Procedures

returns a list of occupations, constructed from N_occ by redistributing electrons wihtin the orbitals given in active_space_MO

in case the number of basis functions has changed dynamically due to out-projection of high overlap matrix eigenvalues we here adapt the wavefunction object, so that we can smoothly continue the calculation

this is for the direct SCF GTO calculation adds all electrostatic interaction matrix elements to the one-electron Hamiltonian H

add one process to S - S%value(k): rate or cross section - S%E_eV(k): transition energy in eV, rounded to nearest integer - S%N_process is increased by one. - S%tag is given by P, V, A, F, 2 (two-photon abs.), 3 (three-photon), ...

calculate the function delta as defined in varshalovich et al. for use in 6-j symbol:

Function to compute the associated Legendre function of x with order l & m ref) Numerical Recipes in Fortran, pp 246-248 $$( l-m ) P^m_l(x) = x ( 2l-1 ) P^m_{l-1}(x) - ( l+m-1 ) P^m_{l-2}(x)$$ $$P^m_m(x) = (-1)^m (2m-1)!! (1-x^2)^(m/2)$$ $$( n!! = 1 \cdot 3 \cdot 5 \cdot ... \cdot m \; \text{where}\; m <= n )$$ $$P^m_{m+1}(x) = x ( 2m+1 ) P^m_m(x)$$

this function calculates all associated Legendre polynomials $P_l^m (x)$ for every l,m < lmax and return them in an array $p(l(l+1)/2+m+1)$ adapted from the numerical recipes implementation

population analysis based on the grid points

from an occupation occ and a new, LC1, and old, LC2, orbital set plus given basis set overlap matrix S12 return the orbital indices that are involved in the changes/rotations of orbitals Ncore is the number of core orbitals

subroutine averaged_occs: returns an array of occupations that describe the occupation patterns which are averaged, using the averaging scheme given in F%av_occ(:) and the occupation N_occ(:)

Broyden-Fletcher-Goldfarb-Shanno (BFGS) method to approximate the inverse hessian by previous step and this step gradient and previous step gradient, and previous estimated hessian

calculates binomial

calculated distances between atoms (employing periodic boundary conditions)

Calculates bond-order btw atoms. -diagonal elements- should be ignored (in practice they are juat the Mulliken charges) revised 13.09.2016 by Ludger: Bond order is $$B_{AB} = \sum_{\mu on atom A} \sum_{\nu on atom B} (DS)_\mu\nu * (DS)_\nu\mu$$ See e.g. INTERNATIONAL JOURNAL OF QUANTUM CHEMISTRY, VOL. XXVI, I S 1-154 ( 1984)

braket(): $< i | O | j >$ where i, j indicate i-th and j-th basis function braket_Laplacian(): $< i | \nabla^2 | j >$

converts a c_char pointer array cStrPtr to a fortran string fStr

calculates the configuration interaction matrix

calculates the configuration interaction matrix (only diagonal elements)

calculates the matrix-matrix product of the CI matrix with coefficients

calculates the product of the CI matrix with a coefficient vector

calculates the absolute square Fourier transform of the electron density on a cartesian reciprocal grid

calculated fluorescence rates between MOs based on independent el. approximation (i.e. based on orbital energies)

calculated fluorescence rates based on configuration interaction calculation print the configurations in the initial state basis (without projection)

calculated fluorescence rates between MOs based on independent el. approximation (i.e. based on orbital energies) but resolves individual spin configurations

calculated fluorescence rates between MOs based on independent el. approximation (i.e. based on orbital energies) but resolves individual spin configurations and considers averaging over configurations

calculates the oscillator strength based on orbital energies

calculates the oscillator strenght employing a CI calculation make natural orbitals

calculates the oscillator strenght based on CIS energies

calculates the transition dipole matrix elements between basisfunctions and stores them in tdipole_bf( mu, nu, p)

calculates inelastic scattering in the Waller_Hartree approximation

calculates Auger-Meitner decay transitions based on orbital energy differences

calculates Auger-Meitner decay transitions based on configuration interaction calculation

calculates Auger-Meitner decay transitions based on orbital energy differences resolving individual spin configurations

calculates Auger-Meitner decay transitions based on orbital energy differences resolving individual spin configurations and considering averaging over configuration

returns two-electron integrals, where the first index has been contracted with LCAO orbital coefficients LC

this subroutine calculates CIS gradient

calculate_dEdR : MAIN routine Jnuc_want(Nwant) : iNuc indices of your need ret(3,Nwant) : contains dE/dR vectors

calculates dipole matrix elements

a wrapper for calculate_dEdR defined in nuclear_derivatives.f90

calculates the RHF/ ROHF / HFS gradient when GTO are used

function returns the electronic entropy - $$f*ln(f) - (1-f)*ln(1-f)$$

return the inactive Fock matrix i.e. from the inactive orbital's density matrix

return the inactive Fock matrix i.e. from the inactive orbital's density matrix in a direct-SCF context (without pre-calculated teints)

calculates the kinetic energy matrix

calculates Koopmann's hole energies in the presence of an electric field

calculates the Koopmann's Theorem gradient assuming that an excited state energy gradient is $$d/dR E_{i->a} = E_\text{gs} - e_i$$

this subroutine calculated the Koopmann's Theorem gradient assuming that an ionized state energy gradient is $$E_{i} = E_\text{gs} - e_i$$ or and anionic state energy is $$E_{a} = E_\text{gs} + e_a$$ it even calculates gradient for excited states assuming that $$E_{i->a} = E_\text{gs} +e_a - e_i,$$ which ignores mutual interaction between orbital a and i

Calculate the value of the Legendre polynomial and its derivative input : n = the order of poly., x = the point to be calculated results : p = P_n(x), dp = P'_n(x)

calculates the gradients of orbital energies

computes the overlap matrix

calculates photoionization cross sections based on orbital energies

computes photoionizatio cross sections employing configuration interaction calculations

calculates photoionization cross sections based on orbital energies resolving individual spin configurations

calculates photoionization cross sections based on orbital energies resolving individual spin configurations and averaging over configurations

computes all two-electron integrals

calculates the external potential energy matrix

computes the interaction matrix elements of the electrons with nuclei potential

calculate all integration weights on Grid3D and

Check if w.f. is normalized $< \psi_k | \psi_k' >$ = 1 if k = k' = 0 if k /= k' with the matrix form wf' wf = I for a real matrix wf^T wf = I for a complex matrix

Use this procedure when you want a check point. But, for initilization, you must put this without counter at the very begining of your program. ex) check_time() ... check_time( c, t ) print *, job_time( c )

checkpoint(): more convenient procedure than check_time() An array time(0:) is a table containing all checkpoint times - time(0) : reference time stamp - time(1) : elapsed time for the first checkpoint - time(2) : elapsed time for the second checkpoint - ... ex) call init_checkpoint( time ) : initialize a time array ... (do some calculations) ... call checkpoint( time ) : make a checkpoint with the time array

performs CI calculations

calculates different types of cis density matrix for stateindex state state: state index 1...n typ (optional argument): 'F' for full density matrix 'P' for particle density matrix 'H' for hole density matrix 'T' for transition-density matrix

calculates cis states

returns the dominant configuration of the CI state with state index state

calculate a clebsch-gordan coefficient $\langle \frac{j_1}{2}, \frac{m_1}{2}, \frac{j_2}{2}, \frac{m_2}{2} | \frac{j}{2}, \frac{m}{2} \rangle$ arguments are integer and twice the true value.

compares two determinants specified by determinant path left_det(1:Nopen) right_det(1:Nopen) and occupation left_occ(:) and right_occ(:), respectively. only_left(1:M) and only_right(1:M) returns the orbital indices of electrons that are only present in left and right determinant only_left_spin(1:M) and only_right_spin(1:M) the respective spin of the electrons coef_sign returns the sign produced by flips to bring the determinants into best congruence. The array length M can be arbitrarily. When the number of electrons only present in one determinant exceeds M comparison is quit and coef_sign=0 is returned

Computes impact ionization doubly differential cross section (DDCS) and cross section integrated over outgoing energy (DCS) Output total and angle-resolved DCS Write a file "DDCS_DATA.dat" containing the DDCS data for post processing The DDCS_DATA.dat file is organised in data blocks separated by a blank line: First block (First line) : N_theta, N_phi Second block (N_theta lines) : List of theta values (rad) Third block (N_phi lines) : List of phi values (rad) Last block (Last line) : N_trans Block 4 to 3+N_trans (1 + N_phi x N_theta lines )contains the DDCS data for all possible transitions. First line contains : indexes of initial and final MO, transition energy (Ha), coresponding outgoing energy (Ha) The next N_phi x N_theta lines contain (loop over phi then theta) : DDCS(theta,phi), DDCS(theta,phi)*dOmega

Computes impact ionization cross sections (DCS) based on the Waller Hartree approximation. Output total and angle-resolved DCS.

angular grid

constructs atom set

construct (or reconstruct) B from A and G all necessary informations are obtained from - A : atomic orbitals of individual atoms - G : molecular grids

construcst an array of CSF types for a given set of occupation configurations, spin multiplicity, and irreducible represention

This function constructs the RHF Fock Matrix assumes that density matrix is symmetric

returns RHF Fock matrix

This procedure constructs the RHF Fock Matrix assumes that density matrix is symmetric

constructs Gaussian basis set

general function to generate grids grid_type= LGL, LG, or FGH mapping_type= linear or algebraic(reciprocal) This function does 1) set parameters of N and grid_type 2) allocate memory for r, rp, D1, D2, and so on 2) make computational grids 3) make real grids by means of a mapping function 4) construct D1 and D2

creates the molecular Grid G

calculates the H0 matrix for given basis B: hkin+v_ext contribution

This function constructs the J Matrix

construct the K matrix

this subroutine returns the matrix $$M_{ij c}= \sum_o^{\text{occ orbitals}} \sum_{k} \frac{d u_{ko} ({\bf R}) }{d {\bf R}} \left( 4 (i j | k o) - (i k | j o) - (i o | j k) \right)$$ it expects to find the MO-space twoelectron integrals in F%teint

this subroutine returns the matrix $$M_{ij c}= \sum_o^{\text{occ orbitals}} \sum^{\text{occ orbitals}}_{p} \frac{d O_{op} ({\bf R}) }{d {\bf R}} \left( 2 (i j | o p) - (i o | j p) \right)$$ it expects to find the MO-space twoelectron integrals in F%teint

this subroutine returns the matrix $$M_{ij c}= \sum_o^{\text{occ orbitals}} \sum^{\text{occ orbitals}}_{p} - \frac{d O_{op} ({\bf R}) }{d {\bf R}} \left( 2 (i j | o p) - (i o | j p) \right)$$ it doesn't need transformed integrals

this subroutine returns the matrix $$M_{ij c}= \sum_o^{\text{occ orbitals}} \sum_{k}^{\text{virt. orbitals}} \frac{d u_{ko} ({\bf R}) }{d {\bf R}} \left( 4 (i j | k o) - (i k | j o) - (i o | j k) \right)$$ it expects to find the MO-space twoelectron integrals in F%teint

this subroutine returns the matrix $$M_{ij c}= \sum_o^{\text{occ orbitals}} \sum_{k}^{\text{virt. orbitals}} \frac{d u_{ko} ({\bf R}) }{d {\bf R}} \left( 4 (i j | k o) - (i k | j o) - (i o | j k) \right)$$ idea: employ a density difference and screen low contributions

prepares the overlap matrix of the current molecular orbitals to a minimal orbital set based on numerical atomic orbitals

returns ROHF Fock matrix for high spin half filled open shell variant=1 (default) Roothan variant=2 Davidson variant=3 Guest & Saunders

returns ROHF Fock matrix for high spin half filled open shell alternative ROHF Fock matrix routine

constructs UHF Fock Matrix

construct the UHF Fock Matrix assumes that density matrix is symmetric

create WF_data from P, A, G, and B

make a table of coupling matrices for a set of transpositions of a set of maximal NopenMax un-paired electrons

contracts a packed storage teints set with dimension n along the first coordinate with coefficients LC $$(cd |ab) = \sum_i C_{id} (ic|ab)$$

contracts a packed storage teints set with dimension n along the first coordinate with coefficients LC $$(ic| ab) = \sum_j C_{jc} (ij| ab)$$

contracts a packed storage teints set with dimension n along the first coordinate with coefficients LC $$(ab |ij) = \sum_l C_{kb} (ka|ij)$$

contracts a packed storage teints set with dimension n along the first coordinate with coefficients LC $$(ij| ka) = \sum_l C_{la} (ij| kl)$$

convert string into lowercase

convert string into uppercase

creates process structure

returns the coefficient for determinant specified by determinant path P in csf specified by path T

returns printable string for a given spin configuration

computes electronic density on the grid density: $$\rho(x) = \sum_{i=1}^{N_\text{MO}} n_i | \psi_i(x) |^2$$ $$\psi_i(x) = \sum_{j=1}^{N_\text{basis}} C_{ij} \phi_j(x)$$ where $phi(x)$ is a basis function

computes the density matrix

spin-alpha density matrix where alpha is the majority spin population

spin-beta density matrix where alpha is the majority spin population

closed shell part density matrix

open shell part of the density matrix

returns a readable string for printing out determinant specified by occupation and determinant path det

Direct inversion of iterative subspace [DIIS] procedure Implement Eqs. (1)-(4) Pulay J. Comp. Chem. 3. 556 (1982) Feb.25.2015 : K.H : (1) Half of operation (w.r. i and j ) can be skipped using symmetry. (2) You do not need full matrix multiplication (dimension=B%N_basis) since we just need the trace this amounts to ~10 sec gain per iteration for 180 atom (660basis) calculations

Feb.25.2015 : K.H : (1) Half of operation (w.r. i and j ) can be skipped using symmetry. (2) You do not need full matrix multiplication (dimension=B%N_basis) since we just need the trace this amounts to ~10 sec gain per iteration for 180 atom (660basis) calculations

for a given MO return the dominant atomic subshell

prints atom set information

print all parameters

print all two-electron integrals

show error message and stop the program

computes the gradient of the exchange potential part for HFS

calculated factorial

factorial(n,m) = m! / n! for n < m = 1 otherwise thus, factorial(1,2) = 2; factorial(2,1) = 1

this subroutine return calculates the chemical potential such that we have sum(F%N_occ) electrons at temperature P%T_eV with orbital energies F%CE fractional occupations are saved in F%f_occ. by Ludger Inhester 26/11/2014 Kota Hanasaki : revision 1406 to adapt to low temperature 01/12/2014 Kota Hanasaki : revision 1413 fix initial guess Mar 2017 Ludger Inhester : cleaned up code, restricted it to obitals F%Ncore+1, choose better variable names if improvement is smaller than this, increase trial step v = TWO* sum( fermi( F%CE(:),T_AU, mu ) ) - Ne Convergence Check...

computes the overlaps to reference orbitals squared

filters set of occupation numbers to meet ras restrictions

filters set of occupation numbers to meet symmetry restrictions

count how many neighboring atoms for each atom and find the bond distances

This subroutines returns a set of coefficients for 'optimal' representing the current Density matrix F%D as a linear combination of the saved density matrices in the F%Dsave array F%D(:,:) = sum_i coef(i) * F%Dsave(:,:,i) + deltaD(:,:) 'optimal' means that the rest is small in that sense that it has minimal number of nonzero elements. To achieve this, we use unconstrained Newton minimizer to minimize the function

find the best permutation of orbitals specified in MO_index such that sum(gross_pop(MO_index(:),nuc_index(:))) is maximized

prints footnote

analysis possible fragments based on bond order values

correlation functional

correlation functional

exchange functional

this subroutine performs a coupled perturbed hartree fock calculation see details in Gerrat & Mills, JCP 49 1719

returns a list of occupations in occs that distributes Nelectot electrons in Norbtot orbitals where the occupations comply with the general active space definitions provides by the list of gass: gass(i)%orbitals(:) is a list defining an orbital subset gass(i)%nmin the minimum number of electrons in this orbital subset gass(i)%nmax the maximum number of electrons in this orbital subset

returns atomic number of a given element

returns charge and mass of a given element

returns the list of combinations k out of 1..n in lexicographical order

given a list of combinations k out of n this subroutine returns the complement combinations

returns eigenvalues of square matrices

computes eigenvalues of square matrices under consideration that there a extra symmetries

if filename2 is specified, override reading of command line options. instead only read from filename2

get some eigenvalues and eigenvectors using davpack

Solve the eigenvalue problem of H(N,N) real symmetric matrix using dsyevx - matrix: real, H(N,N) - eigenvalues: real, E(num) - eigenvectors: real, Z(N,num) - num: number of eigenvectors and eigenvalues with lowest eigenvalues - epsilon: (optional) absolute tolerance Eigenvalues are stored in E(N)

returns the coupling matrix for translocation mu-<nu using the previously prepared table UT

returns the coupling matrix for translocation mu-<nu using the previously prepared table UT

gradient of real sph harmonic

sum the gradients along coordinates x,y,z

gross orbital population analysis

returns ground state occupation numbers

computes gradient of h0 $omp parallel shared(atm, bas, env, first_bf_of_shell) private(sh1,sh2,atom,atom1,atom2,shls,ret1,n1,n2,data1,m1,m2,i,j,factor,my_env,mu) default(none) reduction(+:v,t) $omp do $omp end do $omp end parallel

function has_hole: P: Parameter structure N_occ: occ array oi: orbital index returns TRUE if there might be a statistical chance for a hole in orbital oi otherwise FALSE

index for the upper triangular matrix in a packed storage scheme SEE: http://www.netlib.org/lapack/lug/node123.html optimization: use precalculated array sq_half(i)= $i(i-1)/2$

maps occupation pattern to index

initialize index_packed: precompute index values

print out CPU_time, current time, and hostname With PGI compiler, CPU_time gives real time instead of CPU time. What can I do for it? Huh? USAGE) call init_time() ... call footnote()

prepares an initial electron density F%rho based on atomic electron densities

this is the equivalent to initial_density but returns a density matrix instead of a grid density

integral on the grid

integral on atomic grid

double center grid integral

single center grid integral

function is_equivalent: N_occ: occ array part: orbital index hole: orbital index returns TRUE if orbital part and orbital hole are equivalent otherwise FALSE

returns true if occ1 and occ2 are equivalent through averaging

function is_occupied: P: Parameter structure N_occ: occ array oi: orbital index returns TRUE if there might be a statistical chance for occupancy in orbital oi otherwise FALSE

function is_pair_occupied: P: Parameter structure N_occ: occ array oi: orbital index oj: orbital index returns TRUE if there is a statistical chance that orbital oi and oj are both occupied otherwise FALSE

computes derivatives of J and K

returns job time

Function to calculate 1st derivative of the Legendre f. of x with order n

Function to calculate the Legendre function of x with order n

creates linear interpolation structure

linearly interpolate

localizes orbitals in MO_index array

loewdin population analysis

loewdin population analysis

returns long range interaction energies

Carry out LU-decomposition for A and put the decomposed matrix into M implementation of Crout's method with partial pivoting.

returns the determinant value of LU-decomposed matrix M

Return the inverse matrix of LU-decomposed matrix M using F solve(M, b)

purges LU_matrix structure

Solve a linear equation for b with LU-decomposed matrix M Return the solution for b as a new vector x b : row vector, not column vector, that is, just 1D array of n

computes coupling matrix for transposition uj->ui oj is moved down to position oi

returns the determinant of square matrix A slightly redundant to LU_determiant but allows also to return zero as a determinant

returns the trace of matrix A

computes the trace of the matrix A * B first dimension of A has to match second dimension of B second dimension of A has to match first dimension of B

MCSCF procedure

returns the indices that sorts the array

arg sorts integer array A

arg sorts integer array A

sorts array argument

sorts integer array A

mulliken population analysis

Mulliken atomic population analysis Ref) Mulliken, JCP 23, 1833 (1955)

routine calculates the genealogical pathes for all configuration state functions for given multiplicity multp with Nopen open shells. pathes are returned in arrays csfs(1:Nopen, :) example pathes for multp=2 and Nopen=3 there are two spin configurations: 1,2,1 means spin configuration '//\' 1,0,1 means spin configuration '/\/'

routine calculates the genealogical pathes for all high spin determinants for given multiplicity multp with Nopen open shells. pathes are returned in arrays dets(1:Nopen,:) example pathes for multp=2 and Nopen=3 1,2,1 means spin up up down 1,0,1 means spin up down up -1,0,1 means spin down up up

nuclear repulsion energy

nuclear_repulsion_energy_crystal() for the crystalline case

nuclear repulsion gradient

returns the number of possible spin configurations

randomly selects a CI states according to overlap with given occupation

washes out fractional occupation number for orbitals with simialr energy

calculates squared overlaps of a configuration occ with orbital coefficients LC1 (averaged over spin) to a set of configurations occs with orbital coefficients LC2

returns overlap square between configurations in occ to configurations in occs with different molecular orbitals and basis sets considering only high-spin configurations

returns the final occupation numbers for a specific

returns 0 for unoccupied MO 1 for singly occupied MO 2 for doubly occupied MO 2+i for orbital in average_occ_MO subspace i=1,...n if the orbitals are partially occupied

maps index to occupation pattern

returns the gradient resulting from one-electron contributions (h0)

returns the one-particle density matrix

computes the one-particle density matrix

Operators for efficient matrix multiplication with LAPACK, but better readibility. (It is assumed that the dimensionality fits.)

multiply two real square matrices, the second is transposed A(B*T)= A .mt. B

substract two real square matrices element wise A-B = A .sub. B This functionality is provided by f90. Element wise, arrays (of same dimensionality) can be added, substracted, multiplied and divided. sub_real_matrix_matrix is not a LAPACK feature. (Michael)

multiply two real square matrices, the first is transposed (A*T)B= A .tm. B

Do a Gram-Schmidt orthogonalization for the coefficients given by C according to the metric given by the overlap matrix B%S, i.e. such that < C(:,i) S(:,:) C(:,j) > = delta_i,j B are the basis set parameters

calculated overlaps of configuration state functions for different orbital sets

computes derivative of overlap matrix with respect to geometry

computes the gradient due to basis overlap

computes the gradient due to basis overlap

compute gradient of overlap matrix

calculated overlaps of configuration state functions for different orbital sets

parses an integer array

parses a float array

parses a string array

correlation potential

Nuc-elec potential

calculating the direct Coulomb potential a.k.a. Hartree potential or classical electron-electron repulsion potential

Calculating Coulomb potential on the atomic grid

Yajiang's original implementation for the direct Coulomb potential - potential_J_MC() - potential_J_A_R() - integral_A() - potential_J_A()

exchange potential

General XC functionals potential_x or functional_x( P, G, rho ) potential_c or functional_c( P, G, rho_a, rho_b )

precalculate all 2el matrix elements

subroutine to precalculate all 2el matrix elements

fills the structure indep_mo_pairs containing pairs of mos forming independent SCF optimization parameters

prepares the initial guess

print spherically averaged quantity (potential or density) for each atom with and without multicenter decomposition

print grid parameters and basis function (numerical atomic orbital) info it is supposed to be in read_param.f90, but it requires A(:) and G wait a minute... even it doesn't contain F!!! why should it be here? it has beend moved to basis_func

print out density of states (DOS)

prints info on Gaussian basis set

Print arbitrary N*M matrix

print all parameters

print projection information

Calculate the product of all occupied orbital projections

Calculate the product of all occupied orbital projections

purges angular grid

deallocates array of atom_sets

deallocate basis structure

deallocates and purges array of csf structures

removes linear interpolation structure

purge parameters

purges process data

removes spline interpolation structure

deallocates Umatrixtable structure

purges Wave function data

counts the number of shells for the basis functions

places minimal nao functions and counts the shells

reads molecular grid data from file

read all parameters added 31.Mar.2015: minimize print-outs for dynamical calc

reads LCAO coefficients from file

This function computes the double angular integral (real spherical harmonics)

real Gaunt coefficient

it just reorders the eigenvectors such that C is most similar to a diagonal matrix

calculates $$\frac{d}{dR} \chi_i 1/\vec{R}_i \chi_j$$ where R is given by pos(1:3) and has atom index 'at'

calculates $$\langle \chi_i \frac{d}{dR_i} (1/\vec{R}_i) \cih_j \rangle$$ where R_ is given by pos(1:3)

rotates the orbitals F%LC inside the av_occ(:)%MO subspace such that they best overlap with the orbitals in the av_occ(:)%MO subspace in Cref

rotates orbital pairs by given angle theta

symmetrized two particle density matrix

writes molecular grid data to file

write LCAO coefficients to file

main SCF prodedure if( .NOT. (P%yes_soscf .and. iter>3 .and. ( iSOSCF>0 .or. F%d_rho < P%SOSCF_start(2) )) ) then << this is wrong : we have to check if SOSCF is actually started (see return value of soscf_driver) if( allocated(F%apply_looser_threshold) )then; if( F%apply_looser_threshold(j,i) .and. iter>=acceptable_convergenceAt_old(j,i)-1 .and. iter<acceptable_convergenceAt_old(j,i)+6 )then; write(*,'(a,i4,2f12.6,a,2es12.4,a,2i2)')"#SCF:apply_lower check convergence around expected point:",iter, dE, F%d_rho,"/",eTHR,dTHR,":",j,i endif; endif --- minimal print option for MC ---

performs an analysis of shakeup transitions based on mutual overlap of molecular orbitals

returns a list of occupations, constructed from N_occ by exciting one electron

returns a list of occupations, constructed from N_occ by exciting one electron

solves AxB=D for D

complex spherical harmonics

calculate real spherical harmonics: this is closely connected to sph_harmonics_prefactor() in molecular_grid.f90 sph_harmonics_real( l, m, theta, phi ) := G%Y_lm_prefactor(lm) * (2l+1) / 4pi * associated_Legendre_P( l, |m|, cos(theta) ) * [ ... ] [ ... ] = 1 for m = 0 = cos( |m| * phi ) for m > 0 = sin( |m| * phi ) for m < 0

performs analysis of the spin center

returns a determinant given in genealogical path as alpha and beta strings plus sign

creates spline interpolation structure

spline interpolation

reuse spline structure and fill it with new values

performs stability analysis by inspecting numerical derivatives of the Fock matrix

return the statistical factor that orbital oi is not occupied basically 1 - f_occ/2

return the statistical factor that orbital oi is occupied basically f_occ/2

function stat_factor_pair_occupied P: Parameter structure N_occ: occ array oi: orbital index oj: orbital index returns the statistical factor that a spin-orbital pair oi, oj is occupied

function stat_factor_part_hole_pair P: Parameter structure N_occ: occ array oi: orbital index for particle oj: orbital index for hole returns the statistical factor that a spin-orbital pair oi, oj has an electron in oi and a hole in oj Notes that : stat_factor_part_hole_pair(oi,oj) = stat_factor_occupied(oi) - stat_factor_pair_occupied(oi,oj)

subshell_name: return 1s, 2s, 2p, ... according to (n,l)

sum up process data over equivalent configurations

make some simple tests for the library it checks that the correct order of p basis functions is used

calculates the transition density between two csfs

transform all 2el matrix elements into MO space

transform all active 2el matrix elements into MO space

transform all 2el matrix elements into MO space only those where two electrons are member of the array core and two electrons are in the array virtual (as required for CIS)

transform 2el matrix elements into MO space only those where two electrons are member of the array active

transform 2el matrix elements into MO space only those of type (cc|ij) and (ci|cj) where two electrons are member of the array core

subroutine to transform all 2el matrix elements into MO space where two electrons are member of the array core and two electrons are in the array virtual (as required for CIS)

returns the gradient resulting from two-electron contributions

returns the gradient resulting from two-electron contributions for ROHF

obtain all-active twoelectrons from transformed two-active two-electron integrals

update Atom_set with displacement

update Atom_set with new NAOs it rereads the NAO files assigned in A(:).

update Grid3D with displacement use this subroutine if only molecular geometry in A has been changed. note that A should be already updated with update_Atom_set_with_displacement() in atom_in_mol.f90.

write_density_cube_file(P,A,B,G,F, filename)

!! This routine writes a molden file containing the basis set definition and the linear combination of orbitals !!

get CI state energies

calculated CI state overlaps

returns CIS excitation energies

returns excitation energies based on MO energies

return atomic orbital labels

returns atom labels

return auger decay rates based on configuration interation

returns bond order analysis

compute the CIS gradient

compute CIS gradients and nacs

returns Loewdin population for CIS

returns mulliken population for CIS

returns cis oscillator strengths

compute the cis state transition density

returns energy

returns fluorescence rates

returns geometry

compute the HFS, HF, or ROHF gradient

returns loewdin charges

returns atomic masses

return LCAO coefficients

returns mulliken charges

returns numerical atomic orbital configuration string

returns occupation numbers

returns oscillator strength

return overlap matrix

compute orbital population analysis

compute process data

returns for hole configurations energies and occupation patterns in the presence of electric field based on Koopmann's theorem

returns for hole configurations energies, occupation patterns, gradients and nacs in the presence of electric field based on Koopmann's theorem

initializes calculations performs the first electronic structure calculation

returns orbital energies

returns for particle configurations energies, occupation patterns, gradients and nacs in the presence of electric field based on Koopmann's theorem

clears all the allocated data purges Xdata structure

reads chkpt file

triggers calculation of CI overlap

set CI state index

sets geometry

set irreducible representation

set LCAO coefficients

set spin multiplicity

sets occupation numbers

sets up the xmol object reads in the input parameters and prepares for the calculation

writes chkpt file

writes density cube file

writes mo cub file

writes molden file

translates x,y,z into r, theta, phi