PSI4 Project Logo

DCFT

Performs Density Cumulant Functional Theory computations

General

ALGORITHM

The algorithm to use for the density cumulant and orbital updates in the energy computation. Two-step algorithm (default) is generally more efficient and shows better convergence than simultaneous

  • Type: string
  • Possible Values: TWOSTEP, SIMULTANEOUS
  • Default: TWOSTEP

AO_BASIS

Controls whether to avoid the AO->MO transformation of the two-electron integrals for the four-virtual case (<VV||VV>) by computing the corresponding terms in the AO basis. AO_BASIS = DISK algorithm reduces the memory requirements. It is, however, less efficient due to the extra I/O, so the default algorithm is preferred.

  • Type: string
  • Possible Values: NONE, DISK
  • Default: NONE

CACHELEVEL

Controls how to cache quantities within the DPD library

  • Type: integer
  • Default: 2

DAMPING_PERCENTAGE

The amount (percentage) of damping to apply to the orbital update procedure: 0 will result in a full update, 100 will completely stall the update. A value around 20 (which corresponds to 20% of the previous iteration’s density being mixed into the current iteration) can help in cases where oscillatory convergence is observed.

  • Type: double
  • Default: 0.0

DIIS_MAX_VECS

Maximum number of error vectors stored for DIIS extrapolation

  • Type: integer
  • Default: 6

DIIS_MIN_VECS

Minimum number of error vectors stored for DIIS extrapolation

  • Type: integer
  • Default: 3

DIIS_START_CONVERGENCE

Value of RMS of the density cumulant residual and SCF error vector below which DIIS extrapolation starts. Same keyword controls the DIIS extrapolation for the solution of the response equations.

IGNORE_TAU

Controls whether to ignore terms containing non-idempotent contribution to OPDM or not (for debug puproses only). For practical applications only the default must be used

INTS_TOLERANCE

Minimum absolute value below which integrals are neglected

LAMBDA_MAXITER

Maximum number of density cumulant update micro-iterations per macro-iteration (for ALOGRITHM = TWOSTEP). Same keyword controls the maximum number of density cumulant response micro-iterations per macro-iteration for the solution of the response equations (for RESPONSE_ALOGRITHM = TWOSTEP)

  • Type: integer
  • Default: 50

LOCK_OCC

Controls whether to force the occupation to be that of the SCF guess. For practical applications only the default must be used

MAXITER

Maximum number of macro-iterations for both energy and the solution of the response equations

  • Type: integer
  • Default: 40

MO_RELAX

Controls whether to relax the orbitals during the energy computation or not (for debug puproses only). For practical applications only the default must be used

RESPONSE_ALGORITHM

The algorithm to use for the solution of the response equations for the analytic gradients and properties. Two-step algorithm is generally more efficient than simultaneous and is used by default

  • Type: string
  • Possible Values: TWOSTEP, SIMULTANEOUS
  • Default: TWOSTEP

R_CONVERGENCE

Convergence criterion for the RMS of the residual vector in the density cumulant updates as well as the solution of the density cumulant and orbital response equations. In the orbital updates controls the RMS of the SCF error vector

SCF_MAXITER

Maximum number of orbital update micro-iterations per macro-iteration (for ALOGRITHM = TWOSTEP). Same keyword controls the maximum number of orbital response micro-iterations per macro-iteration for the solution of the response equations (for RESPONSE_ALOGRITHM = TWOSTEP)

  • Type: integer
  • Default: 50

TAU_SQUARED

Controls whether to compute the DCFT energy with the Tau^2 correction to Tau

TIKHONOW_OMEGA

The shift applied to the denominator in the density cumulant update iterations

  • Type: double
  • Default: 0.0

TPDM

Controls whether to compute unrelaxed two-particle density matrix at the end of the energy computation