Input/output from the propagation via RT-BSE method. More...

Functions/Subroutines
subroutine, public	print_rtbse_header_info (rtbse_env)
	Writes the header and basic info to the standard output.

subroutine, public	print_etrs_info (rtbse_env, step, metric)
	Writes the update after single etrs iteration - only for log level > medium.

subroutine, public	print_etrs_info_header (rtbse_env)
	Writes the header for the etrs iteration updates - only for log level > medium.

subroutine, public	print_timestep_info (rtbse_env, step, electron_num_re, convergence, etrs_num, step_walltime)
	Writes the summary line of a completed propagation timestep.

subroutine, public	output_mos_contravariant (rtbse_env, rho, print_key_section)
	Outputs the matrix in MO basis for matrix coefficients corresponding to contravariant operator, i.e. density matrix.

subroutine, public	output_mos_covariant (rtbse_env, ham, print_key_section)
	Outputs the matrix in MO basis for matrix components corresponding to covariant representation, i.e. the Hamiltonian matrix.

subroutine, public	output_field (rtbse_env, append_opt)
	Prints the current field components into a file provided by input.

subroutine, public	read_field (rtbse_env)
	Reads the field from the files provided by input - useful for the continuation run.

subroutine, public	output_moments (rtbse_env, rho)
	Outputs the expectation value of moments from a given density matrix.

subroutine, public	output_restart (rtbse_env, rho, time_index)
	Outputs the restart info (last finished iteration step) + restard density matrix.

subroutine, public	read_restart (rtbse_env)
	Reads the density matrix from restart files and updates the starting time.

subroutine, public	output_restart_linearized (rtbse_env, rho)
	Linearized RT-BSE restart writer. Writes the restart set: lab-frame MO-active density matrices, the .info step index (sim_step = steps completed = the resume step), the once-per-run C_active gauge reference, and the appended RESTART.trace record feeding the FT prefix on continuation. All indices derive from sim_step so the .info/.trace bookkeeping cannot drift apart. (Full RTBSE uses the upstream output_restart above.)

subroutine, public	read_restart_info (rtbse_env)
	Early phase of the restart read: the starting step index from the .info file, the original run's dt peeked from the RESTART.trace header (so ENFORCE_MAX_DT can inherit it), and an SCF_GUESS hygiene check. Runs BEFORE initialize_maximum_timestep; the trace prefix records are loaded separately by read_restart_trace once the trace arrays are sized.

subroutine, public	read_restart_trace (rtbse_env)
	Reads the RESTART.trace prefix (records 1..sim_start) into the in-memory moment/field/ time traces so the continuation FT covers the full history. Header guards: version, n_spin, dims, dt (abort); kick params (warn); eps_active > 0.1 meV (warn). All ranks read (the traces are replicated). Record layout matches write_restart_trace verbatim.

subroutine, public	check_restart_eps_consistency (rtbse_env)
	Compares the original run's active eigenvalues (stashed by read_restart_trace) against the recomputed eps_active, once the Hamiltonian is built. Prints the max deviation and warns above 0.1 meV (GW analytic continuation gives run-to-run QP noise above FP, so the threshold is deliberately loose). Frees the stash. No-op if nothing was stashed.

subroutine, public	read_restart_density (rtbse_env)
	Late phase of the restart read: overwrites rho from the lab-frame restart matrices and sets restart_extracted. MO-active for the linearized path, AO otherwise (cp_fm_read_unformatted aborts on a size mismatch, guarding a changed active window).

subroutine, public	read_restart_c (rtbse_env, c_old, found)
	Reads the previous run's C_active slabs (gauge reference for the restart basis bridge).

Detailed Description

Input/output from the propagation via RT-BSE method.

Author: Stepan Marek (08.24)

Function/Subroutine Documentation

◆ print_rtbse_header_info()

subroutine, public rt_bse_io::print_rtbse_header_info ( type(rtbse_env_type) rtbse_env )

Writes the header and basic info to the standard output.

Parameters

rtbse_env Entry point - rtbse environment

Definition at line 90 of file rt_bse_io.F.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ print_etrs_info()

subroutine, public rt_bse_io::print_etrs_info	(	type(rtbse_env_type)	rtbse_env,
		integer	step,
		real(kind=dp)	metric
	)

Writes the update after single etrs iteration - only for log level > medium.

Parameters

rtbse_env Entry point - rtbse environment

Definition at line 139 of file rt_bse_io.F.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ print_etrs_info_header()

subroutine, public rt_bse_io::print_etrs_info_header ( type(rtbse_env_type) rtbse_env )

Writes the header for the etrs iteration updates - only for log level > medium.

Parameters

rtbse_env Entry point - rtbse environment

Definition at line 156 of file rt_bse_io.F.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ print_timestep_info()

subroutine, public rt_bse_io::print_timestep_info	(	type(rtbse_env_type)	rtbse_env,
		integer	step,
		real(kind=dp), dimension(:), intent(in)	electron_num_re,
		real(kind=dp), optional	convergence,
		integer, optional	etrs_num,
		real(kind=dp), optional	step_walltime
	)

Writes the summary line of a completed propagation timestep.

Parameters

rtbse_env	Entry point - rtbse environment
step	Index of the completed timestep
electron_num_re	Real part of the electron number, one entry per spin channel
convergence	Optional convergence metric reached by the ETRS iteration
etrs_num	Optional number of ETRS iterations used in this timestep
step_walltime	Optional wall time of this timestep, in seconds

Definition at line 176 of file rt_bse_io.F.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ output_mos_contravariant()

subroutine, public rt_bse_io::output_mos_contravariant	(	type(rtbse_env_type)	rtbse_env,
		type(cp_cfm_type), dimension(:), pointer	rho,
		type(section_vals_type), pointer	print_key_section
	)

Outputs the matrix in MO basis for matrix coefficients corresponding to contravariant operator, i.e. density matrix.

Parameters

rtbse_env	Entry point - gwbse environment
rho	Density matrix in AO basis
rtp_section	RTP input section

Definition at line 260 of file rt_bse_io.F.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ output_mos_covariant()

subroutine, public rt_bse_io::output_mos_covariant	(	type(rtbse_env_type)	rtbse_env,
		type(cp_cfm_type), dimension(:), pointer	ham,
		type(section_vals_type), pointer	print_key_section
	)

Outputs the matrix in MO basis for matrix components corresponding to covariant representation, i.e. the Hamiltonian matrix.

Parameters

rtbse_env	Entry point - gwbse environment
cohsex	cohsex matrix in AO basis, covariant representation
rtp_section	RTP input section

Definition at line 326 of file rt_bse_io.F.

Here is the call graph for this function:

◆ output_field()

subroutine, public rt_bse_io::output_field	(	type(rtbse_env_type)	rtbse_env,
		logical, optional	append_opt
	)

Prints the current field components into a file provided by input.

Parameters

rtbse_env	Entry point - gwbse environment
rtp_section	RTP input section

Definition at line 364 of file rt_bse_io.F.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ read_field()

subroutine, public rt_bse_io::read_field ( type(rtbse_env_type) rtbse_env )

Reads the field from the files provided by input - useful for the continuation run.

Parameters

rtbse_env	Entry point - gwbse environment
rtp_section	RTP input section

Definition at line 409 of file rt_bse_io.F.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ output_moments()

subroutine, public rt_bse_io::output_moments	(	type(rtbse_env_type)	rtbse_env,
		type(cp_cfm_type), dimension(:), pointer	rho
	)

Outputs the expectation value of moments from a given density matrix.

Note: Moments matrix is provided by the rtbse_env, uses rho_workspace(1:3)

Parameters

rtbse_env	Entry point - gwbse environment
rho	Density matrix in AO basis
rtp_section	RTP section of the input parameters, where moments destination may be present

Definition at line 447 of file rt_bse_io.F.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ output_restart()

subroutine, public rt_bse_io::output_restart	(	type(rtbse_env_type), pointer	rtbse_env,
		type(cp_cfm_type), dimension(:), pointer	rho,
		integer	time_index
	)

Outputs the restart info (last finished iteration step) + restard density matrix.

Parameters

restart_section	Print key section for the restart files
rho	Density matrix in AO basis
time_index	Time index to be written into the info file

Definition at line 494 of file rt_bse_io.F.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ read_restart()

subroutine, public rt_bse_io::read_restart ( type(rtbse_env_type), pointer rtbse_env )

Reads the density matrix from restart files and updates the starting time.

Parameters

restart_section	Print key section for the restart files
rho	Density matrix in AO basis
time_index	Time index to be written into the info file

Definition at line 538 of file rt_bse_io.F.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ output_restart_linearized()

subroutine, public rt_bse_io::output_restart_linearized	(	type(rtbse_env_type), pointer	rtbse_env,
		type(cp_cfm_type), dimension(:), pointer	rho
	)

Linearized RT-BSE restart writer. Writes the restart set: lab-frame MO-active density matrices, the .info step index (sim_step = steps completed = the resume step), the once-per-run C_active gauge reference, and the appended RESTART.trace record feeding the FT prefix on continuation. All indices derive from sim_step so the .info/.trace bookkeeping cannot drift apart. (Full RTBSE uses the upstream output_restart above.)

Parameters

rtbse_env	RT-BSE environment
rho	Density matrix (lab frame) to store

Definition at line 598 of file rt_bse_io.F.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ read_restart_info()

subroutine, public rt_bse_io::read_restart_info ( type(rtbse_env_type), pointer rtbse_env )

Early phase of the restart read: the starting step index from the .info file, the original run's dt peeked from the RESTART.trace header (so ENFORCE_MAX_DT can inherit it), and an SCF_GUESS hygiene check. Runs BEFORE initialize_maximum_timestep; the trace prefix records are loaded separately by read_restart_trace once the trace arrays are sized.

Parameters

rtbse_env RT-BSE environment

Definition at line 718 of file rt_bse_io.F.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ read_restart_trace()

subroutine, public rt_bse_io::read_restart_trace ( type(rtbse_env_type), pointer rtbse_env )

Reads the RESTART.trace prefix (records 1..sim_start) into the in-memory moment/field/ time traces so the continuation FT covers the full history. Header guards: version, n_spin, dims, dt (abort); kick params (warn); eps_active > 0.1 meV (warn). All ranks read (the traces are replicated). Record layout matches write_restart_trace verbatim.

Parameters

rtbse_env RT-BSE environment

Definition at line 768 of file rt_bse_io.F.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ check_restart_eps_consistency()

subroutine, public rt_bse_io::check_restart_eps_consistency ( type(rtbse_env_type), pointer rtbse_env )

Compares the original run's active eigenvalues (stashed by read_restart_trace) against the recomputed eps_active, once the Hamiltonian is built. Prints the max deviation and warns above 0.1 meV (GW analytic continuation gives run-to-run QP noise above FP, so the threshold is deliberately loose). Frees the stash. No-op if nothing was stashed.

Parameters

rtbse_env RT-BSE environment

Definition at line 873 of file rt_bse_io.F.

Here is the caller graph for this function:

◆ read_restart_density()

subroutine, public rt_bse_io::read_restart_density ( type(rtbse_env_type), pointer rtbse_env )

Late phase of the restart read: overwrites rho from the lab-frame restart matrices and sets restart_extracted. MO-active for the linearized path, AO otherwise (cp_fm_read_unformatted aborts on a size mismatch, guarding a changed active window).

Parameters

rtbse_env RT-BSE environment

Definition at line 898 of file rt_bse_io.F.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ read_restart_c()

subroutine, public rt_bse_io::read_restart_c	(	type(rtbse_env_type), pointer	rtbse_env,
		type(cp_fm_type), dimension(:), pointer	c_old,
		logical, intent(out)	found
	)

Reads the previous run's C_active slabs (gauge reference for the restart basis bridge).

Parameters

rtbse_env	RT-BSE environment
C_old	Caller-created fm array (n_spin) on fm_struct_ao_mo_active, filled on success
found	.TRUE. iff all per-spin C files were present and read

Definition at line 950 of file rt_bse_io.F.

Here is the call graph for this function:

Here is the caller graph for this function:

Functions/Subroutines

Detailed Description

Function/Subroutine Documentation

◆ print_rtbse_header_info()

◆ print_etrs_info()

◆ print_etrs_info_header()

◆ print_timestep_info()

◆ output_mos_contravariant()

◆ output_mos_covariant()

◆ output_field()

◆ read_field()

◆ output_moments()

◆ output_restart()

◆ read_restart()

◆ output_restart_linearized()

◆ read_restart_info()

◆ read_restart_trace()

◆ check_restart_eps_consistency()

◆ read_restart_density()

◆ read_restart_c()