problem in Abinit 2 Aymbo conversion

Forum for discussing issues arising with interfacing the Abinit code (KSS) file with Yambo.

Moderator: Conor Hogan

devsharma
Posts: 28
Joined: Wed Oct 14, 2009 5:43 pm

Re: problem in Abinit 2 Aymbo conversion

Post by devsharma » Sat Nov 07, 2009 6:23 pm

Dear myrta,
hi ,
1) i have used the Abinit version 5.8.3 and change the inputs to very low values , just to check, KSS is converting or not.
Yes it is working.
2) And, I thik the problem is not with the version 5.7.4 , as i reproduced the LiF tutorials.
3) One silly doubt, I produced my KSS file on one pc (one which the YAMBO is not installed ), then I copied KSS file (whole folder) to another PC, where i tried to convert KSS to YAMBO. Is the problem was due to that ??? Means, is it mandatory that both the ABINIT and YAMBO should be installed on the same PC??? As it may happen, that some libraries may conflict. Well , I am not sure about all this and just asking

4) when I tried to run
>..../yambo -F 01_init -J 01_init
I got the error
<---> [01] Job Setup
<---> [02] Input variables setup
<---> [02.01] Unit cells
<---> [02.02] Symmetries

[ERROR] STOP signal received while in :[02.02] Symmetries
[ERROR][SYMs] check the input symmetries!
[dev@dev Abinit]$

Please help here.

5) Is there any direct link , where I can search the earlier pasted problem directly . Directly means like we enter the keyword and get the related topics like google :D .

thanks
take care
Dev Sharma,
Junior Research Fellow
Department of Physics & Astrophysics,
University of Delhi,
India

User avatar
Conor Hogan
Posts: 111
Joined: Tue Mar 17, 2009 12:17 pm
Contact:

Re: problem in Abinit 2 Aymbo conversion

Post by Conor Hogan » Mon Nov 09, 2009 12:03 pm

devsharma wrote: 1) i have used the Abinit version 5.8.3 and change the inputs to very low values , just to check, KSS is converting or not.
Yes it is working.
2) And, I thik the problem is not with the version 5.7.4 , as i reproduced the LiF tutorials.
you should check your system then also with 5.7.4 with the same very low values, to make sure the problem is with 5.7.4. LiF is a very different system.
3) One silly doubt, I produced my KSS file on one pc (one which the YAMBO is not installed ), then I copied KSS file (whole folder) to another PC, where i tried to convert KSS to YAMBO. Is the problem was due to that ??? Means, is it mandatory that both the ABINIT and YAMBO should be installed on the same PC??? As it may happen, that some libraries may conflict. Well , I am not sure about all this and just asking
It depends: if the architectures are the same (you just say PC, so who knows), you should be able to copy the file, e.g. if you copy from 32 bit linux to 32 bit linux. But you should not be able to copy to IBM-AIX for instance, as binary data can be read/written differently by different compilers (look up things like "big-endian", and size of "word"s)


[ERROR] STOP signal received while in :[02.02] Symmetries
[ERROR][SYMs] check the input symmetries!
Well, what are the input symmetries? They should be written in the Abinit output file, or better, if you used netCDF, you could tell us how they are written in the Yambo database using ncdump. In a post to the Abinit forum I saw you reported very strange symmetries there, did you fix that problem?
5) Is there any direct link , where I can search the earlier pasted problem directly . Directly means like we enter the keyword and get the related topics like google :D .
The search bar at the top right of the forum should work...anyway, the forum is not so big at present!

Conor
Dr. Conor Hogan
CNR-ISM, via Fosso del Cavaliere, 00133 Roma, Italy;
Department of Physics and European Theoretical Spectroscopy Facility (ETSF),
University of Rome "Tor Vergata".

devsharma
Posts: 28
Joined: Wed Oct 14, 2009 5:43 pm

Re: problem in Abinit 2 Aymbo conversion

Post by devsharma » Mon Nov 09, 2009 6:13 pm

hi Conor sir,
regarding the symmetry question, I tried ncdump.
Sir you said they are written in the Yambo database.
I think the yambo database are in the created save folder ?? M i right ?? Like
[dev@dev SAVE]$ ls
s.db1 s.kb_pp s.wf
[dev@dev SAVE]$
here s.db1 is the yambo databse.

If i am right upto here, i read the ncdump use and tried
[dev@dev SAVE]$ /home/dev/netcdf-4.0.1/ncdump/ncdump -v symmetries s.db1
but i got error

/home/dev/netcdf-4.0.1/ncdump/ncdump: s.db1: NetCDF: Unknown file format

Please help here.
And , I had corrected my input file , pasted on the Abinit forum.
Also, attached here is the output files. the number of symmetries(nsym) is 32.
Please guide ,

Thanks
With regards
Dev
You do not have the required permissions to view the files attached to this post.
Dev Sharma,
Junior Research Fellow
Department of Physics & Astrophysics,
University of Delhi,
India

User avatar
Conor Hogan
Posts: 111
Joined: Tue Mar 17, 2009 12:17 pm
Contact:

Re: problem in Abinit 2 Aymbo conversion

Post by Conor Hogan » Mon Nov 09, 2009 6:57 pm

Dev, as I wrote:
They should be written in the Abinit output file, or better, if you used netCDF, you could tell us how they are written in the Yambo database using ncdump.
which explains why you get an error trying to use a netCDF program on a binary fortran file.
/home/dev/netcdf-4.0.1/ncdump/ncdump: s.db1: NetCDF: Unknown file format
netCDF format files are written y Yambo as "ns.db1".
Dr. Conor Hogan
CNR-ISM, via Fosso del Cavaliere, 00133 Roma, Italy;
Department of Physics and European Theoretical Spectroscopy Facility (ETSF),
University of Rome "Tor Vergata".

User avatar
myrta gruning
Posts: 241
Joined: Tue Mar 17, 2009 11:38 am
Contact:

Re: problem in Abinit 2 Aymbo conversion

Post by myrta gruning » Tue Nov 10, 2009 12:07 am

Dear Dev

My guess is that the problem you met with the conversion may be due to the fact you generated the KSS on a different computer. Lowering the parameters was just to spedd up the tests, but I don't think they are the problem.

The problem with the symmetry error is due to the fact that you considered non symmorfic symmetries. Only symmorfic symmetries are allowed by yambo.
Change symmorphi2 in your input to 0.

Note that as a consequence you will have more non-equivalent k-points in the irreducible brillouin zone, so you have to change istwfk2 accordingly (e.g. for the 4x4x4 grid is 59*1).

Regards,

m
Dr Myrta Grüning
School of Mathematics and Physics
Queen's University Belfast - Northern Ireland

http://www.researcherid.com/rid/B-1515-2009

devsharma
Posts: 28
Joined: Wed Oct 14, 2009 5:43 pm

Re: problem in Abinit 2 Aymbo conversion

Post by devsharma » Tue Nov 10, 2009 7:10 am

hi myrta mam,
yes, the symmetry problem was due to symmorphi2 option... i have solved it....... but one doubt mam.... how do we chose the value of istwfk2 ?? like istwfk is 59*1 for 4 4 4 grid ??? i serached the log file and read the abinit documentation also ,but didnt get something there which can clear this.
Thnaks a lot
Take care
dev
Dev Sharma,
Junior Research Fellow
Department of Physics & Astrophysics,
University of Delhi,
India

User avatar
Daniele Varsano
Posts: 4048
Joined: Tue Mar 17, 2009 2:23 pm
Contact:

Re: problem in Abinit 2 Aymbo conversion

Post by Daniele Varsano » Tue Nov 10, 2009 8:11 am

Dear Dev,
how do we chose the value of istwfk2 ??
The value of istwk2 is the number of the k-points you have in dataset 2,
multiplied for 1.

Cheers,

Daniele
Dr. Daniele Varsano
S3-CNR Institute of Nanoscience and MaX Center, Italy
MaX - Materials design at the Exascale
http://www.nano.cnr.it
http://www.max-centre.eu/

devsharma
Posts: 28
Joined: Wed Oct 14, 2009 5:43 pm

Re: problem in Abinit 2 Aymbo conversion

Post by devsharma » Tue Nov 10, 2009 10:10 am

hi Daniele sir,
Sir, if these are the number of k points used in dataset 2.
then it should be something like this
See, if i have the output of data set 2 (printed in log and output both), something like below (with any value of istwfk e.g, istwfk 30*1)
i got

DATASET 2 : the unit cell is not primitive
================================================================================
Values of the parameters that define the memory need for DATASET 2.
intxc = 0 ionmov = 0 iscf = -2 xclevel = 1
lmnmax = 4 lnmax = 4 mband = 150 mffmem = 1
P mgfft = 15 mkmem = 16 mpssoang= 4 mpw = 109
mqgrid = 3001 natom = 24 nfft = 2700 nkpt = 16
nloalg = 4 nspden = 1 nspinor = 1 nsppol = 1
nsym = 4 n1xccc = 0 ntypat = 3 occopt = 1

it means that istwk2 should be 16*1 in my input file,
m i right ??
thanks
dev
Dev Sharma,
Junior Research Fellow
Department of Physics & Astrophysics,
University of Delhi,
India

User avatar
andrea marini
Posts: 325
Joined: Mon Mar 16, 2009 4:27 pm
Contact:

Re: problem in Abinit 2 Aymbo conversion

Post by andrea marini » Tue Nov 10, 2009 10:22 am

Dear Dev, I perfectly understand that as a new user of Yambo you may have a ton of questions. However in this forum we cannot help you with abinit related problems. So I kindly ask you to get a little bit more familiar with the GW related features of Abinit by running the 3 tutorials of abinit on GW, KSS and so on

GW1
GW2
GW3

And, please, use the Abinit mailing list to ask any abinit-related question. They will be much more helpfull than any of us.
Andrea MARINI
Istituto di Struttura della Materia, CNR, (Italy)

User avatar
myrta gruning
Posts: 241
Joined: Tue Mar 17, 2009 11:38 am
Contact:

Re: problem in Abinit 2 Aymbo conversion

Post by myrta gruning » Tue Nov 10, 2009 10:30 am

Dear Dev,

You are right, it should be 16*1. I was too quick in scanning the output file and I picked up the wrong number.

Regards,

Myrta
Dr Myrta Grüning
School of Mathematics and Physics
Queen's University Belfast - Northern Ireland

http://www.researcherid.com/rid/B-1515-2009

Post Reply