Dear sir/mam
Greetings for the day !
(1). I am using quantum espresso(QE) to calculate band decomposed charge density. To my knowledge, the kpoint at which we want to calculate the charge density(rho) must be an integer(because whenever i try to calculate rho using kpoint as real numbers e.g 1.1, 1.002, i got error reading in kpoint ) please enlighten me if i am wrong. More importantly please suggest me how can i calculate rho for all the kpoint values( integer as well as fractional).
(2). When i calculate rho for a range of kpoints { let kpoint(1)=1, kpoint(2)=3 } with output file format suitable for Xcrysden type. Three different output files rho1.xsf, rho2.xsf, rho3.xsf for 3 different kpoints 1, 2, and 3 respectively were generated . so how can i merge these files to generate a single output file containing rho for all the kpoints ?
Any suggestion will be helpful for me.
Dear Mousam,
(1) If I understood your question correctly, you are confusing actual kpoints coordinates and kpoint index. The kpoint used in the input file for pp.x is represents 'kpoint index' (i.e. it specifies whether you want to use the first, second or the third kpoint and so on ...) and not actual kpoints. So yes, it has to be an integer. The actual kpoint coordinates are contained in the output/input file for the non-scf calculation (or band structure calculation.). So for example if you want to plot charge density for a certain kpoint coordinate (say 0.0 0.5 0.0), first check its index from your non scf calculation and then use it as kpoint index in pp.x calculation.
(2) QE manual for pp.x briefly describes how to combine two or more files.
NAMELIST &INPUTPP
containing the variables for step (1), followed by
NAMELIST &PLOT
containing the variables for step (2)
The two steps can be performed independently. In order to perform only step (2), leave namelist &INPUTPP blank. In order to perform only step (1), do not specify namelist &PLOT
Intermediate results from step 1 can be saved to disk (see variable filplot in &INPUTPP) and later read in step 2. Since the file with intermediate results is formatted, it can be safely transferred to a different machine. This also allows plotting of a linear combination (for instance, charge differences) by saving two intermediate files and combining them (see variables weight and filepp in &PLOT)
Hope that helps.
Abhishek
Dear Abhishek,
Thanks for your valuable reply. Now i have a better hold on this. Thanks again.
- Badges
- Science topic
- Similar topics
- Filing