Page 1 of 1

Weight of excitation

Posted: Thu Dec 19, 2019 2:19 am
by xtbu
Dear developers,
I am wondering about the wieght in ypp output inducing from ypp -e a as:

Code: Select all

#    Band_V       Band_C       K  ibz       Symm.        Weight       Energy
#    
     32.0000      33.0000     12.0000       6.0000       0.1080       1.4859
     32.0000      33.0000     12.0000       5.0000       0.1080       1.4859
     32.0000      33.0000     12.0000       4.0000       0.1080       1.4859
     31.00        34.00        12.0        6.000       0.6421E-1     1.506   
     31.00        34.00        12.0        5.000       0.6419E-1     1.506   
     31.00        34.00        12.0        4.000       0.6419E-1     1.506
If I want to konw the contributions of the trasition, like 32-33 as above, I need to count the contribution 0.108x3 or just 0.108?
An other thing is the total weight are not equal 1, far less than 1. Why?

Thanks!

xt Bu

Re: Weight of excitation

Posted: Thu Dec 19, 2019 3:18 pm
by Davide Sangalli
Dear Xiangtian Bu,
you should count 0.108x3

The total weight is not 1 because, in the output are reported only the weights of the most relevant transitions.
There is a default threshold at 5% of the total weight. You can lower the threshold via the variable "Weight_treshold"

Best,
D.

Re: Weight of excitation

Posted: Thu Dec 19, 2019 6:28 pm
by Daniele Varsano
Dear Xiangtian Bu,

the variable controlling the threshold of the transition printed in the output is "MinWeight" that will be substituted with "Weight_treshold" in the next release.
Best,
Daniele

Re: Weight of excitation

Posted: Fri Dec 20, 2019 9:36 am
by claudio
Dear Xiangtian Bu,

probably the name Weight is not so correct, it should be better to call it "percentage contribution" to the exciton

best
Claudio

Re: Weight of excitation

Posted: Mon Apr 27, 2020 10:38 am
by xtbu
Thanks, Davide, Daniele and Claudio,
So is the weight stands for |A_eh|^2?Where the A is the prefactor of |S>?
Thanks!

Xiangtian

Re: Weight of excitation

Posted: Mon Apr 27, 2020 11:22 am
by Daniele Varsano
Dear Xiangtian,

yes, the weight is defined as you indicated.

Best,
Daniele

Re: Weight of excitation

Posted: Thu May 21, 2020 10:39 am
by xtbu
Dear all,
I got the exciton dipoles as described here"http://www.yambo-code.org/wiki/index.ph ... e_excitons".
And it is |A_eh<e|D|h>|, and whose square is proportional to oscillator strength. And I obtained the dipoles through Residuals as decroced here as :
Res=8pi/(q^2\Omega Nq)*|BS_R|^2. But why the two method has a different value of dipoles? I think |A_eh<e|D|h>|^2 should be equal as |BS_R|^2 here.

Thanks!

Xiang tiang

Re: Weight of excitation

Posted: Thu May 21, 2020 11:07 am
by Daniele Varsano
Dear Xiang,

I'm not sure I've understood your question,
I think |A_eh<e|D|h>|^2 should be equal as |BS_R|^2 here.
This is correct, these are the weighted transition dipoles, and the residuals are defined as you say, by multiplying them by a prefactor.
Maybe can be useful to look at this thread.

Best,
Daniele

Re: Weight of excitation

Posted: Wed Aug 17, 2022 7:22 am
by pyadav
Dear team Yambo,

How can I get a rough estimate of the size/radius of the lowest energy exciton in any material after BSE calculation.

Thank you,
Pushpendra

Re: Weight of excitation

Posted: Wed Aug 17, 2022 10:03 am
by Daniele Varsano
Dear Pushpendra,

you can have an estimation by plotting the corresponding exictonic wave function, e.g. in 1D where the two other dimensions are averaged and then fit the envelope function with a hydrogenic wave function in the case it applies.

Best,
Daniele