The ABINIT Users Mailing List ( CLOSED )

[Masayoshi Mikami <mmikami@yk.rim.or.jp>]

Hello,

Although I replied to the original post this morning,
I forgot to add "forum@abinit.org" to the "To:" line...

About the description about the elphon, 
please refer to Matthieu's post.

Best wishes,
Masayoshi

Begin forwarded message:
---------------------------------------------------------------------------
From: Masayoshi Mikami <mmikami@rc.m-kagaku.co.jp>
Date: 2004.3.15 11:38:53:JST
To: "Corsin Battaglia" <corsin.battaglia@freesurf.ch>
Subject: Re: [abinit-forum] (fuzzy) RPA susceptibility

Dear Corsin,
I would have comment from user's viewpoint, 
"how to dig in the source archive..."

1) about sus* (e.g. susk)
If you get interested in "susk"
http://www.abinit.org/ABINIT/robodoc_4.0.5/susk.f.html#susk
you might want to refer to "PARENTS" (subroutines that call 
the present subroutine),
and follow the hyperlinks in PARENTS again, again and again.

Then you have a call map:
        susk <- suscep_stat <- suscep <- driver
then, by looking around in the driver.f,
------
elseif(optdriver==2)then
      call status(jdtset_status,filstat,iexit,level,'call suscep   ')
       call suscep(dtfil,dtset,iexit,&
    &   mband,mgfft,mkmem,mpw,natom,nfft,nkpt,&
    &   nspden,nspinor,nsppol,nsym,occ,xred)
------
the key input appears "optdriver 2" as you noticed.

Then, it will be a good idea to try "grep -i optdriver" for Test_*/*
as well as Info/*. You will find the relevant test cases, Test_v2/t57,
Test_v3/t43, Test_v3/t44 (if you look into  v.4.0.5.: for the other 
versions, v.4.1/2/3.x, other test cases can be found. 
So please try "grep optdriver" for Test_*/* in each version.)

In passing,
these test cases can be noticed by "grep -i suscep Test_*/*" 
as well ...
So, it would be a good idea to go through all the Test_* cases
as well as Tutorials on your own  machine.
(It would also help to check the portability/robustness of ABINIT. Thanks !)

2) about elphon.f
As done in 1), you might want to check the PARENTS. 
The anaddb calls this.
Then you will notice in the anaddb code :
-----
!The electron-phonon section is under development !!!!!
     if (anaddb_dtset%elphflag == 1) then
      unitgkk = 90
      open (unit=unitgkk,file='Al_fcc_GKK.bin',&
    &       form='unformatted',status='old')
    !DEBUG
    !write(6,*)' anaddb : before call to elphon ',elphflag
    !  stop
    !ENDDEBUG
       call elphon(acell,amu,atmfrc,blkflg,blkqpt,blknrm,blktyp,blkval,&
                (snip)
-----
Actually, v.4.3.2. is still for development, not for production.
So you might want to wait for the forthcoming v.4.3.x suitable 
for production...

Good continuation & Happy ("abinit") life,
Masayoshi

> On 2004/03/14, at 22:40, Corsin Battaglia wrote:
>> Dear abinit users
  
 Yesterday, I discovered the abinit master index, 
 http://www.abinit.org/ABINIT/robodoc_4.0.2/ABINIT_mi.html.
 It contains the code for the functions suscep, suscep_stat, susk etc. 
 I am interested in calculating these, especially susk. The entry for 
 optdriver in the input variable list of v4.0 mentions
  optdriver=2: susceptibility calculation.
 Until now I have been relying on the abinit tutorials and of course 
 the mailing list. But I can not find any information on how to use 
 the suscep routine. Could anybody tell me, where I can find 
 some hints or recommand me a possibly well-commented input file? 
 Following Masayoshi Mikami's tip, I tried grep -i sus Infos/*, but 
 grep just outputs the information contained in the variable list file 
 keyhr.htm.
  
 Furthmore I found this very cool elphon function. 
 http://www.abinit.org/ABINIT/robodoc_4.3.1/elphon.f.html
 But I have the same problem as with the sus family.
  
 Maybe these *.f files in the master index should be self-explanatory. 
 Not yet for me. So maybe somebody could give me some clues on how to 
 extract the information to get the codes running.
  
 Thanks in advance for your support.
  
 Corsin
---------------------------------------------------------------------------