The RDMFT CECAM presentation gives the value of 0.565 and says, "Value of α is fixed to 0.565 and all calculations for extended systems are performed with this FIXED value. The theory is ab-initio in this sense." The previous slide shows the power functional, while the graph with the quoted text shows the Müller functional, so maybe the quote refers to it.
Finally, in the current Elk version NiO example it's 0.65.
So, which is it?
Also, RDMFT keywords are not in the manual.
Thank you.
Best regards.
Andrew
Edit: In the code, rdmalpha is set to 0.565.
Last edit: Andrew Shyichuk 2021-09-25
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
The value of alpha for solids should be 0.656. This is the value used in Phys. Rev. Lett. 110, 116403 for Mott insulators.
Independently, A. Putaja and E. Räsänen, Phys. Rev. B 84, 035104, found this to be in their range of acceptable values by studying quantum dots.
The earlier value, which was confusingly 0.565, was our best fit to the correlation energy of the homogeneous electron gas. However, this performed poorly for real solids (no band gap for silicon, for example). The value of 0.578 was chosen for finite systems (by fitting to the correlation energies of small molecules).
I'm not sure why the default value in Elk was 0.565. I've now changed it to 0.656 -- thanks for pointing this out.
Regards,
Kay.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Dear Users,
The RDMFT paper by Sharma et al gives RDMFT power functional power parameter α of 0.7 or 0.65 (0.7 looks better from the paper).
Another paper gives the value of 0.578.
The RDMFT CECAM presentation gives the value of 0.565 and says, "Value of α is fixed to 0.565 and all calculations for extended systems are performed with this FIXED value. The theory is ab-initio in this sense." The previous slide shows the power functional, while the graph with the quoted text shows the Müller functional, so maybe the quote refers to it.
Finally, in the current Elk version NiO example it's 0.65.
So, which is it?
Also, RDMFT keywords are not in the manual.
Thank you.
Best regards.
Andrew
Edit: In the code, rdmalpha is set to 0.565.
Last edit: Andrew Shyichuk 2021-09-25
Hi Andrew,
The value of alpha for solids should be 0.656. This is the value used in Phys. Rev. Lett. 110, 116403 for Mott insulators.
Independently, A. Putaja and E. Räsänen, Phys. Rev. B 84, 035104, found this to be in their range of acceptable values by studying quantum dots.
The earlier value, which was confusingly 0.565, was our best fit to the correlation energy of the homogeneous electron gas. However, this performed poorly for real solids (no band gap for silicon, for example). The value of 0.578 was chosen for finite systems (by fitting to the correlation energies of small molecules).
I'm not sure why the default value in Elk was 0.565. I've now changed it to 0.656 -- thanks for pointing this out.
Regards,
Kay.
Dear Kay,
Thank you for clearing that out.
What about scaling? How does RDMFT calculation scale with system size?
Or, at least, what is the approximate ratio between SCF and RDMFT CPU time?
Best regards.
Andrii