BSE spectra considering momentum

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

pangrt
Posts: 26
Joined: Thu Jun 18, 2020 1:38 pm

Re: BSE spectra considering momentum

Post by pangrt » Wed Mar 17, 2021 10:54 am

Dear Daniele,
Thanks for reminding.

Thanks!
Pangrt
RongTianPang, Inner Mongolia University, Hohhot, China :)

pangrt
Posts: 26
Joined: Thu Jun 18, 2020 1:38 pm

Re: BSE spectra considering momentum

Post by pangrt » Thu Apr 22, 2021 2:54 am

Dear Daniele,
I just tried to use the 5.0.1 version to calculate the example and found that the Maximum Residual Value was still around 10^10. And I found that when I use 5.0.0 to calculate the BSE and calculate the exciton property, there will be an error

Code: Select all

[ERROR] STOP signal received while in[06] Excitonic Properties @ Q-index #3
[ERROR] Reading File ./SAVE//ndb.BS_diago_Q3; Variable PARS; NetCDF: Numeric conversion not representable.
Is the output format changed after the update, or is there something wrong with the command?

Thanks!
Pangrt
RongTianPang, Inner Mongolia University, Hohhot, China :)

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

Re: BSE spectra considering momentum

Post by Daniele Varsano » Thu Apr 22, 2021 8:10 am

Dear Pangrt,
thanks for reporting,
indeed, it seems that the correct normalisation of the residual has not percolated. I will fix it soon.
About the format, please update to the 5.0.1, the syntax to sort excitation energies is now:

Code: Select all

ypp -excitons s -BSiq 2
or

Code: Select all

ypp -e s -b 2
Thanks,

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/

pangrt
Posts: 26
Joined: Thu Jun 18, 2020 1:38 pm

Re: BSE spectra considering momentum

Post by pangrt » Fri Apr 23, 2021 11:04 am

Dear Daniele,
I used yambo-5.0.1/bin/ypp -e s -b 2ypp -e s -b 2, and still got the error

Code: Select all

 <---> [01] MPI/OPENMP structure, Files & I/O Directories
 <---> [02] Y(ambo) P(ost)/(re) P(rocessor)
 <---> [03] Core DB
 <---> :: Electrons             : 72.00000
 <---> :: Temperature           : 0.000000 [eV]
 <---> :: Lattice factors       :  6.04860   5.23824  51.02261 [a.u.]
 <---> :: K points              : 102
 <---> :: Bands                 :  400
 <---> :: Symmetries            : 12
 <---> :: RL vectors            : 156219
 <---> [04] K-point grid
 <---> :: Q-points   (IBZ): 102
 <---> :: X K-points (IBZ): 102
 <---> [05] CORE Variables Setup
 <---> [05.01] Unit cells
 <---> [05.02] Symmetries
 <---> [05.03] Reciprocal space
 <---> [05.04] K-grid lattice
 <---> Grid dimensions      :   32   32
 <---> [05.05] Energies & Occupations
 <---> [06] Excitonic Properties @ Q-index #2
[ERROR] STOP signal received while in[06] Excitonic Properties @ Q-index #2
[ERROR] Reading File ./SAVE//ndb.BS_diago_Q2; Variable PARS; NetCDF: Numeric conversion not representable
How can i solve this problem
Thanks!

Pangrt
RongTianPang, Inner Mongolia University, Hohhot, China :)

User avatar
Davide Sangalli
Posts: 610
Joined: Tue May 29, 2012 4:49 pm
Location: Via Salaria Km 29.3, CP 10, 00016, Monterotondo Stazione, Italy
Contact:

Re: BSE spectra considering momentum

Post by Davide Sangalli » Mon Apr 26, 2021 8:56 am

Dear Pangrt,
did you compile in double precision, i.e. with --enable-dp ?
Davide Sangalli, PhD
CNR-ISM, Division of Ultrafast Processes in Materials (FLASHit) and MaX Centre
https://sites.google.com/view/davidesangalli
http://www.max-centre.eu/

pangrt
Posts: 26
Joined: Thu Jun 18, 2020 1:38 pm

Re: BSE spectra considering momentum

Post by pangrt » Wed Apr 28, 2021 1:38 am

Dear Daniele,
I did not use double precision compilation.

Thanks!
Pangrt
RongTianPang, Inner Mongolia University, Hohhot, China :)

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

Re: BSE spectra considering momentum

Post by Daniele Varsano » Wed Apr 28, 2021 12:43 pm

Dear Pangrt,

if not otto large can you post your file:

./SAVE//ndb.BS_diago_Q2

or better, upload somewhere where we can have a look?

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/

pangrt
Posts: 26
Joined: Thu Jun 18, 2020 1:38 pm

Re: BSE spectra considering momentum

Post by pangrt » Thu Apr 29, 2021 8:34 am

Dear Daniele,
My ndb.BS_diago_Q2 file size is 2GB, which makes it difficult to upload.

Thanks!
Pangrt
RongTianPang, Inner Mongolia University, Hohhot, China :)

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

Re: BSE spectra considering momentum

Post by Daniele Varsano » Thu Apr 29, 2021 8:45 am

Dear Pangrt,

can you inspect your database (e.g. using ncdump) and see if you notice anything strange?
BTW the BSE calculation ended correctly and the spectrum you obtained was it reasonable?

can you post here the results of:

Code: Select all

ncdump -v PARS ndb.BS_diago_Q3


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/

pangrt
Posts: 26
Joined: Thu Jun 18, 2020 1:38 pm

Re: BSE spectra considering momentum

Post by pangrt » Fri Apr 30, 2021 9:38 am

Dear Daniele,
By using the ncdump -v PARS ndb.BS_diago_Q3 command. The result is

Code: Select all

netcdf ndb {
dimensions:
	D_0000000003 = 3 ;
	D_0000000001 = 1 ;
	D_0000000002 = 2 ;
	D_0000000100 = 100 ;
	D_0000000035 = 35 ;
	D_0000000005 = 5 ;
	D_0000016384 = 16384 ;
variables:
	float HEAD_VERSION(D_0000000003) ;
	float HEAD_REVISION(D_0000000001) ;
	float SERIAL_NUMBER(D_0000000001) ;
	float SPIN_VARS(D_0000000002) ;
	char K_energies_xc_KIND(D_0000000001, D_0000000100) ;
	float PARS1(D_0000000035) ;
	char W_Cutoff(D_0000000001, D_0000000100) ;
	char W_TDDFT_KERNEL(D_0000000001, D_0000000100) ;
	int W_DbGd(D_0000000002) ;
	byte PERT_SOC(D_0000000001) ;
	float PARS(D_0000000005) ;
	float BS_Energies(D_0000016384, D_0000000002) ;
	float BS_left_Residuals(D_0000016384, D_0000000002) ;
	float BS_right_Residuals(D_0000016384, D_0000000002) ;
	float BS_TABLE(D_0000000003, D_0000016384) ;
	float BS_EIGENSTATES(D_0000016384, D_0000016384, D_0000000002) ;

// global attributes:
		:_NCProperties = "version=2,netcdf=4.7.4,hdf5=1.12.0," ;
data:

 PARS = 1, 16384, _, _, _ ;
}
Thanks!
Pangrt
RongTianPang, Inner Mongolia University, Hohhot, China :)

Locked