Section OT

• Sets the various options for the orbital transformation (OT) method. Default settings already provide an efficient, yet robust method. Most systems benefit from using the FULL_ALL preconditioner combined with a small value (0.001) of ENERGY_GAP. Well-behaved systems might benefit from using a DIIS minimizer. Advantages: It's fast, because no expensive diagonalisation is performed. If preconditioned correctly, method guaranteed to find minimum. Disadvantages: Sensitive to preconditioning. A good preconditioner can be expensive. No smearing, or advanced SCF mixing possible: POOR convergence for metallic systems.  [Edit on GitHub]

• Section path: CP2K_INPUT / FORCE_EVAL / DFT / SCF / DIAGONALIZATION / OT

• This section cannot be repeated.

• This section cites the following references: [VandeVondele2003] [Weber2008]

• Subsections

  • none

• Keywords

  • SECTION_PARAMETERS
  • ALGORITHM
  • BROYDEN_ADAPTIVE_SIGMA
  • BROYDEN_BETA
  • BROYDEN_ENABLE_FLIP
  • BROYDEN_ETA
  • BROYDEN_FORGET_HISTORY
  • BROYDEN_GAMMA
  • BROYDEN_OMEGA
  • BROYDEN_SIGMA
  • BROYDEN_SIGMA_DECREASE
  • BROYDEN_SIGMA_MIN
  • CHOLESKY
  • ENERGIES
  • ENERGY_GAP
  • EPS_IRAC
  • EPS_IRAC_FILTER_MATRIX
  • EPS_IRAC_QUICK_EXIT
  • EPS_IRAC_SWITCH
  • EPS_TAYLOR
  • GOLD_TARGET
  • IRAC_DEGREE
  • LINESEARCH
  • MAX_IRAC
  • MAX_TAYLOR
  • MINIMIZER
  • NONDIAG_ENERGY
  • NONDIAG_ENERGY_STRENGTH
  • N_HISTORY_VEC
  • OCCUPATION_PRECONDITIONER
  • ON_THE_FLY_LOC
  • ORTHO_IRAC
  • PRECONDITIONER
  • PRECOND_SOLVER
  • ROTATION
  • SAFE_DIIS
  • STEPSIZE

• Keyword descriptions

  SECTION_PARAMETERS
  	&OT {Logical}
  	controls the activation of the ot method  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one logical.
  	The lone keyword behaves as a switch to .TRUE.
  	Default value: .FALSE.

  ALGORITHM
  	ALGORITHM {Keyword}
  	Algorithm to be used for OT  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one keyword.
  	Default value: STRICT
  	List of valid keywords: IRAC Orbital Transformation based Iterative Refinement of the Approximative Congruence transformation (OT/IR). STRICT Strict orthogonality: Taylor or diagonalization based algorithm.
  	This keyword cites the following references: [VandeVondele2003] [VandeVondele2005] [Weber2008]

  BROYDEN_ADAPTIVE_SIGMA
  	BROYDEN_ADAPTIVE_SIGMA {Logical}
  	Enable adaptive curvature estimation  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one logical.
  	The lone keyword behaves as a switch to .TRUE.
  	Default value: .TRUE.

  BROYDEN_BETA
  	BROYDEN_BETA {Real}
  	Underrelaxation for the broyden mixer  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one real.
  	Default value: 9.00000000E-001

  BROYDEN_ENABLE_FLIP
  	BROYDEN_ENABLE_FLIP {Logical}
  	Ensure positive definite update  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one logical.
  	The lone keyword behaves as a switch to .TRUE.
  	Default value: .TRUE.

  BROYDEN_ETA
  	BROYDEN_ETA {Real}
  	Dampening of estimated energy curvature.  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one real.
  	Default value: 7.00000000E-001

  BROYDEN_FORGET_HISTORY
  	BROYDEN_FORGET_HISTORY {Logical}
  	Forget history on bad approximation  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one logical.
  	The lone keyword behaves as a switch to .TRUE.
  	Default value: .FALSE.

  BROYDEN_GAMMA
  	BROYDEN_GAMMA {Real}
  	Backtracking parameter  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one real.
  	Default value: 5.00000000E-001

  BROYDEN_OMEGA
  	BROYDEN_OMEGA {Real}
  	Growth limit of curvature.  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one real.
  	Default value: 1.10000000E+000

  BROYDEN_SIGMA
  	BROYDEN_SIGMA {Real}
  	Curvature of energy functional.  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one real.
  	Default value: 2.50000000E-001

  BROYDEN_SIGMA_DECREASE
  	BROYDEN_SIGMA_DECREASE {Real}
  	Reduction of curvature on bad approximation.  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one real.
  	Default value: 7.00000000E-001

  BROYDEN_SIGMA_MIN
  	BROYDEN_SIGMA_MIN {Real}
  	Minimum adaptive curvature.  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one real.
  	Default value: 5.00000000E-002

  CHOLESKY
  	CHOLESKY {Keyword}
  	If FULL_ALL the cholesky decomposition of the S matrix is used. Options on the algorithm to be used.  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one keyword.
  	Default value: REDUCE
  	List of valid keywords: INVERSE Restore uses operator multiply by inverse of the triangular matrix INVERSE_DBCSR Like inverse, but matrix stored as dbcsr, sparce matrix algebra used when possible OFF The cholesky algorithm is not used REDUCE Reduce is called RESTORE Reduce is replaced by two restore

  ENERGIES
  	ENERGIES {Logical}
  	Optimize orbital energies for use in Fermi-Dirac smearing (requires ROTATION and FD smearing to be active).  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one logical.
  	The lone keyword behaves as a switch to .TRUE.
  	Default value: .FALSE.

  ENERGY_GAP
  	ENERGY_GAP {Real}
  	Should be an estimate for the energy gap [a.u.] (HOMO-LUMO) and is used in preconditioning, especially effective with the FULL_ALL preconditioner, in which case it should be an underestimate of the gap (can be a small number, e.g. 0.002). FULL_SINGLE_INVERSE takes it as lower bound (values below 0.05 can cause stability issues). In general, higher values will tame the preconditioner in case of poor initial guesses. A negative value will leave the choice to CP2K depending on type of preconditioner.  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one real.
  	Default value: -1.00000000E+000

  EPS_IRAC
  	EPS_IRAC {Real}
  	Targeted accuracy during the refinement iteration.  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one real.
  	Default value: 1.00000000E-010

  EPS_IRAC_FILTER_MATRIX
  	EPS_IRAC_FILTER_MATRIX {Real}
  	Sets the threshold for filtering the matrices.  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one real.
  	Default value: 0.00000000E+000

  EPS_IRAC_QUICK_EXIT
  	EPS_IRAC_QUICK_EXIT {Real}
  	Only one extra refinement iteration is done when the norm is below this value.  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one real.
  	Default value: 1.00000000E-005

  EPS_IRAC_SWITCH
  	EPS_IRAC_SWITCH {Real}
  	The algorithm switches to the polynomial refinement when the norm is below this value.  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one real.
  	Default value: 1.00000000E-002

  EPS_TAYLOR
  	EPS_TAYLOR {Real}
  	Target accuracy of the taylor expansion for the matrix functions, should normally be kept as is.  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one real.
  	Default value: 1.00000000E-016
  	Alias names for this keyword: EPSTAYLOR

  GOLD_TARGET
  	GOLD_TARGET {Real}
  	Target relative uncertainty in the location of the minimum for LINESEARCH GOLD  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one real.
  	Default value: 1.00000000E-002

  IRAC_DEGREE
  	IRAC_DEGREE {Integer}
  	The refinement polynomial degree (2, 3 or 4).  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one integer.
  	Default value: 4

  LINESEARCH
  	LINESEARCH {Keyword}
  	1D line search algorithm to be used with the OT minimizer, in increasing order of robustness and cost. MINIMIZER CG combined with LINESEARCH GOLD should always find an electronic minimum. Whereas the 2PNT minimizer is almost always OK, 3PNT might be needed for systems in which successive OT CG steps do not decrease the total energy.  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one keyword.
  	Default value: 2PNT
  	List of valid keywords: 2PNT extrapolate based on 2 points 3PNT ... or on 3 points GOLD perform 1D golden section search of the minimum (very expensive) NONE take fixed length steps
  	Alias names for this keyword: LINE_SEARCH

  MAX_IRAC
  	MAX_IRAC {Integer}
  	Maximum allowed refinement iteration.  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one integer.
  	Default value: 50

  MAX_TAYLOR
  	MAX_TAYLOR {Integer}
  	Maximum order of the Taylor expansion before diagonalisation is preferred, for large parallel runs a slightly higher order could sometimes result in a small speedup.  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one integer.
  	Default value: 4

  MINIMIZER
  	MINIMIZER {Keyword}
  	Minimizer to be used with the OT method  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one keyword.
  	Default value: CG
  	List of valid keywords: BROYDEN Broyden mixing approximating the inverse Hessian CG Conjugate Gradients: most reliable, use for difficult systems. The total energy should decrease at every OT CG step if the line search is appropriate. DIIS Direct inversion in the iterative subspace: less reliable than CG, but sometimes about 50% faster SD Steepest descent: not recommended

  NONDIAG_ENERGY
  	NONDIAG_ENERGY {Logical}
  	Add a non-diagonal energy penalty (FD smearing)  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one logical.
  	The lone keyword behaves as a switch to .TRUE.
  	Default value: .FALSE.

  NONDIAG_ENERGY_STRENGTH
  	NONDIAG_ENERGY_STRENGTH {Real}
  	The prefactor for the non-diagonal energy penalty (FD smearing)  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one real.
  	Default value: 1.00000000E+000

  N_HISTORY_VEC
  	N_HISTORY_VEC {Integer}
  	Number of history vectors to be used with DIIS or BROYDEN  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one integer.
  	Default value: 7
  	Alias names for this keyword: NDIIS, N_DIIS, N_BROYDEN

  OCCUPATION_PRECONDITIONER
  	OCCUPATION_PRECONDITIONER {Logical}
  	Preconditioner with the occupation numbers (FD smearing)  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one logical.
  	The lone keyword behaves as a switch to .TRUE.
  	Default value: .FALSE.

  ON_THE_FLY_LOC
  	ON_THE_FLY_LOC {Logical}
  	On the fly localization of the molecular orbitals. Can only be used with OT/IRAC.  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one logical.
  	Default value: .FALSE.

  ORTHO_IRAC
  	ORTHO_IRAC {Keyword}
  	The orthogonality method.  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one keyword.
  	Default value: CHOL
  	List of valid keywords: CHOL Cholesky. LWDN Loewdin. POLY Polynomial.

  PRECONDITIONER
  	PRECONDITIONER {Keyword}
  	Type of preconditioner to be used with all minimization schemes. They differ in effectiveness, cost of construction, cost of application. Properly preconditioned minimization can be orders of magnitude faster than doing nothing.  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one keyword.
  	Default value: FULL_KINETIC
  	List of valid keywords: FULL_ALL Most effective state selective preconditioner based on diagonalization, requires the ENERGY_GAP parameter to be an underestimate of the HOMO-LUMO gap. This preconditioner is recommended for almost all systems, except very large systems where make_preconditioner would dominate the total computational cost. FULL_KINETIC Cholesky inversion of S and T, fast construction, robust, and relatively good, use for very large systems. FULL_SINGLE Based on H-eS diagonalisation, not as good as FULL_ALL, but somewhat cheaper to apply. FULL_SINGLE_INVERSE Based on H-eS cholesky inversion, similar to FULL_SINGLE in preconditioning efficiency but cheaper to construct, might be somewhat less robust. Recommended for large systems. FULL_S_INVERSE Cholesky inversion of S, not as good as FULL_KINETIC, yet equally expensive. NONE skip preconditioning
  	This keyword cites the following references: [Schiffmann2015] [VandeVondele2003] [Weber2008]

  PRECOND_SOLVER
  	PRECOND_SOLVER {Keyword}
  	How the preconditioner is applied to the residual.  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one keyword.
  	Default value: DEFAULT
  	List of valid keywords: DEFAULT the default DIRECT Cholesky decomposition followed by triangular solve (works for FULL_KINETIC/SINGLE_INVERSE/S_INVERSE) INVERSE_CHOLESKY Cholesky decomposition followed by explicit inversion (works for FULL_KINETIC/SINGLE_INVERSE/S_INVERSE) INVERSE_UPDATE Performs a Hotelling update of the inverse if a previous preconditioner is present. Mainly useful for GPU accelerated systems (works for FULL_KINETIC/SINGLE_INVERSE/S_INVERSE)

  ROTATION
  	ROTATION {Logical}
  	Introduce additional variables so that rotations of the occupied subspace are allowed as well, only needed for cases where the energy is not invariant under a rotation of the occupied subspace such as non-singlet restricted calculations or fractional occupations.  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one logical.
  	The lone keyword behaves as a switch to .TRUE.
  	Default value: .FALSE.

  SAFE_DIIS
  	SAFE_DIIS {Logical}
  	Reject DIIS steps if they point away from the minimum, do SD in that case.  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one logical.
  	Default value: .TRUE.
  	Alias names for this keyword: SAFER_DIIS

  STEPSIZE
  	STEPSIZE {Real}
  	Initial stepsize used for the line search, sometimes this parameter can be reduced to stabilize DIIS or to improve the CG behavior in the first few steps. The optimal value depends on the quality of the preconditioner. A negative values leaves the choice to CP2K depending on the preconditioner.  [Edit on GitHub]
  	This keyword cannot be repeated and it expects precisely one real.
  	Default value: -1.00000000E+000