rim analysing

[nrbsh]

Dear yambo users and developers

I would like to use RIM analyzer.
What is the difference of following the the process explained in http://www.yambo-code.org/input_file/ypp/ypp_rim.php
and using yambo -c input file?

Thanks,
Zara
Tehran University

[Daniele Varsano]

Dear Zara,
The method described in the link, which use the ypp (yambo post processing) it is a trick to speed up the convergence of optical spectra by averaging on the energy of random k-points (without the need to calculate the corresponding wave function).

The RIM, random integration method (ypp -c) or (-r for the 4.0 version) is explained here and also in the yambo paper and it is needed to better evaluate Coulomb integral in the IBZ. This is somehow mandatory for non 3D sampling of the BZ.

Best,
Daniele

[nrbsh]

Dear Daniele,
Thank you very much for your quick and helpful response.

Using yambo-4.0.2-rev.90, I setup coulomb cutoff for BS calculation in the 2D system (first I found static dielectric function using yambo -b and then BS calculation using yambo -c -o b -k sex -y d).

I have ~ 40 bohr vacuum in z direction and found that if I get z-cutoff ~ 19 bohr my gap is better that z-cutoff~35-38.

Now my question is about the hidden factor 2, mentioned it in the forum for the older versions of yambo package.
Does it fix in the yambo-4.0.2?

The RIM parameters in my input file:
rim_cut # [R RIM CUT] Coulomb potential
RandQpts=1500000 # [RIM] Number of random q-points in the BZ
RandGvec= 11 RL # [RIM] Coulomb interaction RS components
CUTGeo= "box z" # [CUT] Coulomb Cutoff geometry: box/cylinder/sphere X/Y/Z/XY..
% CUTBox
0.00 | 0.00 | 19.00 | # [CUT] [au] Box sides
%


Best regards,
Zara
Tehran Univrsity

[Daniele Varsano]

Dear Zara,
The factor 2 should be there, unless someone changed it, I still need to fix it, and I'll do asap.

(first I found static dielectric function using yambo -b and then BS calculation using yambo -c -o b -k sex -y d).

Anyway I would also use the cutoff potential in the calculation of the screening:

Code: Select all

>yambo -c -b 

If you got a better agreement with the experiment could be hazard. You can check that your calculation should be "stable" varying the zcut a little bit around 35-38, while should vary much more around 19 bohr.

Best,
Daniele

[shaldar]

Dear Daniele,

Can you please let me know if the factor of 2 is still there in the recent version ( GPL Version 4.0.2 Revision 96)

Regards

Soumyajyoti

[Daniele Varsano]

Dear Soumyajyoti,
yes it is, and I think I will keep as it is. So the correct way to use the box is to set the x/y/z_cut just s bit smaller than dimension of your cell.

Best,
Daniele

[shaldar]

Dear Daniele,

Thanks for the clarification

Regards
Soumyajyoti