Electron-phonon interactions from Monte-Carlo sampling - Revision history

Karsai: /* Full MC sampling */

2025-06-05T11:35:55Z

Full MC sampling

← Older revision		Revision as of 11:35, 5 June 2025
Line 50:		Line 50:
	The tag {{TAG\|TEBEG}}=0 is also needed to choose the temperature at which the sampling is run.		The tag {{TAG\|TEBEG}}=0 is also needed to choose the temperature at which the sampling is run.

	Additionally the {{TAG\|PHON_LBOSE}} can be set ''.TRUE.'' or ''.FALSE.'' (default {{TAG\|PHON_LBOSE}}=''.TRUE.''), which selects Bose-Einstein or Maxwell-Boltzmann statistics, respectively.		Additionally, the {{TAG\|PHON_LBOSE}} can be set ''.TRUE.'' or ''.FALSE.'' (default {{TAG\|PHON_LBOSE}}=''.TRUE.''), which selects Bose-Einstein or Maxwell-Boltzmann statistics, respectively.

	A sample {{TAG\|INCAR}} file for 0 K looks like the following:		A sample {{TAG\|INCAR}} file for 0 K looks like the following:

Karsai: /* ZG configuration (one-shot sampling) */

2025-06-05T11:35:21Z

ZG configuration (one-shot sampling)

← Older revision		Revision as of 11:35, 5 June 2025
Line 25:		Line 25:
	== ZG configuration (one-shot sampling) ==		== ZG configuration (one-shot sampling) ==

	M. Zacharias and F. Giustino{{cite\|zacharias:prb:2016}} introduced a one-shot method (named ZG configuration after the authors). This method is an approximation to full MC sampling. It only uses a single distorted structure and hence it is computationally much cheaper than the full MC sampling. It retains an accuracy very close to the full MC sampling for converged ~~super cell~~ sizes. For example we showed that for the zero-point renormalization of the band gap the accuracy is within 5 meV between the ZG configurations and the full MC sampling{{cite\|karsai:njp:2018}}. Hence we suggest ~~to use~~ this method preferably, ~~where~~ convergence of the ~~super cell~~ size is hard to achieve or the 5 meV accuracy is enough.		M. Zacharias and F. Giustino{{cite\|zacharias:prb:2016}} introduced a one-shot method (named ZG configuration after the authors). This method is an approximation to full MC sampling. It only uses a single distorted structure and hence it is computationally much cheaper than the full MC sampling. It retains an accuracy very close to the full MC sampling for converged supercell sizes. For example, we showed that for the zero-point renormalization of the band gap, the accuracy is within 5 meV between the ZG configurations and the full MC sampling{{cite\|karsai:njp:2018}}. Hence we suggest using this method preferably, when convergence of the supercell size is hard to achieve or the 5 meV accuracy is enough.

	To select the ZG configuration {{TAG\|PHON_NSTRUCT}}=0 has to be set in the {{TAG\|INCAR}} file.		To select the ZG configuration {{TAG\|PHON_NSTRUCT}}=0 has to be set in the {{TAG\|INCAR}} file.

Karsai: /* ZG configuration (one-shot sampling) */

2025-06-05T11:33:49Z

ZG configuration (one-shot sampling)

← Older revision		Revision as of 11:33, 5 June 2025
Line 25:		Line 25:
	== ZG configuration (one-shot sampling) ==		== ZG configuration (one-shot sampling) ==

	M. Zacharias and F. Giustino{{cite\|zacharias:prb:2016}} introduced a one-shot method (named ZG configuration after the authors). This method is an approximation to full MC sampling. It only uses a single distorted structure and hence it is ~~orders of magnitude faster~~ than the full MC sampling. It retains an accuracy very close to the full MC sampling for converged super cell sizes. For example we showed that for the zero-point renormalization of the band gap the accuracy is within 5 meV between the ZG configurations and the full MC sampling{{cite\|karsai:njp:2018}}. Hence we suggest to use this method preferably, where convergence of the super cell size is hard to achieve or the 5 meV accuracy is enough.		M. Zacharias and F. Giustino{{cite\|zacharias:prb:2016}} introduced a one-shot method (named ZG configuration after the authors). This method is an approximation to full MC sampling. It only uses a single distorted structure and hence it is computationally much cheaper than the full MC sampling. It retains an accuracy very close to the full MC sampling for converged super cell sizes. For example we showed that for the zero-point renormalization of the band gap the accuracy is within 5 meV between the ZG configurations and the full MC sampling{{cite\|karsai:njp:2018}}. Hence we suggest to use this method preferably, where convergence of the super cell size is hard to achieve or the 5 meV accuracy is enough.

	To select the ZG configuration {{TAG\|PHON_NSTRUCT}}=0 has to be set in the {{TAG\|INCAR}} file.		To select the ZG configuration {{TAG\|PHON_NSTRUCT}}=0 has to be set in the {{TAG\|INCAR}} file.

Karsai: /* Step-by-step instructions */

2025-06-05T11:32:29Z

Step-by-step instructions

← Older revision		Revision as of 11:32, 5 June 2025
Line 17:		Line 17:
	where NUMBER runs from 1 to {{TAG\|PHON_NSTRUCT}}.		where NUMBER runs from 1 to {{TAG\|PHON_NSTRUCT}}.

	'''Step 2''': Run calculation for the previously created {{TAG\|POSCAR}} files on the desired observable. These calculations can be anything that is suitable for a MC sum <math>\langle O(T)\rangle = \frac{1}{n} \sum\limits_{i=1}^{n} O(x_{T}^{\textrm{MC,i}})</math>, for example band gap calculations, absorption spectra calculations, etc.		'''Step 2''': Run calculation for the previously created {{TAG\|POSCAR}} files on the desired observable. These calculations can be anything that is suitable for an MC sum <math>\langle O(T)\rangle = \frac{1}{n} \sum\limits_{i=1}^{n} O(x_{T}^{\textrm{MC,i}})</math>, for example, band gap calculations, absorption spectra calculations, etc.

	'''Step 3 (optional)''': Calculate the desired observable for the original "pristine" supercell. This step can be necessary when changes to an observable due to electron-phonon interactions are required. An example of such a calculation is the calculation of the band-gap renormalization due to electron-phonon interactions.		'''Step 3 (optional)''': Calculate the desired observable for the original "pristine" supercell. This step can be necessary when changes to an observable due to electron-phonon interactions are required. An example of such a calculation is the calculation of the band-gap renormalization due to electron-phonon interactions.

Karsai at 11:31, 5 June 2025

2025-06-05T11:31:48Z

← Older revision		Revision as of 11:31, 5 June 2025
Line 1:		Line 1:
	{{NB\|mind\|This feature is only available from VASP 6.0 or higher.}}		{{NB\|mind\|This feature is only available from VASP 6.0 or higher.}}

	For the theory on electron-interactions from Monte-Carlo (MC) sampling please click [[Electron-phonon interactions theory\|here]].		For the theory on electron-phonon interactions from Monte-Carlo (MC) sampling please click [[Electron-phonon interactions theory\|here]].

	First of all this method needs a sufficiently large ~~super cell~~.		First of all this method needs a sufficiently large supercell.
	It also involves phonon calculations for the <math>\Gamma</math> point (see {{TAG\|Phonons from finite differences}}). So many tags in the {{TAG\|INCAR}} will be used from the phonon calculations.		It also involves phonon calculations for the <math>\Gamma</math> point (see {{TAG\|Phonons from finite differences}}). So many tags in the {{TAG\|INCAR}} will be used from the phonon calculations.

Karsai at 11:30, 5 June 2025

2025-06-05T11:30:33Z

Karsai: /* Step-by-step instructions */

2025-05-16T14:00:13Z

Step-by-step instructions

← Older revision		Revision as of 14:00, 16 May 2025
Line 11:		Line 11:

	== Step-by-step instructions ==		== Step-by-step instructions ==
	'''Step 1''': Run a single calculation to create the {{TAG\|POSCAR}} file(s) with special positions either belonging to the ZG configuration method or Monte Carlo sampling. To enable electron-phonon interactions from MC methods {{TAG\|PHON_LMC}}=''.TRUE.'' has to be set in the {{TAG\|INCAR}} file. Also {{TAG\|IBRION}}=6 has to be selected (the Monte-Carlo (MC) method is currently only implemented for {{TAG\|IBRION}}=6). The description of the remaining required {{TAG\|INCAR}} tags for each method is given below. Both methods produce {{TAG\|POSCAR}} files with different distorted Wycoff positions but unchanged Brillouin ~~matrixThe~~. The ZG configuration method produces one structure for each temperature defined in {{TAG\|PHON_TLIST}}. The files are labeled as		'''Step 1''': Run a single calculation to create the {{TAG\|POSCAR}} file(s) with special positions either belonging to the ZG configuration method or Monte Carlo sampling. To enable electron-phonon interactions from MC methods {{TAG\|PHON_LMC}}=''.TRUE.'' has to be set in the {{TAG\|INCAR}} file. Also, {{TAG\|IBRION}}=6 has to be selected (the Monte-Carlo (MC) method is currently only implemented for {{TAG\|IBRION}}=6). The description of the remaining required {{TAG\|INCAR}} tags for each method is given below. Both methods produce {{TAG\|POSCAR}} files with different distorted Wycoff positions but unchanged Brillouin matrix. The ZG configuration method produces one structure for each temperature defined in {{TAG\|PHON_TLIST}}. The files are labeled as
	POSCAR.T=TEMP.		POSCAR.T=TEMP.
	The MC sampling code produces many {{TAG\|POSCAR}} files at a given temperature defined by {{TAG\|TEBEG}}. The files are labeled as		The MC sampling code produces many {{TAG\|POSCAR}} files at a given temperature defined by {{TAG\|TEBEG}}. The files are labeled as
Line 19:		Line 19:
	'''Step 2''': Run calculation for the previously created {{TAG\|POSCAR}} files on the desired observable. These calculations can be anything that is suitable for a Monte Carlo sum <math>\langle O(T)\rangle = \frac{1}{n} \sum\limits_{i=1}^{n} O(x_{T}^{\textrm{MC,i}})</math>, for example band gap calculations, absorption spectra calculations, etc.		'''Step 2''': Run calculation for the previously created {{TAG\|POSCAR}} files on the desired observable. These calculations can be anything that is suitable for a Monte Carlo sum <math>\langle O(T)\rangle = \frac{1}{n} \sum\limits_{i=1}^{n} O(x_{T}^{\textrm{MC,i}})</math>, for example band gap calculations, absorption spectra calculations, etc.

	'''Step 3 (optional)''': Calculate the desired observable for the original "pristine" supercell. This step can be necessary when changes of an observable due to electron-phonon interactions is required. An example ~~for~~ such a calculation is the calculation of the band-gap renormalization due to electron-phonon interactions.		'''Step 3 (optional)''': Calculate the desired observable for the original "pristine" supercell. This step can be necessary when changes to an observable due to electron-phonon interactions are required. An example of such a calculation is the calculation of the band-gap renormalization due to electron-phonon interactions.

	'''Step 4 (optional)''': If Monte Carlo sampling was used average the results over the number of structures created in step 1 ({{TAG\|PHON_NSTRUCT}}).		'''Step 4 (optional)''': If Monte Carlo sampling was used, average the results over the number of structures created in step 1 ({{TAG\|PHON_NSTRUCT}}).

	== ZG configuration (one-shot sampling) ==		== ZG configuration (one-shot sampling) ==

Karsai at 13:39, 16 May 2025

2025-05-16T13:39:50Z

← Older revision		Revision as of 13:39, 16 May 2025
Line 1:		Line 1:
	{{NB\|mind\|This feature is only available from VASP 6.0 or higher.}}		{{NB\|mind\|This feature is only available from VASP 6.0 or higher.}}

	For the theory on electron-interactions from Monte-Carlo sampling please click ~~here~~ [[Electron-phonon interactions theory\|here]].		For the theory on electron-interactions from Monte-Carlo sampling please click [[Electron-phonon interactions theory\|here]].

	First of all this method needs a sufficiently large super cell.		First of all this method needs a sufficiently large super cell.

Karsai at 13:39, 16 May 2025

2025-05-16T13:39:41Z

← Older revision		Revision as of 13:39, 16 May 2025
Line 1:		Line 1:
	{{NB\|mind\|This feature is only available from VASP 6.0 or higher.}}		{{NB\|mind\|This feature is only available from VASP 6.0 or higher.}}

	For the theory on electron-interactions from Monte-Carlo sampling ~~see~~ [[Electron-phonon interactions theory\|here]].		For the theory on electron-interactions from Monte-Carlo sampling please click here [[Electron-phonon interactions theory\|here]].

	First of all this method needs a sufficiently large super cell.		First of all this method needs a sufficiently large super cell.

Karsai at 13:39, 16 May 2025

2025-05-16T13:39:19Z

← Older revision		Revision as of 13:39, 16 May 2025
Line 1:		Line 1:
	{{NB\|mind\|This feature is only available from VASP 6.0 or higher.}}		{{NB\|mind\|This feature is only available from VASP 6.0 or higher.}}

	For the theory on electron-interactions from Monte-Carlo sampling see [[~~here\|~~Electron-phonon interactions theory]].		For the theory on electron-interactions from Monte-Carlo sampling see [[Electron-phonon interactions theory\|here]].

	First of all this method needs a sufficiently large super cell.		First of all this method needs a sufficiently large super cell.

← Older revision		Revision as of 11:35, 5 June 2025
Line 50:		Line 50:
	The tag {{TAG\|TEBEG}}=0 is also needed to choose the temperature at which the sampling is run.		The tag {{TAG\|TEBEG}}=0 is also needed to choose the temperature at which the sampling is run.

	Additionally the {{TAG\|PHON_LBOSE}} can be set ''.TRUE.'' or ''.FALSE.'' (default {{TAG\|PHON_LBOSE}}=''.TRUE.''), which selects Bose-Einstein or Maxwell-Boltzmann statistics, respectively.		Additionally, the {{TAG\|PHON_LBOSE}} can be set ''.TRUE.'' or ''.FALSE.'' (default {{TAG\|PHON_LBOSE}}=''.TRUE.''), which selects Bose-Einstein or Maxwell-Boltzmann statistics, respectively.

	A sample {{TAG\|INCAR}} file for 0 K looks like the following:		A sample {{TAG\|INCAR}} file for 0 K looks like the following:

← Older revision		Revision as of 11:30, 5 June 2025
Line 1:		Line 1:
	{{NB\|mind\|This feature is only available from VASP 6.0 or higher.}}		{{NB\|mind\|This feature is only available from VASP 6.0 or higher.}}

	For the theory on electron-interactions from Monte-Carlo sampling please click [[Electron-phonon interactions theory\|here]].		For the theory on electron-interactions from Monte-Carlo (MC) sampling please click [[Electron-phonon interactions theory\|here]].

	First of all this method needs a sufficiently large super cell.		First of all this method needs a sufficiently large super cell.
Line 11:		Line 11:

	== Step-by-step instructions ==		== Step-by-step instructions ==
	'''Step 1''': Run a single calculation to create the {{TAG\|POSCAR}} file(s) with special positions either belonging to the ZG configuration method or ~~Monte Carlo~~ sampling. To enable electron-phonon interactions from MC methods {{TAG\|PHON_LMC}}=''.TRUE.'' has to be set in the {{TAG\|INCAR}} file. Also, {{TAG\|IBRION}}=6 has to be selected (the ~~Monte-Carlo (MC) method is~~ currently only implemented for {{TAG\|IBRION}}=6). The description of the remaining required {{TAG\|INCAR}} tags for each method is given below. Both methods produce {{TAG\|POSCAR}} files with different distorted Wycoff positions but unchanged Brillouin matrix. The ZG configuration method produces one structure for each temperature defined in {{TAG\|PHON_TLIST}}. The files are labeled as		'''Step 1''': Run a single calculation to create the {{TAG\|POSCAR}} file(s) with special positions either belonging to the ZG configuration method or MC sampling. To enable electron-phonon interactions from MC methods {{TAG\|PHON_LMC}}=''.TRUE.'' has to be set in the {{TAG\|INCAR}} file. Also, {{TAG\|IBRION}}=6 has to be selected (the sampling methods are currently only implemented for {{TAG\|IBRION}}=6). The description of the remaining required {{TAG\|INCAR}} tags for each method is given below. Both methods produce {{TAG\|POSCAR}} files with different distorted Wycoff positions but unchanged Brillouin matrix. The ZG configuration method produces one structure for each temperature defined in {{TAG\|PHON_TLIST}}. The files are labeled as
	POSCAR.T=TEMP.		POSCAR.T=TEMP.
	The MC sampling code produces many {{TAG\|POSCAR}} files at a given temperature defined by {{TAG\|TEBEG}}. The files are labeled as		The MC sampling code produces many {{TAG\|POSCAR}} files at a given temperature defined by {{TAG\|TEBEG}}. The files are labeled as
Line 17:		Line 17:
	where NUMBER runs from 1 to {{TAG\|PHON_NSTRUCT}}.		where NUMBER runs from 1 to {{TAG\|PHON_NSTRUCT}}.

	'''Step 2''': Run calculation for the previously created {{TAG\|POSCAR}} files on the desired observable. These calculations can be anything that is suitable for a ~~Monte Carlo~~ sum <math>\langle O(T)\rangle = \frac{1}{n} \sum\limits_{i=1}^{n} O(x_{T}^{\textrm{MC,i}})</math>, for example band gap calculations, absorption spectra calculations, etc.		'''Step 2''': Run calculation for the previously created {{TAG\|POSCAR}} files on the desired observable. These calculations can be anything that is suitable for a MC sum <math>\langle O(T)\rangle = \frac{1}{n} \sum\limits_{i=1}^{n} O(x_{T}^{\textrm{MC,i}})</math>, for example band gap calculations, absorption spectra calculations, etc.

	'''Step 3 (optional)''': Calculate the desired observable for the original "pristine" supercell. This step can be necessary when changes to an observable due to electron-phonon interactions are required. An example of such a calculation is the calculation of the band-gap renormalization due to electron-phonon interactions.		'''Step 3 (optional)''': Calculate the desired observable for the original "pristine" supercell. This step can be necessary when changes to an observable due to electron-phonon interactions are required. An example of such a calculation is the calculation of the band-gap renormalization due to electron-phonon interactions.

	'''Step 4 (optional)''': If ~~Monte Carlo~~ sampling was used, average the results over the number of structures created in step 1 ({{TAG\|PHON_NSTRUCT}}).		'''Step 4 (optional)''': If MC sampling was used, average the results over the number of structures created in step 1 ({{TAG\|PHON_NSTRUCT}}).

	== ZG configuration (one-shot sampling) ==		== ZG configuration (one-shot sampling) ==

	M. Zacharias and F. Giustino{{cite\|zacharias:prb:2016}} introduced a one-shot method (named ZG configuration after the authors). This method is an approximation to full ~~Monte Carlo~~ sampling. It only uses a single distorted structure and hence it is orders of magnitude faster than the full MC sampling. It retains an accuracy very close to the full MC sampling for converged super cell sizes. For example we showed that for the zero-point renormalization of the band gap the accuracy is within 5 meV between the ZG configurations and the full MC sampling{{cite\|karsai:njp:2018}}. Hence we suggest to use this method preferably, where convergence of the super cell size is hard to achieve or the 5 meV accuracy is enough.		M. Zacharias and F. Giustino{{cite\|zacharias:prb:2016}} introduced a one-shot method (named ZG configuration after the authors). This method is an approximation to full MC sampling. It only uses a single distorted structure and hence it is orders of magnitude faster than the full MC sampling. It retains an accuracy very close to the full MC sampling for converged super cell sizes. For example we showed that for the zero-point renormalization of the band gap the accuracy is within 5 meV between the ZG configurations and the full MC sampling{{cite\|karsai:njp:2018}}. Hence we suggest to use this method preferably, where convergence of the super cell size is hard to achieve or the 5 meV accuracy is enough.

	To select the ZG configuration {{TAG\|PHON_NSTRUCT}}=0 has to be set in the {{TAG\|INCAR}} file.		To select the ZG configuration {{TAG\|PHON_NSTRUCT}}=0 has to be set in the {{TAG\|INCAR}} file.