I/O routines used by MULTIDIP. More...

Data Types
type	KMatrix
	K-matrix file. More...
type	MappedMatrix
	Auxiliary data structure for matrix (potentially memory-mapped, or distributed) More...
type	MolecularData
	RMT molecular data file. More...
type	ScatAkCoeffs
	Photoionization wave function coefficients. More...

Functions/Subroutines
subroutine	load_mapped_matrix (this, filename, u, offset, rows, cols)
	Read or map a matrix from file.
subroutine	destruct_mapped_matrix (this)
	Finalize a (potentially mapped) matrix object.
subroutine	read_kmatrices (km, lukmt, nkset)
	Read K-matrix files.
subroutine	read_kmatrix_file (km)
	Read K-matrix file.
subroutine	reset_kmatrix_position (km, skip)
	Reset I/O pointer to start of K-matrices.
subroutine	get_kmatrix (km, kmat, skip)
	Read single K-matrix from the K-matrix file.
subroutine	read_wfncoeffs (ak, lusct)
	Read wave function coefficients from files.
subroutine	read_bndcoeffs (bnd, ei, lubnd, mgvn0, stot0)
	Read inner bound wave function coefficients.
subroutine	read_wfncoeffs_file (ak)
	Read wave function coefficients for a single symmetry from a file.
subroutine	reset_wfncoeffs_position (ak, skip)
	Reset I/O pointer to start of Ak-coeffs.
subroutine	get_wfncoeffs (ak, e, re_ak, im_ak, skip)
	Read single Ak-matrix from the Ak-coeffs file.
subroutine	read_molecular_data (moldat, filename, mpiio, read_wmat2)
	Read RMT molecular_data file.
subroutine	read_target_properties (lutarg, prop, etarg)
	Read ion transition dipoles.
subroutine	setup_angular_integrals (moldat)
	Store the angular integrals in a more convenient shape.
subroutine	get_diptrans (moldat, i, iidip, ifdip)
	Return dipole transition descriptors.
subroutine	apply_dipole_matrix (moldat, component, irrpair, transp, nf, nn, x, y)
	Multiply by dipole matrix.
subroutine	assert_status (message, errcode)
	Check for error code.
integer function	clip (x, a, b)
	Clamp value in range.
subroutine	apply_boundary_amplitudes (moldat, irr, transp, x, y)
	Multiply by boundary amplitudes.
subroutine	scale_boundary_amplitudes (moldat, irr, v, vw)
	Scale boundary amplitudes matrix by a diagonal matrix.
subroutine	write_energy_grid (escat)
	Write photoelectron energies to file.
subroutine	write_partial_wave_moments (moldat, m, nesc, suffix)
	Write partial wave moments.
subroutine	write_raw_dipoles (m, chains, nesc, stem)
	Write raw transition dipole moments.
subroutine	write_rsolve_dipoles (moldat, m, chains, escat, lu_pw_dipoles)
	Write transition dipole moments in RSOLVE format.
subroutine	write_cross_section (cs, nesc, erange, filename)
	Write cross sections to a file.

Variables
integer	myproc = 1
integer	nprocs = 1

Detailed Description

I/O routines used by MULTIDIP.

Author: J Benda

Date: 2020 - 2022

This module contains routines that read the necessary input files (K-matrices, scattering coefficients, molecular_data files) and return the needed subset of data.

Function/Subroutine Documentation

◆ apply_boundary_amplitudes()

subroutine multidip_io::apply_boundary_amplitudes	(	type(moleculardata), intent(in)	moldat,
		integer, intent(in)	irr,
		character(len=1), intent(in)	transp,
		real(wp), dimension(:,:), intent(inout)	x,
		real(wp), dimension(:,:), intent(inout)	y )

Multiply by boundary amplitudes.

Author: J Benda

Date: 2020 - 2023

Multiply real matrix X by the matrix of boundary amplitudes and store the result into the real matrix Y. When transp is 'T', then the elements of the matrix X are expected to correspond to outer region channels, while the elements of the matrix Y to the inner region states. Otherwise, the opposite is assumed.

Definition at line 1193 of file multidip_io.F90.

Here is the caller graph for this function:

◆ apply_dipole_matrix()

subroutine multidip_io::apply_dipole_matrix	(	type(moleculardata), intent(in)	moldat,
		integer, intent(in)	component,
		integer, intent(in)	irrpair,
		character(len=1), intent(in)	transp,
		integer, intent(in)	nf,
		integer, intent(in)	nn,
		real(wp), dimension(:, :, :), intent(inout)	x,
		real(wp), dimension(:, :, :), intent(inout)	y )

Multiply by dipole matrix.

Author: J Benda

Date: 2020 - 2022

Multiply matrix 'X' by the dipole matrix corresponding to dipole component 'I' and transition index 's'. Store the result in matrix 'Y'. Use 'nn' rows of 'X', and 'nf' rows of 'Y'.

The matrix argument X and Y have the following one-dimensional distribution among processes:

   !>            +---+---+---+---+
   !>            |   |   |   |   |
   !>            |   |   |   |   |
   !>            | 0 | 1 | 2 | 3 |
   !>            |   |   |   |   |
   !>            |   |   |   |   |
   !>            +---+---+---+---+
   !>

However, when ScaLAPACK multiplication is desired, the arrays need to be redistributed into the standard two-dimensional block-cyclic form

   !>            +-+-+-+-+-+-+-+-+
   !>            |0|1|2|3|0|1|2|3|
   !>            +-+-+-+-+-+-+-+-+
   !>            |1|2|3|0|1|2|3|0|
   !>            +-+-+-+-+-+-+-+-+
   !>            |2|3|0|1|2|3|0|1|
   !>            +-+-+-+-+-+-+-+-+
   !>            |3|0|1|2|3|0|1|2|
   !>            +-+-+-+-+-+-+-+-+
   !>

and back. This is accomplished by the subroutine pdgemr2d. Note however, that this subroutine is prone to failure when the matrices exceed 4-byte addressable number of elements. Working with such matrices then requires patched version of ScaLAPACK that use long integers internally.

Definition at line 1051 of file multidip_io.F90.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ assert_status()

subroutine multidip_io::assert_status	(	character(len=*), intent(in)	message,
		integer, intent(in)	errcode )

Check for error code.

Author: J Benda

Date: 2022 - 2024

If the provided error code is non-zero, print the error message and abort.

Definition at line 1157 of file multidip_io.F90.

Here is the caller graph for this function:

◆ clip()

integer function multidip_io::clip	(	integer	x,
		integer	a,
		integer	b )

Clamp value in range.

Author: J Benda

Date: 2022

Definition at line 1175 of file multidip_io.F90.

Here is the caller graph for this function:

◆ destruct_mapped_matrix()

subroutine multidip_io::destruct_mapped_matrix ( type(mappedmatrix), intent(inout) this )

Finalize a (potentially mapped) matrix object.

Author: J Benda

Date: 2021 - 2022

Safely deallocates the matrix. If it has been mapped, the pointer is only nullified; unmapping happens automatically in the desctructor of the file_mapping type.

Definition at line 280 of file multidip_io.F90.

Here is the caller graph for this function:

◆ get_diptrans()

subroutine multidip_io::get_diptrans	(	type(moleculardata), intent(in)	moldat,
		integer, intent(in)	i,
		integer, dimension(:), intent(inout), allocatable	iidip,
		integer, dimension(:), intent(inout), allocatable	ifdip )

Return dipole transition descriptors.

Author: J Benda

Date: 2020

Based on the Cartesian component index I = 1, 2, 3, (re)allocate the arrays 'iidip' and 'ifdip' and fill them with initial/final irreducible representations pairs for dipole matrices present in the molecular_data file structure 'moldat'.

Definition at line 993 of file multidip_io.F90.

Here is the caller graph for this function:

◆ get_kmatrix()

subroutine multidip_io::get_kmatrix	(	class(kmatrix), intent(in)	km,
		real(wp), dimension(:, :), intent(inout), allocatable	kmat,
		integer, intent(in), optional	skip )

Read single K-matrix from the K-matrix file.

Author: J Benda

Date: 2021 - 2024

Assuming that the K-matrix file associated with this object is correctly positioned, read the next K-matrix record. Symmetrize the K-matrix and store it into the allocatable two-dimensional array kmat (re/allocate as necessary).

Definition at line 466 of file multidip_io.F90.

◆ get_wfncoeffs()

subroutine multidip_io::get_wfncoeffs	(	class(scatakcoeffs), intent(in)	ak,
		real(wp), intent(inout)	e,
		real(wp), dimension(:, :), intent(inout), allocatable	re_ak,
		real(wp), dimension(:, :), intent(inout), allocatable	im_ak,
		integer, intent(in), optional	skip )

Read single Ak-matrix from the Ak-coeffs file.

Author: J Benda

Date: 2021 - 2024

Assuming that the Ak-coeffs file associated with this object is correctly positioned, read the next Ak-matrix record. Also returns the scattering energy in atomic units.

Definition at line 684 of file multidip_io.F90.

◆ load_mapped_matrix()

subroutine multidip_io::load_mapped_matrix	(	class(mappedmatrix), intent(inout)	this,
		character(len=*), intent(in)	filename,
		integer(int32), intent(in)	u,
		integer(int64), intent(in)	offset,
		integer(int32), intent(in)	rows,
		integer(int32), intent(in)	cols )

Read or map a matrix from file.

Author: J Benda

Date: 2021 - 2024

Depending on the compilation settings, either map the given section of the file to memory, or simply allocate and read the chunk. Optionally, read and distribute the matrix in the block-cyclic way.

Definition at line 158 of file multidip_io.F90.

Here is the call graph for this function:

◆ read_bndcoeffs()

subroutine multidip_io::read_bndcoeffs	(	real(wp), dimension(:), intent(out)	bnd,
		real(wp), intent(out)	ei,
		integer, intent(in)	lubnd,
		integer, intent(in)	mgvn0,
		integer, intent(in)	stot0 )

Read inner bound wave function coefficients.

Authors: J Benda

Date: 2023

Read the inner region expansion coefficients of the bound state as calculated by BOUND.

Parameters

[out]	bnd	Vector of inner region expansion coefficients to fill.
[out]	Ei	Bound state energy as calculated by BOUND.
[in]	lubnd	File unit number with the (unformatted) BOUND output.
[in]	mgvn0	Expected MGVN (to check).
[in]	stot0	Expected total spin (to check).

Definition at line 554 of file multidip_io.F90.

Here is the caller graph for this function:

◆ read_kmatrices()

subroutine multidip_io::read_kmatrices	(	type(kmatrix), dimension(:), intent(inout), allocatable	km,
		integer, dimension(:), intent(in)	lukmt,
		integer, dimension(:), intent(in)	nkset )

Read K-matrix files.

Author: J Benda

Date: 2020 - 2024

Read all needed K-matrix files. These are needed to calculate photionization coefficients (Ak) during the calculation. Also perform consistency check between the energy samples in individual files.

Definition at line 310 of file multidip_io.F90.

Here is the caller graph for this function:

◆ read_kmatrix_file()

subroutine multidip_io::read_kmatrix_file ( class(kmatrix), intent(inout) km )

Read K-matrix file.

Author: J Benda

Date: 2021 - 2024

Read metadata from a K-matrix file, count K-matrices and store the unit and dataset for later use in retrieval of the K-matrices themselves. The K-matrices are not being read into memory here to avoid exhausting RAM (particularly in parallel mode) if the K-matrices are large.

Definition at line 353 of file multidip_io.F90.

◆ read_molecular_data()

subroutine multidip_io::read_molecular_data	(	type(moleculardata), intent(inout)	moldat,
		character(*), intent(in)	filename,
		logical, intent(in)	mpiio,
		logical, intent(in)	read_wmat2 )

Read RMT molecular_data file.

Author: J Benda

Date: 2020 - 2023

Read data from the RMT molecular_data file. This includes in particular both the inner and outer region transition dipole elements, as well as for instance Gaunt angular integrals for all partial waves used.

Definition at line 728 of file multidip_io.F90.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ read_target_properties()

subroutine multidip_io::read_target_properties	(	integer, intent(in)	lutarg,
		real(wp), dimension(:,:,:), intent(inout), allocatable	prop,
		real(wp), dimension(:), intent(inout), allocatable	etarg )

Read ion transition dipoles.

Author: J Benda

Date: 2021

Read ion target transition dipoles from a DENPROP file.

Definition at line 892 of file multidip_io.F90.

Here is the caller graph for this function:

◆ read_wfncoeffs()

subroutine multidip_io::read_wfncoeffs	(	type(scatakcoeffs), dimension(:), intent(inout), allocatable	ak,
		integer, dimension(:), intent(in)	lusct )

Read wave function coefficients from files.

Author: J Benda

Date: 2020

Read wave function (Ak) coefficients from an unformatted file written by RSOLVE. This is an optional functionality meant for dabugging. It is recommended that the Ak coefficients are calculated by the present program instead to avoid storing large datasets on disk or in memory.

Definition at line 528 of file multidip_io.F90.

Here is the caller graph for this function:

◆ read_wfncoeffs_file()

subroutine multidip_io::read_wfncoeffs_file ( class(scatakcoeffs), intent(inout) ak )

Read wave function coefficients for a single symmetry from a file.

Author: J Benda

Date: 2021

Definition at line 595 of file multidip_io.F90.

◆ reset_kmatrix_position()

subroutine multidip_io::reset_kmatrix_position	(	class(kmatrix), intent(in)	km,
		integer, intent(in), optional	skip )

Reset I/O pointer to start of K-matrices.

Author: J Benda

Date: 2021

Rewind the unit to the start of the associated dataset and read through to the very beginning of the actual K-matrix data. Also skip the given number of leading K-matrices. This prepares the file for the subsequent calls to get_kmatrix.

Definition at line 431 of file multidip_io.F90.

◆ reset_wfncoeffs_position()

subroutine multidip_io::reset_wfncoeffs_position	(	class(scatakcoeffs), intent(in)	ak,
		integer, intent(in), optional	skip )

Reset I/O pointer to start of Ak-coeffs.

Author: J Benda

Date: 2021

Rewind the unit to the start of the associated dataset and read through to the very beginning of the actual Ak-coeffs data. Also skip the given number of leading Ak-coeffs. This prepares the file for the subsequent calls to get_wfncoeffs.

Definition at line 649 of file multidip_io.F90.

◆ scale_boundary_amplitudes()

subroutine multidip_io::scale_boundary_amplitudes	(	type(moleculardata), intent(in)	moldat,
		integer, intent(in)	irr,
		real(wp), dimension(:), intent(in)	v,
		real(wp), dimension(:, :), intent(inout)	vw )

Scale boundary amplitudes matrix by a diagonal matrix.

Authors: J Benda

Date: 2020 - 2024

Multiply boundary amplitudes for each inner region state by an element of the provided vector 'v'. Writes the result into 'vw'.

Definition at line 1240 of file multidip_io.F90.

Here is the caller graph for this function:

◆ setup_angular_integrals()

subroutine multidip_io::setup_angular_integrals ( type(moleculardata), intent(inout) moldat )

Store the angular integrals in a more convenient shape.

Author: J Benda

Date: 2020 - 2024

Convert the Gaunt integral storage from a list to an easily addressable matrix.

Definition at line 957 of file multidip_io.F90.

Here is the caller graph for this function:

◆ write_cross_section()

subroutine multidip_io::write_cross_section	(	real(wp), dimension(:, :), intent(in)	cs,
		integer, intent(in)	nesc,
		integer, dimension(2), intent(in)	erange,
		character(len=*), intent(in)	filename )

Write cross sections to a file.

Author: J Benda

Date: 2020 - 2024

Write cross sections to a file. The first column of the array is expected to be the photon energy.

Definition at line 1553 of file multidip_io.F90.

Here is the caller graph for this function:

◆ write_energy_grid()

subroutine multidip_io::write_energy_grid ( real(wp), dimension(:), intent(in) escat )

Write photoelectron energies to file.

Author: J Benda

Date: 2023

Write the column list of photoelectron scattering energies to file. Atomic units are used.

Definition at line 1271 of file multidip_io.F90.

Here is the caller graph for this function:

◆ write_partial_wave_moments()

subroutine multidip_io::write_partial_wave_moments	(	type(moleculardata), intent(in)	moldat,
		complex(wp), dimension(:, :, :), intent(in), allocatable	m,
		integer, intent(in)	nesc,
		character(len=*), intent(in)	suffix )

Write partial wave moments.

Author: J Benda

Date: 2020 - 2024

Produce tables with the multi-photon ionisation matrix elements per partial wave.

Definition at line 1290 of file multidip_io.F90.

Here is the caller graph for this function:

◆ write_raw_dipoles()

subroutine multidip_io::write_raw_dipoles	(	complex(wp), dimension(:, :, :, :), intent(in), allocatable	m,
		integer, dimension(:, :), intent(in), allocatable	chains,
		integer, intent(in)	nesc,
		character(len=*), intent(in)	stem )

Write raw transition dipole moments.

Author: J Benda

Date: 2021 - 2024

Produce tables with the multi-photon ionisation matrix elements per partial wave and per ionisation chain.

Definition at line 1347 of file multidip_io.F90.

Here is the caller graph for this function:

◆ write_rsolve_dipoles()

subroutine multidip_io::write_rsolve_dipoles	(	type(moleculardata), intent(in)	moldat,
		complex(wp), dimension(:, :, :, :), intent(in), allocatable	m,
		integer, dimension(:, :), intent(in), allocatable	chains,
		real(wp), dimension(:), intent(in)	escat,
		integer, intent(in)	lu_pw_dipoles )

Write transition dipole moments in RSOLVE format.

Author: J Benda, Z Masin

Date: 2021 - 2024

Produce tables with 1-photon ionisation matrix elements in the format of RSOLVE.

Definition at line 1420 of file multidip_io.F90.

Here is the caller graph for this function:

Variable Documentation

◆ myproc

integer multidip_io::myproc = 1

Definition at line 48 of file multidip_io.F90.

◆ nprocs

integer multidip_io::nprocs = 1

Definition at line 49 of file multidip_io.F90.

Data Types

Functions/Subroutines

Variables

Detailed Description

Function/Subroutine Documentation

◆ apply_boundary_amplitudes()

◆ apply_dipole_matrix()

◆ assert_status()

◆ clip()

◆ destruct_mapped_matrix()

◆ get_diptrans()

◆ get_kmatrix()

◆ get_wfncoeffs()

◆ load_mapped_matrix()

◆ read_bndcoeffs()

◆ read_kmatrices()

◆ read_kmatrix_file()

◆ read_molecular_data()

◆ read_target_properties()

◆ read_wfncoeffs()

◆ read_wfncoeffs_file()

◆ reset_kmatrix_position()

◆ reset_wfncoeffs_position()

◆ scale_boundary_amplitudes()

◆ setup_angular_integrals()

◆ write_cross_section()

◆ write_energy_grid()

◆ write_partial_wave_moments()

◆ write_raw_dipoles()

◆ write_rsolve_dipoles()

Variable Documentation

◆ myproc

◆ nprocs