Input-file format

The input file consists of several parameter blocks.

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

Block name	dcore_pre	dcore	dcore_check	dcore_anacont	dcore_spectrum
[model]	Yes	Yes	Yes	Yes	Yes
[pre]	Yes
[system]		Yes	Yes	Yes	Yes
[impurity_solver]		Yes			Yes
[control]		Yes
[post]			Yes	Yes	Yes
[post.check]			Yes
[post.anacont]				Yes
[post.anacont.pade]				Yes
[post.anacont.spm]				Yes
[post.spectrum]					Yes
[mpi]		Yes			Yes

For example, we can see that dcore_pre reads only [model] and [pre] blocks. Therefore, dcore_pre needs to be re-executed only when the parameters in [model] and [pre] blocks 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”, “file” (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).
interaction_file	String	None	list of filenames that contain U tensor for each inequivalent shell
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

[pre] block

This block includes parameters that controls the behavior of dcore_pre.

Name	Type	Default	Description
skip_lattice	Bool	False	Skip generation of lattice model including H(k) in dcore_pre. If True, data in seedname.h5/dft_input are kept.
skip_umat	Bool	False	Skip generation of U matrix in dcore_pre. If True, data in seedname.h5/DCore/Umat are kept.
skip_local_potential	Bool	False	Skip generation of local potential in dcore_pre. If True, data in seedname.h5/DCore/LocalPotential are kept.
overwrite	Bool	True	Whether seedname.h5 is overwritten. If False, the existing seedname.h5 is repalced.

[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.

[post] block

This block includes parameters that are solely used by dcore_anacont and dcore_spectrum.

Name	Type	Default	Description
dir_post	String	post	Directory to which results of dcore_post are stored.
dir_work	String	work/post	Directory to which temporary files of dcore_post are stored.

[post.anacont] block

This block includes parameters that are solely used by dcore_anacont.

Name	Type	Default	Description
solver	String	algorithm	Algorithm for analytic continuation
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
show_result	Bool	False	plot result of analytic continuation
save_result	Bool	False	plot result of analytic continuation

[post.anacont.pade] block

This block includes parameters that are solely used by dcore_anacont.

Name	Type	Default	Description
iomega_max	Float	0.0	Cut-off frequency of the Matsubara frequency
n_min	Integer	0	lower bound of the number of used Matsubara frequency
n_max	Integer	100000000	upper bound of the number of used Matsubara frequency
eta	Float	0.01	Imaginary Frequency shift to avoid divergence

[post.anacont.spm] block

This block includes parameters that are solely used by dcore_anacont.

Name	Type	Default	Description
n_matsubara	Integer	100000	number of tau points
n_tau	Integer	-1	number of tau points
n_tail	Integer	10	number of matsubara points for tail-fitting
n_sv	Integer	50	number of singular values to be used
lambda	Float	1e-05	coefficient of L1 regularization
solver	String		solver to be used
verbose_opt	Bool	False	show optimization progress
show_fit	Bool	False	plot result of tail-fitting

[post.spectrum] block

This block includes parameters that are solely used by dcore_spectrum.

[post.check] block

This block includes parameters that are solely used by dcore_check.

Name	Type	Default	Description
omega_check	Float	0.0	Maximum frequency for dcore_check. If not specified, a fixed number of Matsubara points are taken.

[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>.