[DFTB-Plus-User] TDDFTB excitation window and OS filter

[Fahri Alkan]

Hello all,
I am using DFTB+ to calculate the absorption spectra for very large
systems. For such systems, one often needs to employ energy window (EW) and
oscillator strength (OS) filter approximations to reduce the size of
one-particle transition space. However, I believe there is a bug in the
code when you invoke both of these filters. For a test system (Ag55
cluster), SPX.DAT shows 58286 transitions without any filter. When I apply
10^-3 OS filter alone, this number reduces to 14260. When I apply 0.1 au EW
alone (with 100 Nr of excitations), one-particle transition space reduces
to only 1759. If I use them together, what I expect reading from the manual
is that one-particle transition space should reduce from 1759 to further
some number as a result of OS filter. However, SPX.DAT shows 15549
excitations! (Even larger than the both numbers I get from filters alone).
I am using the compiled binary version of DFTB-19.1. I haven't encounter
any other problems with the program. My excitation block in the input is as
follows:
ExcitedState = Casida {
 NrOfExcitations=100
 Symmetry=singlet
 WriteTransitions=no
 WriteSPTransitions=yes
 WriteMulliken=no
 WriteCoefficients=no
 WriteEigenvectors=no
 WriteTransitionDipole=no
 OscillatorWindow = 0.00100000000000000
 EnergyWindow = 0.10000000000000000
}
Is this a bug? Any comments on this will be appropriated greatly.
Thanks

-- 

Asst. Prof. Fahri Alkan

Department of Material Science & Nanotechnology Engineering
Abdullah Gül University
Kayseri, Turkey
Office: BA102
Email: fahri.alkan at agu.edu.tr
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.zfn.uni-bremen.de/pipermail/dftb-plus-user/attachments/20191025/6a22dcc8/attachment.html>