Bethe-Salpeter solver: SLEPC - Revision history

Fulvio: /* Links */

2023-05-19T16:59:12Z

Links

← Older revision		Revision as of 16:59, 19 May 2023
Line 127:		Line 127:
	* Alternative Bethe-Salpeter equation solver: [[Bethe-Salpeter solver: diagonalization \| diagonalization]]		* Alternative Bethe-Salpeter equation solver: [[Bethe-Salpeter solver: diagonalization \| diagonalization]]
	* Alternative Bethe-Salpeter equation solver: [[Bethe-Salpeter solver: Lanczos-Haydock\|Haydock]]		* Alternative Bethe-Salpeter equation solver: [[Bethe-Salpeter solver: Lanczos-Haydock\|Haydock]]
			* Back to [[Rome 2023#Tutorials]]
	* Back to [[Calculating optical spectra including excitonic effects: a step-by-step guide]] tutorial		* Back to [[Calculating optical spectra including excitonic effects: a step-by-step guide]] tutorial
	* [[Tutorials\|Back to tutorials menu]]		* [[Tutorials\|Back to tutorials menu]]

Fulvio: /* Choice of input parameters */

2021-04-03T14:13:14Z

Choice of input parameters

← Older revision		Revision as of 14:13, 3 April 2021
Line 55:		Line 55:

	Therefore we are trying to describe just the dominant feature in the spectrum.		Therefore we are trying to describe just the dominant feature in the spectrum.

			There is also an additional parameter defining the maximum number of iterations:
			BSSSlepcMaxIt=0
			If we keep this equal to zero, the code will use an internally defined value.

	==Bethe-Salpeter solver runlevel==		==Bethe-Salpeter solver runlevel==

Fulvio: /* Summary */

2021-04-01T18:26:47Z

Summary

← Older revision		Revision as of 18:26, 1 April 2021
Line 116:		Line 116:
	From this tutorial you've learned:		From this tutorial you've learned:
	* How to compute the optical spectrum by using the SLEPc solver within the Bethe-Salpeter equation framework		* How to compute the optical spectrum by using the SLEPc solver within the Bethe-Salpeter equation framework
	* How to correctly describe a subset of the excitonic eigenvalues and eigenvectors without diagonalizing the full ~~excitonic~~ Hamiltonian.		* How to correctly describe a subset of the excitonic eigenvalues and eigenvectors without diagonalizing the full two-particle Hamiltonian.

	==Links==		==Links==

Fulvio: /* Including more peaks */

2021-04-01T18:25:39Z

Including more peaks

← Older revision		Revision as of 18:25, 1 April 2021
Line 108:		Line 108:
	...		...
	plot 'o-3D_BSE.eps_q1_diago_bse' u 1:2 w l lc rgb 'black' t 'Diagonalization', 'o-3D_BSE-NEig4.eps_q1_slepc_bse' u 1:2 w l lw 2 lc rgb 'blue' t 'Slepc 4 states', 'o-3D_BSE-NEig13.eps_q1_slepc_bse' u 1:2 w l lw 2 lc rgb 'red' t 'Slepc 13 states', 'o-3D_BSE-NEig21.eps_q1_slepc_bse' u 1:2 w l lw 2 lc rgb 'orange' t 'Slepc 21 states'		plot 'o-3D_BSE.eps_q1_diago_bse' u 1:2 w l lc rgb 'black' t 'Diagonalization', 'o-3D_BSE-NEig4.eps_q1_slepc_bse' u 1:2 w l lw 2 lc rgb 'blue' t 'Slepc 4 states', 'o-3D_BSE-NEig13.eps_q1_slepc_bse' u 1:2 w l lw 2 lc rgb 'red' t 'Slepc 13 states', 'o-3D_BSE-NEig21.eps_q1_slepc_bse' u 1:2 w l lw 2 lc rgb 'orange' t 'Slepc 21 states'
	[[~~file~~:~~03_bse_haydock_high~~.png\|none\|600px]]		[[File:03 bse slepc 21.png\|none\|600px]]
	We see that we obtain an accurate description of the relevant part of the absorption spectrum of bulk hBN by just including 21 states. The diagonalization solver needs 432 states instead, because this is the size of the two-particle Hamiltonian in our calculation (keep in mind that this is far from being numerically converged, and the correct absorption spectrum looks quite different).		We see that we obtain an accurate description of the relevant part of the absorption spectrum of bulk hBN by just including 21 states. The diagonalization solver needs 432 states instead, because this is the size of the two-particle Hamiltonian in our calculation (keep in mind that this is far from being numerically converged, and the correct absorption spectrum looks quite different).

Fulvio: /* Bethe-Salpeter solver runlevel */

2021-04-01T18:24:46Z

Bethe-Salpeter solver runlevel

← Older revision		Revision as of 18:24, 1 April 2021
Line 89:		Line 89:
	...		...
	plot 'o-3D_BSE.eps_q1_diago_bse' u 1:2 w l lc rgb 'black' t 'Diagonalization', 'o-3D_BSE-NEig4.eps_q1_slepc_bse' u 1:2 w l lw 2 lc rgb 'blue' t 'Slepc 4 states'		plot 'o-3D_BSE.eps_q1_diago_bse' u 1:2 w l lc rgb 'black' t 'Diagonalization', 'o-3D_BSE-NEig4.eps_q1_slepc_bse' u 1:2 w l lw 2 lc rgb 'blue' t 'Slepc 4 states'
	[[~~file~~:~~03_bse_haydock~~.png\|none\|600px]]		[[File:03 bse slepc 4.png\|none\|600px]]
	As expected, by requesting the first 4 states we are able to reproduce the first and strongest absorption peak.		As expected, by requesting the first 4 states we are able to reproduce the first and strongest absorption peak.

Fulvio: /* Prerequisites */

2021-04-01T18:23:20Z

Prerequisites

← Older revision		Revision as of 18:23, 1 April 2021
Line 2:		Line 2:

	==Prerequisites==		==Prerequisites==
	[[File:Yambo-handbook-~~v4.1.2~~-p-16.png\|thumb\|Cheatsheet on SLEPc\|150px]]		[[File:Yambo-handbook-v5-p-21.png\|thumb\|Cheatsheet on SLEPc\|150px]]
	* You must first complete the [[Static screening]] and [[Bethe-Salpeter kernel]] modules in the [[Calculating optical spectra including excitonic effects: a step-by-step guide\|step-by-step guide]]		* You must first complete the [[Static screening]] and [[Bethe-Salpeter kernel]] modules in the [[Calculating optical spectra including excitonic effects: a step-by-step guide\|step-by-step guide]]
	* Optionally, in order to compare results with the diagonalization solver, you may first complete the [[Bethe-Salpeter solver: diagonalization]] module.		* Optionally, in order to compare results with the diagonalization solver, you may first complete the [[Bethe-Salpeter solver: diagonalization]] module.

Fulvio at 18:17, 1 April 2021

2021-04-01T18:17:45Z

← Older revision		Revision as of 18:17, 1 April 2021
Line 1:		Line 1:
	In this module you learn how to obtain the optical absorption spectrum of bulk BN within the Bethe-Salpeter equation (BSE) framework for a previously calculated Bethe-Salpeter kernel. You will learn how to by-pass the diagonalization of the full two-particle Hamiltonian and only obtain eigenvalues and eigenvectors for a selected number of states, using the ~~SLEPc solver.~~[https://slepc.upv.es SLEPc ~~library~~]<ref>V. Hernandez, J.E. Roman and V. Vidal in '' ACM Transactions on Mathematical Software'', '''31''' 315 (2005)</ref>		In this module you learn how to obtain the optical absorption spectrum of bulk BN within the Bethe-Salpeter equation (BSE) framework for a previously calculated Bethe-Salpeter kernel. You will learn how to by-pass the diagonalization of the full two-particle Hamiltonian and only obtain eigenvalues and eigenvectors for a selected number of states, using the [https://slepc.upv.es SLEPc] solver.<ref>V. Hernandez, J.E. Roman and V. Vidal in '' ACM Transactions on Mathematical Software'', '''31''' 315 (2005)</ref>

	==Prerequisites==		==Prerequisites==

Fulvio at 18:16, 1 April 2021

2021-04-01T18:16:42Z

← Older revision		Revision as of 18:16, 1 April 2021
Line 1:		Line 1:
	In this module you learn how to obtain the optical absorption spectrum of bulk BN within the Bethe-Salpeter equation (BSE) framework for a previously calculated Bethe-Salpeter kernel. You will learn how to by-pass the diagonalization of the full two-particle Hamiltonian and only obtain eigenvalues and eigenvectors for a selected number of states, using the SLEPc solver.		In this module you learn how to obtain the optical absorption spectrum of bulk BN within the Bethe-Salpeter equation (BSE) framework for a previously calculated Bethe-Salpeter kernel. You will learn how to by-pass the diagonalization of the full two-particle Hamiltonian and only obtain eigenvalues and eigenvectors for a selected number of states, using the SLEPc solver.[https://slepc.upv.es SLEPc library]<ref>V. Hernandez, J.E. Roman and V. Vidal in '' ACM Transactions on Mathematical Software'', '''31''' 315 (2005)</ref>

	==Prerequisites==		==Prerequisites==
Line 126:		Line 126:
	* [[Tutorials\|Back to tutorials menu]]		* [[Tutorials\|Back to tutorials menu]]
	* [[Modules\|Back to technical modules menu]]		* [[Modules\|Back to technical modules menu]]

			==References==

Fulvio: /* Summary */

2021-04-01T18:14:40Z

Summary

← Older revision		Revision as of 18:14, 1 April 2021
Line 116:		Line 116:
	From this tutorial you've learned:		From this tutorial you've learned:
	* How to compute the optical spectrum by using the SLEPc solver within the Bethe-Salpeter equation framework		* How to compute the optical spectrum by using the SLEPc solver within the Bethe-Salpeter equation framework
	* How to correctly describe a ~~section~~ of the excitonic eigenvalues and eigenvectors without diagonalizing the full excitonic Hamiltonian.		* How to correctly describe a subset of the excitonic eigenvalues and eigenvectors without diagonalizing the full excitonic Hamiltonian.

	==Links==		==Links==

Fulvio: /* Including more peaks */

2021-04-01T18:13:43Z

Including more peaks

← Older revision		Revision as of 18:13, 1 April 2021
Line 109:		Line 109:
	plot 'o-3D_BSE.eps_q1_diago_bse' u 1:2 w l lc rgb 'black' t 'Diagonalization', 'o-3D_BSE-NEig4.eps_q1_slepc_bse' u 1:2 w l lw 2 lc rgb 'blue' t 'Slepc 4 states', 'o-3D_BSE-NEig13.eps_q1_slepc_bse' u 1:2 w l lw 2 lc rgb 'red' t 'Slepc 13 states', 'o-3D_BSE-NEig21.eps_q1_slepc_bse' u 1:2 w l lw 2 lc rgb 'orange' t 'Slepc 21 states'		plot 'o-3D_BSE.eps_q1_diago_bse' u 1:2 w l lc rgb 'black' t 'Diagonalization', 'o-3D_BSE-NEig4.eps_q1_slepc_bse' u 1:2 w l lw 2 lc rgb 'blue' t 'Slepc 4 states', 'o-3D_BSE-NEig13.eps_q1_slepc_bse' u 1:2 w l lw 2 lc rgb 'red' t 'Slepc 13 states', 'o-3D_BSE-NEig21.eps_q1_slepc_bse' u 1:2 w l lw 2 lc rgb 'orange' t 'Slepc 21 states'
	[[file:03_bse_haydock_high.png\|none\|600px]]		[[file:03_bse_haydock_high.png\|none\|600px]]
	We see that we obtain an accurate description of the relevant part of the absorption spectrum of bulk hBN by just including 21 states. The diagonalization solver ~~instead includes~~ 432 states because this is the size of the ~~excitonic~~ Hamiltonian in our calculation (keep in mind that this is far from being numerically converged, and the correct absorption spectrum looks quite different).		We see that we obtain an accurate description of the relevant part of the absorption spectrum of bulk hBN by just including 21 states. The diagonalization solver needs 432 states instead, because this is the size of the two-particle Hamiltonian in our calculation (keep in mind that this is far from being numerically converged, and the correct absorption spectrum looks quite different).

	In a real calculation for a large system, when full diagonalization is not an option, you will typically first obtain the full absorption spectrum using the Lanczos-Haydock solver, then switch to the SLEPc ~~for~~ the exciton states relevant to ~~you~~, while using the previously obtained full spectrum as a reference as we did in this tutorial.		In a real calculation for a large system, when full diagonalization is not an option, you will typically first obtain the full absorption spectrum using the Lanczos-Haydock solver, then switch to the SLEPc to focus on the exciton states relevant to your analysis, while using the previously obtained full spectrum as a reference as we did in this tutorial.

	==Summary==		==Summary==