How to merge QP database with app

[jasonhan0710]

Dear Yambo developers,

I have noticed that we can merge the el-ph QP database with the one of ppa. First, I got the el-ph QP database as the tutorials http://www.yambo-code.org/wiki/index.ph ... n_Coupling said. Then I get the ppa QP database in another directory with the same k grid and band index used in el-ph QP calculation. After that, I tried to generate the input file with the command "ypp -qpdb m", and changed the input file as

Code: Select all

#
#       Version 5.0.4 Revision 19595 Hash 896bffc02
#                        Branch is
#               MPI+OpenMP+SLK+HDF5_IO Build
#                http://www.yambo-code.org
#
QPDBs                            # [R] Quasiparticle Databases
QPDB_merge                       # [R] Mergering
%Actions_and_names               # [QPDB] Format is "what"|"OP"|"prefactor"|"DB"|. OP can be +/-/x(only for Z)
"E" |"+" |"1" |"./EH/ndb.QP" |
"E" |"+" |"1" |"../yambo/T300/ndb.QP" |
%

However, after I run ypp, there is a warning in the report file, which reports

Code: Select all

 <---> Merging...
 <---> [WARNING]Differenent number of descriptions in the QP DBs

 <---> QP DBs with different  "GW approximation"
 <---> Different values:  Fan + DW  != PPA
 <--->detected 1 family(ies) of QP databases...
 <---> [WARNING]Differenent number of descriptions in the QP DBs

 <---> QP DBs with different  "GW approximation"
 <---> Different values:  Fan + DW  != PPA
 <--->done
 <---> I/O...done
 <---> [06] Timing Overview
 <---> [07] Memory Overview
 <---> [08] Game Over & Game summary

And, when I used the new QP database to get the absorption spectrum, it gave a total different result with the one scissor operation gave. As shown in the figure (see the attachment). And the BSE calculation warns that the system become a metallic system. The report files of QP merge and BSE calculation are also attached.

Is it a right way to merge the QP database?

Best,
Jason

[Daniele Varsano]

Dear Jason,
The syntax of the ypp input is correct.
I've not much experience on that, but it seems that something went wrong: what you can do is to check the resulting QP energies in the BSE report file after the merged QP has been read: they are reported as "Eqp @ K", these should be compared with the KS energies and the QP corrections you calculated previously reported in the corresponding reports and outputs.

Best,
Daniele

[jasonhan0710]

Dear Daniele,

Thank you for your reply.

I will check the merged QP database manually.

Best,
Jason