About lifetime

Concerns issues with computing quasiparticle corrections to the DFT eigenvalues - i.e., the self-energy within the GW approximation (-g n), or considering the Hartree-Fock exchange only (-x)

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

Post Reply
lesheng
Posts: 26
Joined: Thu Feb 07, 2013 11:39 pm

About lifetime

Post by lesheng » Tue May 27, 2014 9:42 pm

Hi all,

I am now trying to calculate the QP lifetime of a system, and I noticed that there is a keyword "LifeTrCG" in the input file. My question is how can I interpret this variable, and will this value affect the calculation time significantly?

Thanks for your attention,

Lesheng Li
PhD candidate
209 South Rd, Chapel Hill, NC 27599
Department of Chemistry
The University of North Carolina at Chapel Hill
Email: leshengl@live.unc.edu

User avatar
Davide Sangalli
Posts: 614
Joined: Tue May 29, 2012 4:49 pm
Location: Via Salaria Km 29.3, CP 10, 00016, Monterotondo Stazione, Italy
Contact:

Re: About lifetime

Post by Davide Sangalli » Tue May 27, 2014 10:28 pm

Dear Lesheng Li,
the varable controls how the poles of the response function are described.

To compute the lifetimes yambo performes a GW calculation solving the frequency integral on the real axis. The response function enters in the integral.
By default poles are grouped together only if they are exactly degenerate in energy: LifeTrCG=100. (%)

If you reduce the value of the variable the poles will be grouped also if they are not exacty degenerate. In practice the variables ccontrols how much the code is sensitive to the energy dfference between two poles. 100% is the maximun sensitivity the code has. Reducing it, you could speed up your calculation but you should not to reduce it too much ...

Best,
Davide
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/

lesheng
Posts: 26
Joined: Thu Feb 07, 2013 11:39 pm

Re: About lifetime

Post by lesheng » Tue May 27, 2014 10:42 pm

Davide Sangalli wrote:Dear Lesheng Li,
the varable controls how the poles of the response function are described.

To compute the lifetimes yambo performes a GW calculation solving the frequency integral on the real axis. The response function enters in the integral.
By default poles are grouped together only if they are exactly degenerate in energy: LifeTrCG=100. (%)

If you reduce the value of the variable the poles will be grouped also if they are not exacty degenerate. In practice the variables ccontrols how much the code is sensitive to the energy dfference between two poles. 100% is the maximun sensitivity the code has. Reducing it, you could speed up your calculation but you should not to reduce it too much ...

Best,
Davide

Hi Davide,

Thanks for your reply. I know this is a silly question, but how should I know which point will be the best one... I mean in the tutorial it uses 30, is there any specific reason to set this value as 30?

Plus, I was trying to calculate it with 30, but it shows me an error as following:

<03s> [01] Files & I/O Directories
<04s> [02] CORE Variables Setup
<04s> [02.01] Unit cells
<08s> [02.02] Symmetries
<08s> [02.03] RL shells
<08s> [02.04] K-grid lattice
<08s> [02.05] Energies [ev] & Occupations
<08s> [03] Transferred momenta grid
<08s> [03.01] G0W0 on the real axis
<08s> [03.02] Lifetimes Transitions Selector
<08s> [Life-CG] R(p) Tot o/o(of R) : 16368 317515 29
<08s> [03.03] Dynamical Dielectric Matrix
[ERROR] STOP signal received while in :[03.03] Dynamical Dielectric Matrix
[ERROR]Mem All. failed. Element X require 9.00073 [Gb]

Do you have any suggestion on this?

Thanks,
Lesheng Li
PhD candidate
209 South Rd, Chapel Hill, NC 27599
Department of Chemistry
The University of North Carolina at Chapel Hill
Email: leshengl@live.unc.edu

Post Reply