KPOINTS

Controls Brillouin-zone sampling with k-points. [Edit on GitHub]

Keywords

Keyword descriptions

DEBUG_FULL_KPOINT_SYMMETRY

Type: logical
Default: T
Lone keyword: T
Usage: DEBUG_FULL_KPOINT_SYMMETRY

Description: Use full atomic k-point symmetry also for DEBUG finite-difference points. This is enabled by default so analytical and finite-difference evaluations use the symmetry of their current geometry. Disable it to restrict DEBUG finite-difference energies to inversion/time-reversal symmetry. [Edit on GitHub]

EPS_SYMMETRY

Type: real
Default: 1.00000000E-006
Aliases: EPS_GEO
Usage: EPS_SYMMETRY

Description: Accuracy in k-point symmetry determination. EPS_GEO is accepted as an alias. [Edit on GitHub]

FULL_GRID

Type: logical
Default: F
Lone keyword: T
Usage: FULL_GRID

Description: Use the full, non-symmetry-reduced k-point grid. [Edit on GitHub]

GAMMA_CENTERED

Type: logical
Default: F
Lone keyword: T
Usage: GAMMA_CENTERED

Description: Generate a gamma-centered variant of the Monkhorst-Pack or MacDonald mesh. This shifts the original mesh so it can include the Gamma point, and makes sense only when an even number of subdivisions is used. For MacDonald meshes, the explicit shift is applied after the gamma-centering shift. [Edit on GitHub]

INVERSION_SYMMETRY_ONLY

Type: logical
Default: F
Lone keyword: T
Usage: INVERSION_SYMMETRY_ONLY

Description: Restrict k-point reduction to k-space inversion (time-reversal) symmetry. [Edit on GitHub]

KPOINT

Type: real[4]
Repeatable: yes
Usage: KPOINT x y z w

Description: Specify kpoint coordinates and weight. [Edit on GitHub]

LATTICE_FFT

Type: enum
Default: AUTO
Usage: LATTICE_FFT AUTO

Description: Control the FFT-based transformation of real-space matrices to complete regular k-point grids. AUTO uses the lattice FFT for at least 27 k points when the grid and memory requirements are satisfied. ON bypasses this performance threshold. OFF always uses the established direct phase sum. Unsupported, incomplete, or oversized grids retain the direct fallback. [Edit on GitHub]

Valid values:

  • AUTO Select the FFT from a conservative cost threshold.

  • ON Prefer the FFT whenever the grid and memory checks allow it.

  • OFF Always use the direct real-cell to k-point transform.

PARALLEL_GROUP_SIZE

Type: integer
Default: -1
Usage: PARALLEL_GROUP_SIZE

Description: Number of MPI processes to be used for a single k-point. This number must divide the total number of processes. The number of groups must divide the total number of kpoints. Value=-1 (smallest possible number of processes per group, satisfying the constraints). Value=0 (all processes). Value=n (exactly n processes). [Edit on GitHub]

SCHEME

Type: string[ ]
Usage: SCHEME {KPMETHOD} {integer} {integer} ..

Description: K-point generation scheme. Available options are:

  • NONE

  • GAMMA

  • MONKHORST-PACK

  • MACDONALD

  • GENERAL

For MONKHORST-PACK the number of k points in all 3 dimensions has to be supplied along with the keyword. For MACDONALD also the list of shifts. E.g. MONKHORST-PACK 12 12 8, MACDONALD 4 4 4 0.25 0.25 0.25. GENERAL uses explicitly listed k-points. If symmetry reduction is requested, the explicit set must be equally weighted and closed under the selected operations. [Edit on GitHub]

References: Monkhorst1976, MacDonald1978

SYMMETRY

Type: logical
Default: F
Lone keyword: T
Usage: SYMMETRY

Description: Use symmetry to reduce the number of kpoints. [Edit on GitHub]

SYMMETRY_BACKEND

Type: enum
Default: K290
Usage: SYMMETRY_BACKEND K290

Description: Select the backend used to provide and apply atomic k-point symmetry operations. K290 is the established default. SPGLIB uses the symmetry operations returned by SPGLIB, including their fractional translations. This applies to Monkhorst-Pack, MacDonald, and closed GENERAL k-point sets. If SYMMETRY_REDUCTION_METHOD is not specified, it follows the selected backend. [Edit on GitHub]

Valid values:

  • K290 Use the existing K290 k-point symmetry backend.

  • SPGLIB Use SPGLIB symmetry operations as k-point symmetry backend.

SYMMETRY_REDUCTION_METHOD

Type: enum
Default: K290
Usage: SYMMETRY_REDUCTION_METHOD K290

Description: Select the method used to reduce Monkhorst-Pack and MacDonald k-point meshes when atomic symmetry is enabled. K290 is the established default. SPGLIB uses the symmetry operations returned by SPGLIB for the k-point reduction. GENERAL k-point lists can be reduced when the explicit set is equally weighted and closed under the selected operations. With SYMMETRY_BACKEND K290 this can be used as a comparison mode using K290 operations for SPGLIB-generated mappings. [Edit on GitHub]

Valid values:

  • K290 Use the existing K290 k-point symmetry reduction method.

  • SPGLIB Use SPGLIB symmetry operations for k-point reduction.

UNITS

Type: string
Default: B_VECTOR
Usage: UNITS

Description: Special k-points are defined either in units of reciprocal lattice vectors or in Cartesian coordinates in units of 2Pi/len. B_VECTOR: in multiples of the reciprocal lattice vectors (b). CART_ANGSTROM: In units of 2Pi/Angstrom. CART_BOHR: In units of 2Pi/Bohr. [Edit on GitHub]

VERBOSE

Type: logical
Default: F
Lone keyword: T
Usage: VERBOSE

Description: Verbose output information. [Edit on GitHub]

WAVEFUNCTIONS

Type: enum
Default: COMPLEX
Usage: WAVEFUNCTIONS REAL

Description: Select whether real or complex wavefunctions should be used when allowed by the k-point set. REAL wavefunctions can only represent Gamma or special k-points and symmetry operations with real Bloch phases. Use COMPLEX for general atomic k-point symmetries with nontrivial phases. [Edit on GitHub]

Valid values:

  • REAL Use real wavefunctions (if possible by kpoints specified).

  • COMPLEX Use complex wavefunctions (default).