Input-file format

The input file consists of six parameter blocks named [model], [system], [impurity_solver], [control], [tool], and [mpi].

The following table shows which blocks are used by each program.

	dcore_pre	dcore	dcore_check	dcore_post
[model]	Yes	Yes	Yes	Yes
[system]		Yes	Yes	Yes
[impurity_solver]		Yes		Yes
[control]		Yes
[tool]			Yes	Yes
[mpi]		Yes		Yes

For example, we can see that dcore_pre reads only [model] block. Therefore, dcore_pre needs to be re-executed only when the parameters in [model] block are changed.

The parameters included in each block are explained below.

[model] block

This block includes parameters for defining a model to be solved.

Name	Type	Default	Description
seedname	String	dcore	Name of the system. The model HDF5 file will be seedname.h5.
lattice	String	chain	Chosen from “chain”, “square”, “cubic”, “bethe”, “wannier90”, and “external”
t	Float	1.0	Transfer integral (Nearest neighbor)
t’	Float	0.0	Transfer integral (Second nearest)
nelec	Float	1.0	Number of electrons per unit cell.
norb	String	1	Number of orbitals at each inequivalent shell (integers separated by commas or spaces.)
ncor	Integer	1	Number of correlated shells in a unit cell (for lattice = wannier90).
corr_to_inequiv	String	None	Mapping from correlated shells to equivalent shells (for lattice = wannier90)
bvec	String	[(1.0,0.0,0.0),(0.0,1.0,0.0),(0.0,0.0,1.0)]	Reciprocal lattice vectors in arbitrary unit.
nk	Integer	8	Number of k along each line. This automatically sets nk0=nk1=nk2=nk. This parameter and (nk0, nk1, nk2) are mutually exclusive.
nk0	Integer	0	Number of k along b_0 (for lattice = wannier90, external)
nk1	Integer	0	Number of k along b_1 (for lattice = wannier90, external)
nk2	Integer	0	Number of k along b_2 (for lattice = wannier90, external)
spin_orbit	Bool	False	Whether the spin-orbit case.
interaction	String	kanamori	Chosen from “slater_uj”, “slater_f”, “kanamori”, “respack” (See below)
density_density	Bool	False	If true, only the density-density part of the interaction is used (See below).
kanamori	String	None	U (Diagonal Coulomb pot.), U’ (Off-diagonal Coulomb pot.) and J (Hund coupling) (See below).
slater_f	String	None	Angular momentum, Slater integrals F (See below).
slater_uj	String	None	Angular momentum, Slater integrals in U and J (See below).
slater_basis	String	cubic	Basis of the Slater interaction (See below).
local_potential_matrix	String	None	dict of {ish: ‘filename’} to specify local potential matrix of ish-th shell
local_potential_factor	String	1.0	Prefactors to the local potential matrix (float or list with len=ncor)

See separate pages from the link below for detailed descriptions of some parameters.

lattice parameter

interaction parameter

slater_basis parameter

local_potential_* parameters

[system] block

This block includes thermodynamic parameters and some technical parameters such as the number of Matsubara frequencies.

Name	Type	Default	Description
beta	Float	1.0	Inverse temperature. This parameter is overwritten, if T is given.
T	Float	-1.0	Temperature. If this parameter is given, beta is overwritten by 1/T.
n_iw	Integer	2048	Number of Matsubara frequencies
fix_mu	Bool	False	Whether or not to fix chemical potential to a given value.
mu	Float	0.0	Initial chemical potential.
prec_mu	Float	0.0001	Threshold for calculating chemical potential with the bisection method.
with_dc	Bool	False	Whether or not to use double-counting (DC) correction (See below)
dc_type	String	HF_DFT	Chosen from ‘HF_DFT’ (default), ‘HF_imp’, ‘FLL’
dc_orbital_average	Bool	False	If true, the DC correction is averaged over orbitals in each shell. Off-diagonal components are dropped.
no_tail_fit	Bool	False	Compute Matsubara summation without fitting high-frequency moments.

If the parameter with_dc is specified to True, the following part of the self-energy is subtracted to avoid the double-counting error of the self-energy.

\[\Sigma_{i, \alpha \sigma \beta \sigma'}^{\rm dc-imp} = \delta_{\sigma \sigma'} \sum_{\gamma \delta \sigma_1} U_{\alpha \gamma \beta \delta} \langle c_{\gamma \sigma_1}^\dagger c_{\delta \sigma_1}\rangle_0 - \sum_{\gamma \delta} U_{\alpha \gamma \delta \beta} \langle c_{\gamma \sigma'}^\dagger c_{\delta \sigma}\rangle_0,\]

where \(\langle \cdots \rangle_0\) indicates the expectation value at the initial (Kohn-Sham) state.

[impurity_solver] block

This block specifies an impurity solver to be used and necessary parameters for running the solver program.

Name	Type	Default	Description
name	String	null	Name of impurity solver. Available options are null, TRIQS/cthyb, TRIQS/hubbard-I, ALPS/cthyb, ALPS/cthyb-seg, pomerol.
basis_rotation	String	None	You can specify either ‘Hloc’, ‘None’, or the location of a file.

Additionally, we have to specify solver-dependent parameters in the way like n_cycles{int} = 500000. For details, see the reference manual for each solver.

[control] block

This block includes parameters that control the self-consistency loop of DMFT.

Name	Type	Default	Description
max_step	Integer	100	Maximum steps of DMFT loops
sigma_mix	Float	0.5	Mixing parameter for self-energy
restart	Bool	False	Whether or not restart from a previous calculation stored in a HDF file.
initial_static_self_energy	String	None	dict of {ish: ‘filename’} to specify initial value of the self-energy of ish-th shell. The file format is the same as local_potential_matrix.
initial_self_energy	String	None	Filename containing initial self-energy in the same format as sigma.dat generated by dcore_check.
time_reversal	Bool	False	If true, an average over spin components are taken to set Sz = 0.
time_reversal_transverse	Bool	False	If true, the self-energy is symmetrized so that Sx=Sy=0.
symmetry_generators	String	None	Generators for symmetrization of self-energy.
n_converge	Integer	1	The DMFT loop is terminated if the convergence criterion defined with converge_tol is satisfied n_converge times consecutively.
converge_tol	Float	0.0	Tolerance in the convergence check. The chemical potential and the renormalization factor are examined.

[tool] block

This block includes parameters that are solely used by dcore_post.

Name	Type	Default	Description
nnode	Integer	0	[NOT USED] Number of node for the k path
nk_line	Integer	8	Number of k along each line
knode	String	[(G,0.0,0.0,0.0),(X,1.0,0.0,0.0)]	The name and the fractional coordinate of each k-node.
omega_min	Float	-1.0	Minimum value of real frequency
omega_max	Float	1.0	Max value of real frequency
Nomega	Integer	100	Number of real frequencies
broadening	Float	0.0	An additional Lorentzian broadening
eta	Float	0.0	Imaginary frequency shift for the Pade approximation
omega_pade	Float	1e+20	Cutoff frequency for the Pade approximation. Data in [-i omega_pade, i omega_pade] is used.
n_pade_min	Integer	20	Minimum number of Matsubara frequencies used for Pade approximation.
n_pade_max	Integer	-1	Maximum number of Matsubara frequencies used for Pade approximation. If negative, this will be replaced with n_iw in [system] block.
post_dir	String	post	Directory to which results of dcore_post are stored.
omega_check	Float	0.0	Maximum frequency for dcore_check. If not specified, a fixed number of Matsubara points are taken.
nk_mesh	Integer	0	Number of k points along each axis for computation of A(k,omega) on a 3D mesh
nk0_mesh	Integer	0	Number of k points along b_0 for computation of A(k,omega) on a 3D mesh
nk1_mesh	Integer	0	Number of k points along b_1 for computation of A(k,omega) on a 3D mesh
nk2_mesh	Integer	0	Number of k points along b_2 for computation of A(k,omega) on a 3D mesh
Lambda_IR	Float	10000.0	IR parameter for computing G(k,iw)
cutoff_IR	Float	1e-05	IR cutoff parameter for computing G(k,iw)
cutoff_IR_2P	Float	1e-05	IR cutoff parameter for computing two-particle quantities
gk_smpl_freqs	String	dense	Sampling fermionic frequencies for dcore_gk. dense or the name of a text file that contains sampling frequencies.For gk_smpl_freqs = dense, all fermionic frequencies in the range of [-n_iw, n_iw] are used. A valid text file must contain the number of sampling frequencies in its first line.Each line after the first line must contain two integer numbers: The first one is an sequential index of a sampling frequency (start from 0), and the second one is a fermionic sampling frequency (i.e, -1, 0, 1).

[mpi] block

This block includes parameters which are read by dcore and dcore_post.

Name	Type	Default	Description
command	String	mpirun -np #	Command for executing a MPI job. # will be relaced by the number of processes.

When an option -DMPIEXEC=<MPIRUN> is passed to the cmake command, The default value of command will be replaced with <MPIRUN>.