Question on Yambo5.10 tutorial

You can find here problems arising when using old releases of Yambo (< 5.0). Issues as parallelization strategy, performance issues and other technical aspects.

Moderators: Davide Sangalli, andrea.ferretti, myrta gruning, andrea marini, Daniele Varsano, Conor Hogan

Locked
lysea82
Posts: 2
Joined: Tue May 31, 2022 7:11 am

Question on Yambo5.10 tutorial

Post by lysea82 » Sat Jun 04, 2022 4:15 am

Dear Yambo community

I compiled Yambo5.1.0 on my Centos7 workstation with Intel ONEAPI 2021.3.0. I am now following the tutorial on calculation of Si band gaps (http://www.yambo-code.org/wiki/index.php?title=Silicon)

I could repeat the Hatree Fock section, but failed to repeat the "COHSEX without empty bands" secton. In the "W-size convergence" subsection, no matter what the "NGsBlkXs" was, I always got the same "[Hatree-Fock] Direct Gap", which was about 8.320753 eV. This result was quite different from the one of the tutorial.

And I noticed in the "parse_gap.sh" script that, the band gap was defined as

Code: Select all

......
   VBMo=`grep " 1.000" $i/References/$1 | grep " 4.0000" | grep " 0.000" | awk '{print $3}'`
   VBM=`grep " 1.000" $i/References/$1 | grep " 4.0000" | grep " 0.000" | awk '{print $4}'`
   CBMo=`grep " 2.576" $i/References/$1 | grep " 5.000" | awk '{print $3}'`
   CBM=`grep " 2.576" $i/References/$1 | grep " 5.000" | awk '{print $4}'`
   case "$1" in 
    *-01HF*)
      echo  ${NK[$j]} $VBMo $VBM $CBMo $CBM | awk '{print $1 " " $5-$3 }' >>  "$2"
      ;;
    *)
      echo  ${NK[$j]} $VBMo $VBM $CBMo $CBM | awk '{print $1 " " $4+$5 }' >>  "$2"
      ;;
   esac
......
Should the band gap be $4+$5-($2+$3)?

Thanks!
Last edited by lysea82 on Fri Jun 10, 2022 12:10 pm, edited 1 time in total.
Dr. Ying Liu
School of Information Sci. Tech., Fudan University
220 Handan Road
Shanghai, China

User avatar
Daniele Varsano
Posts: 3773
Joined: Tue Mar 17, 2009 2:23 pm
Contact:

Re: Question on Yambo5.10 tutorial

Post by Daniele Varsano » Mon Jun 06, 2022 5:35 pm

Dear lysea82,

please sign your post with your name and affiliation, this is a rule of the forum and you can add your signature once for all in your user profile.
no matter what the "NGsBlkXs" was, I always got the same "[Hatree-Fock] Direct Gap", which was about 8.320753 eV.
Please note that NGsBlkXs do not affect the HF gap, which does not depend on the size of the dielectric matrix.

About the parse script, please note they are very old, it is possible that there are mistakes or that they are not suitable anymore with the new format output and should be corrected. Just pay attention that HF and COhsex output it is possible they do not contain the same information (Eqp and correction) anyway this is reported in the legend.

Best,

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/

lysea82
Posts: 2
Joined: Tue May 31, 2022 7:11 am

Re: Question on Yambo5.10 tutorial

Post by lysea82 » Fri Jun 10, 2022 12:13 pm

Dear Daniele

Thank you very much for your replies. Also I added my signature to the post.

Regards
Dr. Ying Liu
School of Information Sci. Tech., Fudan University
220 Handan Road
Shanghai, China

Locked