X grid is not uniform

[chinaye]

Dear Developers,
We do not know how to solve the error " [RL indx] X grid is not uniform. Gamma point only. " The ground-state calculation was carried out by using Abinit-5.7.2. Thank you very much.

[andrea marini]

We do not know how to solve the error " [RL indx] X grid is not uniform. Gamma point only. " The ground-state calculation was carried out by using Abinit-5.7.2. Thank you very much.

Dear Jianfei Ye,

we cannot help using the little information you provide. Please use the attachments to post

• The input file (and pseudo files) of abinit
• the input/report files generated by Yambo

Cheers

[chinaye]

Dear Jianfei Ye,

we cannot help using the little information you provide. Please use the attachments to post

• The input file (and pseudo files) of abinit
• the input/report files generated by Yambo

Cheers

Dear andrea marini,
The attachment, files. rar, includes the input files of abinit and files generated by Yambo. Thank you very much.

[myrta gruning]

Dear Jianfei Ye,

the error you found it is due to the fact that Yambo needs a uniform grid. The grid generated by your abinit input is not uniform.
The shifts (shiftk and nshiftk input variables in ABINIT) have to be chosen properly in order to have a uniform grid. See http://www.abinit.org/documentation/hel ... tml#shiftk.
Your choice

Code: Select all

nshiftk2 4
shiftk2  0.5 0.5 0.5  # These shifts will be the same for all grids
         0.5 0.0 0.0  
         0.0 0.5 0.0
         0.0 0.0 0.5

is fine just for an FCC but it produces a non uniform grid in your case.
By running with

Code: Select all

nshiftk2 1
shiftk2  0.0 0.0 0.0 #NO SHIFT

or

Code: Select all

nshiftk2 1
shiftk2  0.0 0.0 0.5 #

I avoided the problem you met.
Note that also the default shiftk (the one that is chosen if you not specify shiftk in your input), 0.5 0.5 0.5, will generate a non uniform grid!!!

I do not know if there are still other shiftk's that give a uniform grid that may be "better" from a numerical point of view, but with the two here you should be able to run your job.

Cheers,
m

[chinaye]

Dear Myrta Grüning,
thank you very much. Your answer helps me a lot.