put the extracted directory in the home repertory (for example), 

To install the package execute the following command:
============
      username@machinename:~/home$     gunzip test_conv.tar.gz
      username@machinename:~/home$     tar xvf   test_conv.tar
      username@machinename:~/home$     cd test_conv/
      username@machinename:~/home/test_conv$   ./install_all

To uninstall the package type the following command:
============
username@machinename:~/home/test_conv$   ./Uninstall_all

Execution:

The K-point optimization.
==========================

Execution:
========
 - Create a directory for your material(you will find the example of MgO in the repertory Example )
 - Put the struct file in the created repertory
 - Execute the command

    username@machinename:~/MgO$    kpt_setup_lapw     # for setup

    username@machinename:~/MgO$     kpt_opt_lapw     # intialization  and Scf calculation

    username@machinename:~/MgO$     kpt_ana_lapw      # analysis and plotting E=f(kpoint)

Note:
===============================================================================================
- if the energy is not converged the script will exit by default after 15 iterations (cycle).

- you can change the value of the error of energy in the file  difenergy  during the execution of
  the script  kpt_opt_lapw, because the script take the value each cycle then the script do
  the convergence test.

- After analysis the results , if you want more point (or otherwise improve the result),
  run directly the command : kpt_opt_lapw   without initialization,
  in this case the script will be start at the last end point .
 
-Before run  kpt_opt_lapw ,  you can :
 Modify option of  init_lapw  by edit     kpt_init_lapw  file
   (by :" emac kpt_init_lapw "  or by "gdit kpt_init_lapw ")
 Modify option of  run_lapw   by edit     kpt_run_lapw  file
   The SCF run by the default command: run_lapw -ec 0.0001  -i 40 )
 Modify the  convergence error by edit  difenergy   file

 -SCF calculation are saved in directory  SCF_kpt_"number_of_input_Kpoint"
  for each kpoint
==============================================================================================
Note: You can check the video to see the execution




The rKmax optimization.
==========================

Execution:
========
 - Create a directory for your material(you will find the example of MgO in the repertory Example )
 - Put the struct file in the created repertory
 - Execute the command

    username@machinename:~/MgO$    rkmax_setup     # for setup

    username@machinename:~/MgO$     rkmax_opt      # intialization  and Scf calculation

    username@machinename:~/MgO$     rkmax_ana     # analysis and plotting E=f(rkmax)

Note:
===============================================================================================
-  the script will exit if the energy is converged or the rkmax value is equal 10.5 

- you can change the value of the error of energy in the file  difenergy  during the execution of
  the script  rkmax_opt_lapw, because the script take the value each cycle then the script do
  the convergence test.

- After analysis the results , if you want more point (or otherwise improve the result),
  run directly the command : rkmax_opt   without initialization,
 
  in this case the script will be start at the last end point (if rkmax< 10.5)  .
 
-Before run  rkmax_opt_lapw ,  you can :

 Modify option of  init_lapw  by edit     rkmax_init_lapw  file
   (by :" emac rkmax_init_lapw "  or by "gdit rkmax_init_lapw ")
 Modify option of  run_lapw   by edit     rkmax_run_lapw  file
   The SCF run by the default command: run_lapw -ec 0.0001  -i 40 )
 Modify the  convergence error by edit  difenergy   file

 -SCF calculation are saved in directory  SCF_rkmax_"number_of_input_rkmax"
  for each rkmax
==============================================================================================

The Gmax optimization.
==========================

Execution:
========
 - Create a directory for your material(you will find the example of MgO in the repertory Example )
 - Put the struct file in the created repertory
 - Execute the command

    username@machinename:~/MgO$    gmax_setup     # setup

    username@machinename:~/MgO$     gmax_opt      # intialization  and Scf calculation

    username@machinename:~/MgO$     gmax_ana     # analysis and plotting E=f(gmax)

Note:
===============================================================================================
-  the script will exit if the energy is converged or the gmax value is equal 20.  .

- you can change the value of the error of energy in the file  difenergy  during the execution of
  the script  gmax_opt_lapw, because the script take the value each cycle then the script do
  the convergence test.

- After analysis the results , if you want more point (or otherwise improve the result),
  run directly the command : gmax_opt   without initialization,
 
  in this case the script will be start at the last end point (if gmax< 20.5)  .
 
-Before run  gmax_opt_lapw ,  you can :

 Modify option of  init_lapw_silent  by edit     gmax_init_lapw  file
       (by :" emac gmax_init_lapw "  or by "gdit gmax_init_lapw ")
       for help : init_lapw_silent -h
      
 Modify option of  run_lapw   by edit     gmax_run_lapw  file
   The SCF run by the default command: run_lapw -ec 0.0001  -i 40 )
 Modify the  convergence error by edit  difenergy   file

 -SCF calculation are saved in directory  SCF_gmax_"number_of_input_gmax"
  for each gmax
==============================================================================================

The RMT optimization.
==========================

Execution:
========
 - Create a directory for your material(you will find the example of MgO in the repertory Example )
 - Put the struct file in the created repertory
 - Execute the command

    username@machinename:~/MgO$    rmt_setup     # setup

    username@machinename:~/MgO$     rmt_opt      # intialization  and Scf calculation

    username@machinename:~/MgO$     rmt_ana     # analysis and plotting E=f(rmt)

Note:
===============================================================================================
-  the script will be exit after 8 cycles, you can run again rmt_opt_lapw for more cycle (X8)  .

-  After analysis the results , if you want more point (or otherwise improve the result),
  run directly the command : rmt_opt   without initialization(rmt_setup),
 
  in this case the script will be start at the last  point .  .
 
-Before run  rmt_opt_lapw ,  you can :

 Modify option of  init_lapw_silent  by edit     rmt_init_lapw  file
       (by :" emac rmt_init_lapw "  or by "gdit rmt_init_lapw ")
       for help : init_lapw_silent -h
      
 Modify option of  run_lapw   by edit     rmt_run_lapw  file
   The SCF run by the default command: run_lapw -ec 0.0001  -i 40 )


 -SCF calculation are saved in directory  SCF_rmt_"number_of_cycle"
  for each rmt
==============================================================================================
Note: You can check the video to see the execution



General Note
============
For more points, you can run directly the command line: 'XXX'_opt_lapw 
in this case, the calculation will be started from the last point, the new calculation
will be added to the old (update ).

 if you run  the command line   'XXX'_setup_lapw  the old calculation will be
 formated and the calculation will be started from the first point

'XXX' mean the parameter (kpt  rkmax  gmax rmt )

7 comments:

  1. Hii,
    Many thanks for this script.
    What is the proper way to optimize these parameters?
    Should we first optimze for k-points and then use these optimised k-points for RKmax and then use this RKmax for Gmax optimization and so on/

    Sincerely
    Bhamu

    ReplyDelete
  2. This comment has been removed by the author.

    ReplyDelete
  3. I cannot find the script. All I see is an image of MgO convergence test from the 4shared link. Please help

    ReplyDelete
    Replies
    1. another link https://drive.google.com/file/d/0B-QmTukdEFfQSDFqeG4tLTFzQW8/view

      Delete
    2. When I write ./install_all I get this error message : set: Syntax Error.
      the same thing as for the above ./install_c , please help

      Delete
  4. many thanks for this program, but I think there is a problem when launching the command rmt_opt_lapw

    ReplyDelete