nrgljubljana_interface::nrg_params_t

#include <nrgljubljana_interface/params.hpp>

Detailed Description

Low-level NRG parameters.

Definition at line 154 of file params.hpp.

Public Attributes
double	accumulation = 0.0
	Shift of the accumulation points for binning.
double	betabar = 1.0
	Parameter \( \bar{\beta} \) for thermodynamics.
size_t	bins = 1000
	Number of bins per decade for spectral data.
bool	broaden = false
	Enable broadening of spectra.
double	broaden_min_ratio = 3.0
	Auto-tune the broaden_min parameter.
bool	calc0 = true
	Perform calculations at the 0-th iteration.
bool	cfs = false
	Perform a CFS (complete Fock space) calculation.
bool	cfsgt = false
	Compute the CFS greater correlation function.
bool	cfsls = false
	Compute the CFS lesser correlation function.
bool	checkdiag = false
	Test the diagonalisation results.
bool	checkrho = false
	Test that \( \mathrm{tr}(\rho) = 1 \).
bool	checksumrules = false
	Check operator sum rules.
double	chitp = 1.0
	Parameter \( p \) for \( \chi(T) \) calculations.
std::string	diag = "default"
	Eigensolver routine (dsyev|dsyevr|zheev|zheevr|default).
double	diagratio = 1.0
	Ratio of eigenstates computed in partial diagonalisation.
int	diagth = 1
	Number of diagonalisation threads.
double	discard_immediately = 1e-16
	Peak clipping on the fly.
double	discard_trim = 1e-16
	Peak clipping at the end of the run.
std::string	discretization = "Z"
	Discretization scheme.
bool	dm = false
	Compute density matrices.
bool	dmnrg = false
	Perform a DMNRG (density-matrix NRG) calculation.
bool	dmnrgmats = false
	Perform the DMNRG calculation on the Matsubara axis.
bool	done = true
	Create a DONE file.
size_t	dsyevrlimit = 100
	Minimal matrix size for dsyevr.
bool	dump_f = false
	Dump \( \langle f \rangle \) matrix elements.
bool	dumpabs = false
	Dump in terms of absolute energies.
size_t	dumpannotated = 0
	Number of eigenvalues to dump.
size_t	dumpdiagonal = 0
	Dump diagonal matrix elements.
bool	dumpenergies = false
	Dump all energies to a file.
bool	dumpgroups = true
	Dump by grouping degenerate states.
size_t	dumpprecision = 8
	Number of digits of precision used when dumping.
bool	dumpscaled = true
	Dump using omega_N energy units.
bool	dumpsubspaces = false
	Save detailed subspace info.
double	emax = -1.0
	Upper binning limit.
double	emin = -1.0
	Lower binning limit.
bool	fdm = true
	Perform an FDM (full-density-matrix) calculation.
bool	fdmexpv = true
	Calculate expectation values using the FDM algorithm.
size_t	fdmexpvn = 0
	Iteration at which the expectation values are evaluated.
bool	fdmgt = false
	Compute the FDM greater correlation function.
bool	fdmls = false
	Compute the FDM lesser correlation function.
bool	fdmmats = false
	Perform the FDM calculation on the Matsubara axis.
bool	finite = false
	Perform a Costi-Hewson-Zlatic finite-T calculation.
bool	finitemats = false
	Perform a \( T > 0 \) calculation on the Matsubara axis.
double	fixeps = 1e-15
	Threshold value for eigenvalue splitting corrections.
int	forcestop = -1
	Force stop at the given iteration (-1: disabled).
double	goodE = 2.0
	Energy window parameter for patching.
double	grouptol = 1e-6
	Energy tolerance for considering two states as degenerate.
double	gtp = 0.7
	Parameter \( p \) for \( G(T) \) calculations.
bool	lastall = false
	Keep all states in the last iteration for DMNRG.
bool	lastalloverride = false
	Override the automatic lastall setting.
double	linstep = 0
	Bin width for the linear mesh.
std::string	log = ""
	List of tokens defining what to log.
bool	logall = false
	Log everything.
size_t	logenumber = 10
	Number of eigenvalues to show for log=e.
size_t	mats = 100
	Number of Matsubara points to collect.
int	mMAX = -1
	Number of sites in the star representation ( \( -1 \): automatically determined).
size_t	Ninit = 0
	Number of initial Wilson chain operators.
int	Nmax = -1
	Number of sites in the Wilson chain ( \( -1 \): automatically determined).
bool	NN1 = false
	Perform N/N+1 patching.
bool	NN2avg = false
	Average over even and odd N/N+2 spectra.
bool	NN2even = true
	Use even iterations in N/N+2 patching.
double	NNtanh = 0.0
	Parameter \( a \) in the \( \tanh[a(x-0.5)] \) window function.
bool	noimag = true
	Do not output the imaginary parts of expectation values.
double	omega0 = -1.0
	Smallest energy scale in the problem, \( \omega_0 \).
double	omega0_ratio = 1.0
	Sets \( \omega_0 = \mathtt{omega0\_ratio} \times T \).
size_t	prec_custom = 10
	Precision of columns in the 'custom' output file.
size_t	prec_td = 10
	Precision of columns in the 'td' output file.
size_t	prec_xy = 10
	Precision of the spectral function output.
size_t	preccpp = 2000
	Precision for tridiagonalisation.
bool	reim = false
	Output the imaginary parts of the correlators.
bool	removefiles = true
	Remove temporary data files.
bool	restart = true
	Restart the calculation to achieve the truncation goal.
double	restartfactor = 2.0
	Rescale factor used when restart is true.
bool	resume = false
	Attempt to restart the calculation.
double	safeguard = 1e-5
	Additional states to keep in case of a near degeneracy.
size_t	safeguardmax = 200
	Maximal number of additional states to keep.
bool	savebins = true
	Save binned (unbroadened) data.
std::string	specgt = ""
	Conductance curves to compute.
std::string	speci1t = ""
	\( I_1 \) curves to compute.
std::string	speci2t = ""
	\( I_2 \) curves to compute.
std::string	stopafter = ""
	Stop the calculation at a given point.
std::string	strategy = "kept"
	Recalculation strategy.
bool	substeps = false
	Use the interleaved diagonalization scheme.
std::string	tri = "old"
	Tridiagonalisation approach.
bool	v3mm = false
	Compute the 3-leg vertex on the Matsubara/Matsubara axis.
size_t	width_custom = 16
	Width of columns in the 'custom' output file.
size_t	width_td = 16
	Width of columns in the 'td' output file.
double	xmax = -1.0
	Largest \( x \) in the discretization ODE solver ( \( -1 \): automatically determined).
double	z = 1.0
	Parameter \( z \) (twist) in the logarithmic discretization.
size_t	zheevrlimit = 100
	Minimal matrix size for zheevr.

Friends
void	h5_read (h5::group h5group, std::string subgroup_name, nrg_params_t &sp)
	Read nrg_params_t from HDF5.
void	h5_write (h5::group h5group, std::string subgroup_name, nrg_params_t const &sp)
	Write nrg_params_t to HDF5.

---

The documentation for this struct was generated from the following file:

• nrgljubljana_interface/params.hpp