Locating iku on BZ

[sitangshu]

Dear Yambo guys,

I have a basic question: How to locate internal k points (iku) on a Brillouin zone?

For example: I have this snippet and I want to locate what point on the BZ electron dispersion this Q11 can be found.

[03] Transferred momenta grid
=============================
IBZ Q-points : 19
BZ Q-points : 144

Q [00011] : 0.08333 0.37500 0.00000 (iku) * weight 0.08333

With regards,
Sitangshu

[claudio]

Dear Sitangshu

the iku unit is an internal units of Yambo.

I advise you to use the cc units that means Cartesian Coordinates (CC):

0.000000 0.256363 0.000000 (cc)

in these units the q-points are in bohr-radius-1

let me know if it helps

best
Claudio

[Daniele Varsano]

Dear Sitangshu,

I do agree with Claudio,
anyway, you can also use ypp utility:

Code: Select all

ypp -k q  

and there you can set input q point in iku and have the output in rlu or cc

Best,
Daniele

[sitangshu]

Dear Professor Claudio,

Thank you for your quick reply. I analyzed an exciton transition, and saw the iku listed above o be a main contribution. Strangely, this is not in the high symmetry points. So, I thought to see the transition in the electron band structure. Thus I suppose "rlu" might be more favorable option...

So, I also followed Professor Daniele's suggestion ypp -k q option and put in input:

cooIn= "iku" # Points coordinates (in) cc/rlu/iku/alat
cooOut= "rlu" # Points coordinates (out) cc/rlu/iku/alat
%Qpts # Q points list
0.08333 |0.37500 |0.00000 |

The run gives multiple values with a warning... What I am missing? Are these supposed to be same listed in o.bands_interpolated file?

<---> [06.01] == Built-in IBZ K-grid [rlu] ==
0.000000000 0.000000000 0.000000000 1.000000000
0.000000000 0.083333328 0.000000000 6.000000000
0.000000000 0.166666657 0.000000000 6.000000000
0.000000000 0.250000000 0.000000000 6.000000000
0.000000000 0.333333313 0.000000000 6.000000000
0.000000000 0.416666687 0.000000000 6.000000000
0.000000000 -0.500000000 0.000000000 3.000000000
0.083333343 0.083333328 0.000000000 6.000000000
0.083333343 0.166666672 0.000000000 12.000000000
0.083333343 0.249999985 0.000000000 12.000000000
0.083333343 0.333333313 0.000000000 12.000000000
0.083333343 0.416666687 0.000000000 12.000000000
0.166666687 0.166666657 0.000000000 6.000000000
0.166666687 0.249999970 0.000000000 12.000000000
0.166666687 0.333333343 0.000000000 12.000000000
0.166666687 0.416666657 0.000000000 6.000000000
0.250000000 0.249999970 0.000000000 6.000000000
0.250000000 0.333333343 0.000000000 12.000000000
0.333333373 0.333333313 0.000000000 2.000000000
<---> [06.02] == Q-grid analyzer ==
<---> [WARNING] Point already in the grid. Listing internal Q-points.

Thank you and regards,
Sitangshu

[Daniele Varsano]

Dear Sitangshu,
the warining is harmless, it just stell you that the point you indicated in input is present in your gris (as expected).
In the output, you have your grid printed in rlu where you should recognize the k/q point you are looking at.

Please note that in the BSE you are including k points from a regular grid and not a path along with high symmetry points.

Best,

Daniele

[claudio]

Dear Sitangshu,

consider also yambopy to analyze the transitions that compose an exciton
and plot them along the band structure:

http://www.yambo-code.org/wiki/index.ph ... ocal_space

best
Claudio

[sitangshu]

Dear Professor Claudio,

Many thanks for the link.... I went through the coding and obtained the less-decent looking plot. This is the best I can do by optimizing the BZ size, etc. But it seems that the exciton weights and "qpts" are less in numbers. Probably this would be the reason why less dots are here in the plot...?
PS: I did a BSE mapping 12 12 1 ---> 72 72 1 DG.



Regards,
Sitangshu

[amolina]

Dear Sitangshu,

I guess yambopy is plotting the exciton weigths in the coarse grid of 12x12x1. You have also the option of plotting the weigths in the band structure, using the interpolation tool.

Cheers,
Alejandro.

[sitangshu]

Dear Alejandro,

Thanks for your reply. I also thought so that weights are plotted on the coarse grid. I suppose the code has to be modified in some sense so that it takes the results from double-grid?

Regards,
Sitangshu

[haseebphysics1]

Dear Developers,

It has been a useful thread to read. I was able to convert k-point using ypp. However, I have a generic question.

Q: Yambo outputs some K-points in negative values too, like -0.499999881 -0.499999970 0.000000000 (taken in fractional/rlu). Since I had calculated bandstructure with the only +k (along the k-path of BZ), so can I neglect these negative values in the coordinates of reciprocal space? and if time-reversal symmetry holds and can be seen in the report file? or will there be any other change also to compare the bands plot with the Yambo k-points?

Many thanks,
Haseeb