LCPQ
/
quantum_package
mirror of https://github.com/LCPQ/quantum_package
10
0
Fork 0
Code Issues Releases Wiki Activity

modifs in doc

This commit is contained in:
Emmanuel Giner 2019-01-11 19:37:06 +01:00
parent a30f05b660
commit 3f7cc67646
30 changed files with 686 additions and 771 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
TODO
View File

@ -12,6 +12,8 @@ Problem with EMSL
* Creer une page web pas trop degueu et la mettre ici : http://lcpq.github.io/quantum_package
* Creer une page avec la liste de tous les exectuables
# Exterieur

22
docs/ref Normal file
View File

@ -0,0 +1,22 @@
* See docs/source/_static/links.rst file for many useful macros
* option provider :option:`name_of_module provider`
* subroutine :c:func:`my_subroutine`
* module :ref:`module`
* provider :c:data:`my_subroutine`
* qp_command :ref:`qp_command`
* linux command :command:`qp_command`
* linux command with option :command:`qp_command -o`
* environment variable :envvar:`VARIABLE`
* file :file:`filename`
* for a file to be referenced, type
.. _myref:
in the head of the file.
(see https://www.sphinx-doc.org/en/1.7/markup/inline.html)
Then, to reference for "myref" just type :ref:`myref`
* roles like warning, note, etc : see http://www.sphinx-doc.org/en/1.6/markup/para.html
* URL : `<https://gitlab.com/eginer/qp_plugins_eginer>`_
or use `IRPF90`_ and define
_IRPF90: http://irpf90.ups-tlse.fr
somewhere
* References of published results with QP should be added into docs/source/research.bib in bibtex
format

59
docs/source/modules/cis.rst
View File

@ -8,7 +8,34 @@
cis
===
This module contains a CIS program, built by setting the following rules:
This module contains a CIS program.
The user point of view
----------------------
The :command:`cis` program performs the CI of the ROHF-like + all single excitations on top of it.
This program can be very useful to :
* **Ground state calculations**: generate a guess for the ground state wave function if one is not sure that the :c:func:`scf` program gave the lowest SCF solution. In combination with :c:func:`save_natorb` it can produce new |MOs| in order to reperform an :c:func:`scf` optimization.
* **Excited states calculations**: generate guess for all the :option:`determinants n_states` wave functions, that will be used by the :c:func:`fci` program.
The main keywords/options to be used are:
* :option:`determinants n_states` : number of states to consider for the |CIS| calculation
* :option:`determinants s2_eig` : force all states to have the desired value of :math:`S^2`
* :option:`determinants expected_s2` : desired value of :math:`S^2`
The programmer point of view
----------------------------
This module have been built by setting the following rules:
* The only generator determinant is the Hartree-Fock (single-reference method)
* All generated singly excited determinants are included in the wave function (no perturbative
@ -42,7 +69,35 @@ Subroutines / functions
File: :file:`cis.irp.f`
Configuration Interaction with Single excitations.
Configuration Interaction with Single excitations.
This program takes a reference Slater determinant of ROHF-like occupancy,
and performs all single excitations on top of it, disregarding spatial symmetry and compute the "n_states" lowest eigenstates of that CI matrix. (see :option:`determinants n_states`)
This program can be useful in many cases:
* GROUND STATE CALCULATION: to be sure to have the lowest scf solution, perform an :c:func:`scf` (see the :ref:`hartree_fock` module), then a :c:func:`cis`, save the natural orbitals (see :c:func:`save_natorb`) and reperform an :c:func:`scf` optimization from this MO guess
* EXCITED STATES CALCULATIONS: the lowest excited states are much likely to be dominanted by single-excitations. Therefore, running a :c:func:`cis` will save the "n_states" lowest states within the CIS space in the EZFIO folder, which can afterward be used as guess wave functions for a further multi-state fci calculation if you specify "read_wf" = True before running the fci executable (see :option:`determinants read_wf`). Also, if you specify "s2_eig" = True, the cis will only retain states having the good value :math:`S^2` value (see :option:`determinants s2_eig` and :option:`determinants expected_s2`). If "s2_eig" = False, it will take the lowest n_states, whatever multiplicity they are.
Note: if you would like to discard some orbitals, use :ref:`qp_set_mo_class` to specify:
* "core" orbitals which will be always doubly occupied
* "act" orbitals where an electron can be either excited from or to
* "del" orbitals which will be never occupied

55
docs/source/modules/cisd.rst
View File

@ -8,7 +8,32 @@
cisd
====
This module contains a CISD program, built by setting the following rules:
This module contains a CI of single and double excitations.
The user point of view
----------------------
The :command:`cisd` program performs the CI of the ROHF-like + all single and double excitations on top of it.
This program can be very useful to :
* **Ground state calculations**: generate a guess for the ground state wave function if one is not sure that the :c:func:`scf` program gave the lowest SCF solution. In combination with :c:func:`save_natorb` it can produce new |MOs| in order to reperform an :c:func:`scf` optimization.
* **Excited states calculations**: generate guess for all the :option:`determinants n_states` wave functions, that will be used by the :c:func:`fci` program.
The main keywords/options to be used are:
* :option:`determinants n_states` : number of states to consider for the |cisd| calculation
* :option:`determinants s2_eig` : force all states to have the desired value of :math:`S^2`
* :option:`determinants expected_s2` : desired value of :math:`S^2`
The programmer point of view
----------------------------
This module have been built by setting the following rules:
* The only generator determinant is the Hartree-Fock (single-reference method)
* All generated determinants are included in the wave function (no perturbative
@ -42,7 +67,33 @@ Subroutines / functions
File: :file:`cisd.irp.f`
Configuration Interaction with Single and Double excitations.
Configuration Interaction with Single and Double excitations.
This program takes a reference Slater determinant of ROHF-like occupancy,
and performs all single and double excitations on top of it, disregarding spatial symmetry and compute the "n_states" lowest eigenstates of that CI matrix (see :option:`determinants n_states`).
This program can be useful in many cases:
* GROUND STATE CALCULATION: if even after a :c:func:`cis` calculation, natural orbitals (see :c:func:`save_natorb`) and then :c:func:`scf` optimization, you are not sure to have the lowest scf solution, do the same strategy with the :c:func:`cisd` executable instead of the :c:func:`cis` exectuable to generate the natural orbitals as a guess for the :c:func:`scf`.
* EXCITED STATES CALCULATIONS: the lowest excited states are much likely to be dominanted by single- or double-excitations. Therefore, running a :c:func:`cisd` will save the "n_states" lowest states within the CISD space in the EZFIO folder, which can afterward be used as guess wave functions for a further multi-state fci calculation if you specify "read_wf" = True before running the fci executable (see :option:`determinants read_wf`). Also, if you specify "s2_eig" = True, the cisd will only retain states having the good value :math:`S^2` value (see :option:`determinants expected_s2` and :option:`determinants s2_eig`). If "s2_eig" = False, it will take the lowest n_states, whatever multiplicity they are.
Note: if you would like to discard some orbitals, use :ref:`qp_set_mo_class` to specify:
* "core" orbitals which will be always doubly occupied
* "act" orbitals where an electron can be either excited from or to
* "del" orbitals which will be never occupied

80
docs/source/modules/density_for_dft.rst
View File

@ -39,60 +39,6 @@ Providers
---------
.. c:var:: one_body_dm_alpha_ao_for_dft
.. code:: text
double precision, allocatable :: one_body_dm_alpha_ao_for_dft (ao_num,ao_num,N_states)
double precision, allocatable :: one_body_dm_beta_ao_for_dft (ao_num,ao_num,N_states)
File: :file:`density_for_dft.irp.f`
one body density matrix on the AO basis based on one_body_dm_mo_alpha_for_dft
.. c:var:: one_body_dm_average_mo_for_dft
.. code:: text
double precision, allocatable :: one_body_dm_average_mo_for_dft (mo_num,mo_num)
File: :file:`density_for_dft.irp.f`
.. c:var:: one_body_dm_beta_ao_for_dft
.. code:: text
double precision, allocatable :: one_body_dm_alpha_ao_for_dft (ao_num,ao_num,N_states)
double precision, allocatable :: one_body_dm_beta_ao_for_dft (ao_num,ao_num,N_states)
File: :file:`density_for_dft.irp.f`
one body density matrix on the AO basis based on one_body_dm_mo_alpha_for_dft
.. c:var:: one_body_dm_mo_alpha_for_dft
.. code:: text
double precision, allocatable :: one_body_dm_mo_alpha_for_dft (mo_num,mo_num,N_states)
File: :file:`density_for_dft.irp.f`
density matrix for alpha electrons in the MO basis used for all DFT calculations based on the density
.. c:var:: one_body_dm_mo_alpha_one_det
.. code:: text
@ -107,19 +53,6 @@ Providers
.. c:var:: one_body_dm_mo_beta_for_dft
.. code:: text
double precision, allocatable :: one_body_dm_mo_beta_for_dft (mo_num,mo_num,N_states)
File: :file:`density_for_dft.irp.f`
density matrix for beta electrons in the MO basis used for all DFT calculations based on the density
.. c:var:: one_body_dm_mo_beta_one_det
.. code:: text
@ -134,19 +67,6 @@ Providers
.. c:var:: one_body_dm_mo_for_dft
.. code:: text
double precision, allocatable :: one_body_dm_mo_for_dft (mo_num,mo_num,N_states)
File: :file:`density_for_dft.irp.f`
.. c:var:: one_e_dm_alpha_ao_for_dft
.. code:: text

204
docs/source/modules/determinants.rst
View File

@ -550,196 +550,6 @@ Providers
.. c:var:: one_body_dm_ao_alpha
.. code:: text
double precision, allocatable :: one_body_dm_ao_alpha (ao_num,ao_num)
double precision, allocatable :: one_body_dm_ao_beta (ao_num,ao_num)
File: :file:`density_matrix.irp.f`
One body density matrix on the |AO| basis : :math:`\rho_{AO}(\alpha), \rho_{AO}(\beta)` .
.. c:var:: one_body_dm_ao_beta
.. code:: text
double precision, allocatable :: one_body_dm_ao_alpha (ao_num,ao_num)
double precision, allocatable :: one_body_dm_ao_beta (ao_num,ao_num)
File: :file:`density_matrix.irp.f`
One body density matrix on the |AO| basis : :math:`\rho_{AO}(\alpha), \rho_{AO}(\beta)` .
.. c:var:: one_body_dm_dagger_mo_spin_index
.. code:: text
double precision, allocatable :: one_body_dm_dagger_mo_spin_index (mo_num,mo_num,N_states,2)
File: :file:`density_matrix.irp.f`
.. c:var:: one_body_dm_mo
.. code:: text
double precision, allocatable :: one_body_dm_mo (mo_num,mo_num)
File: :file:`density_matrix.irp.f`
One-body density matrix
.. c:var:: one_body_dm_mo_alpha
.. code:: text
double precision, allocatable :: one_body_dm_mo_alpha (mo_num,mo_num,N_states)
double precision, allocatable :: one_body_dm_mo_beta (mo_num,mo_num,N_states)
File: :file:`density_matrix.irp.f`
:math:`\alpha` and :math:`\beta` one-body density matrix for each state
.. c:var:: one_body_dm_mo_alpha_average
.. code:: text
double precision, allocatable :: one_body_dm_mo_alpha_average (mo_num,mo_num)
double precision, allocatable :: one_body_dm_mo_beta_average (mo_num,mo_num)
File: :file:`density_matrix.irp.f`
:math:`\alpha` and :math:`\beta` one-body density matrix for each state
.. c:var:: one_body_dm_mo_beta
.. code:: text
double precision, allocatable :: one_body_dm_mo_alpha (mo_num,mo_num,N_states)
double precision, allocatable :: one_body_dm_mo_beta (mo_num,mo_num,N_states)
File: :file:`density_matrix.irp.f`
:math:`\alpha` and :math:`\beta` one-body density matrix for each state
.. c:var:: one_body_dm_mo_beta_average
.. code:: text
double precision, allocatable :: one_body_dm_mo_alpha_average (mo_num,mo_num)
double precision, allocatable :: one_body_dm_mo_beta_average (mo_num,mo_num)
File: :file:`density_matrix.irp.f`
:math:`\alpha` and :math:`\beta` one-body density matrix for each state
.. c:var:: one_body_dm_mo_diff
.. code:: text
double precision, allocatable :: one_body_dm_mo_diff (mo_num,mo_num,2:N_states)
File: :file:`density_matrix.irp.f`
Difference of the one-body density matrix with respect to the ground state
.. c:var:: one_body_dm_mo_spin_index
.. code:: text
double precision, allocatable :: one_body_dm_mo_spin_index (mo_num,mo_num,N_states,2)
File: :file:`density_matrix.irp.f`
.. c:var:: one_body_single_double_dm_mo_alpha
.. code:: text
double precision, allocatable :: one_body_single_double_dm_mo_alpha (mo_num,mo_num)
double precision, allocatable :: one_body_single_double_dm_mo_beta (mo_num,mo_num)
File: :file:`density_matrix.irp.f`
:math:`\alpha` and :math:`\beta` one-body density matrix for each state
.. c:var:: one_body_single_double_dm_mo_beta
.. code:: text
double precision, allocatable :: one_body_single_double_dm_mo_alpha (mo_num,mo_num)
double precision, allocatable :: one_body_single_double_dm_mo_beta (mo_num,mo_num)
File: :file:`density_matrix.irp.f`
:math:`\alpha` and :math:`\beta` one-body density matrix for each state
.. c:var:: one_body_spin_density_ao
.. code:: text
double precision, allocatable :: one_body_spin_density_ao (ao_num,ao_num)
File: :file:`density_matrix.irp.f`
One body spin density matrix on the |AO| basis : :math:`\rho_{AO}(\alpha) - \rho_{AO}(\beta)`
.. c:var:: one_body_spin_density_mo
.. code:: text
double precision, allocatable :: one_body_spin_density_mo (mo_num,mo_num)
File: :file:`density_matrix.irp.f`
:math:`\rho(\alpha) - \rho(\beta)`
.. c:var:: one_e_dm_ao_alpha
.. code:: text
@ -3755,20 +3565,6 @@ Subroutines / functions
.. c:function:: test_det
.. code:: text
subroutine test_det
File: :file:`test_det.irp.f`
.. c:function:: u_0_s2_u_0
.. code:: text

141
docs/source/modules/dft_utils_in_r.rst
View File

@ -31,6 +31,7 @@ Providers
double precision, allocatable :: aos_grad_in_r_array (ao_num,n_points_final_grid,3)
double precision, allocatable :: aos_grad_in_r_array_transp (n_points_final_grid,ao_num,3)
double precision, allocatable :: aos_grad_in_r_array_transp_xyz (3,n_points_final_grid,ao_num)
File: :file:`ao_in_r.irp.f`
@ -49,6 +50,26 @@ Providers
double precision, allocatable :: aos_grad_in_r_array (ao_num,n_points_final_grid,3)
double precision, allocatable :: aos_grad_in_r_array_transp (n_points_final_grid,ao_num,3)
double precision, allocatable :: aos_grad_in_r_array_transp_xyz (3,n_points_final_grid,ao_num)
File: :file:`ao_in_r.irp.f`
aos_grad_in_r_array(i,j,k) = value of the kth component of the gradient of ith ao on the jth grid point
aos_grad_in_r_array_transp(i,j,k) = value of the kth component of the gradient of jth ao on the ith grid point
k = 1 : x, k= 2, y, k 3, z
.. c:var:: aos_grad_in_r_array_transp_xyz
.. code:: text
double precision, allocatable :: aos_grad_in_r_array (ao_num,n_points_final_grid,3)
double precision, allocatable :: aos_grad_in_r_array_transp (n_points_final_grid,ao_num,3)
double precision, allocatable :: aos_grad_in_r_array_transp_xyz (3,n_points_final_grid,ao_num)
File: :file:`ao_in_r.irp.f`
@ -195,126 +216,6 @@ Providers
.. c:var:: one_body_dm_alpha_at_r
.. code:: text
double precision, allocatable :: one_body_dm_alpha_at_r (n_points_final_grid,N_states)
double precision, allocatable :: one_body_dm_beta_at_r (n_points_final_grid,N_states)
File: :file:`dm_in_r.irp.f`
one_body_dm_alpha_at_r(i,istate) = n_alpha(r_i,istate) one_body_dm_beta_at_r(i,istate) = n_beta(r_i,istate) where r_i is the ith point of the grid and istate is the state number
.. c:var:: one_body_dm_beta_at_r
.. code:: text
double precision, allocatable :: one_body_dm_alpha_at_r (n_points_final_grid,N_states)
double precision, allocatable :: one_body_dm_beta_at_r (n_points_final_grid,N_states)
File: :file:`dm_in_r.irp.f`
one_body_dm_alpha_at_r(i,istate) = n_alpha(r_i,istate) one_body_dm_beta_at_r(i,istate) = n_beta(r_i,istate) where r_i is the ith point of the grid and istate is the state number
.. c:var:: one_body_grad_2_dm_alpha_at_r
.. code:: text
double precision, allocatable :: one_dm_and_grad_alpha_in_r (4,n_points_final_grid,N_states)
double precision, allocatable :: one_dm_and_grad_beta_in_r (4,n_points_final_grid,N_states)
double precision, allocatable :: one_body_grad_2_dm_alpha_at_r (n_points_final_grid,N_states)
double precision, allocatable :: one_body_grad_2_dm_beta_at_r (n_points_final_grid,N_states)
File: :file:`dm_in_r.irp.f`
one_dm_and_grad_alpha_in_r(1,i,i_state) = d\dx n_alpha(r_i,istate) one_dm_and_grad_alpha_in_r(2,i,i_state) = d\dy n_alpha(r_i,istate) one_dm_and_grad_alpha_in_r(3,i,i_state) = d\dz n_alpha(r_i,istate) one_dm_and_grad_alpha_in_r(4,i,i_state) = n_alpha(r_i,istate) one_body_grad_2_dm_alpha_at_r(i,istate) = d\dx n_alpha(r_i,istate)^2 + d\dy n_alpha(r_i,istate)^2 + d\dz n_alpha(r_i,istate)^2 where r_i is the ith point of the grid and istate is the state number
.. c:var:: one_body_grad_2_dm_beta_at_r
.. code:: text
double precision, allocatable :: one_dm_and_grad_alpha_in_r (4,n_points_final_grid,N_states)
double precision, allocatable :: one_dm_and_grad_beta_in_r (4,n_points_final_grid,N_states)
double precision, allocatable :: one_body_grad_2_dm_alpha_at_r (n_points_final_grid,N_states)
double precision, allocatable :: one_body_grad_2_dm_beta_at_r (n_points_final_grid,N_states)
File: :file:`dm_in_r.irp.f`
one_dm_and_grad_alpha_in_r(1,i,i_state) = d\dx n_alpha(r_i,istate) one_dm_and_grad_alpha_in_r(2,i,i_state) = d\dy n_alpha(r_i,istate) one_dm_and_grad_alpha_in_r(3,i,i_state) = d\dz n_alpha(r_i,istate) one_dm_and_grad_alpha_in_r(4,i,i_state) = n_alpha(r_i,istate) one_body_grad_2_dm_alpha_at_r(i,istate) = d\dx n_alpha(r_i,istate)^2 + d\dy n_alpha(r_i,istate)^2 + d\dz n_alpha(r_i,istate)^2 where r_i is the ith point of the grid and istate is the state number
.. c:var:: one_dm_alpha_in_r
.. code:: text
double precision, allocatable :: one_dm_alpha_in_r (n_points_integration_angular,n_points_radial_grid,nucl_num,N_states)
double precision, allocatable :: one_dm_beta_in_r (n_points_integration_angular,n_points_radial_grid,nucl_num,N_states)
File: :file:`dm_in_r.irp.f`
.. c:var:: one_dm_and_grad_alpha_in_r
.. code:: text
double precision, allocatable :: one_dm_and_grad_alpha_in_r (4,n_points_final_grid,N_states)
double precision, allocatable :: one_dm_and_grad_beta_in_r (4,n_points_final_grid,N_states)
double precision, allocatable :: one_body_grad_2_dm_alpha_at_r (n_points_final_grid,N_states)
double precision, allocatable :: one_body_grad_2_dm_beta_at_r (n_points_final_grid,N_states)
File: :file:`dm_in_r.irp.f`
one_dm_and_grad_alpha_in_r(1,i,i_state) = d\dx n_alpha(r_i,istate) one_dm_and_grad_alpha_in_r(2,i,i_state) = d\dy n_alpha(r_i,istate) one_dm_and_grad_alpha_in_r(3,i,i_state) = d\dz n_alpha(r_i,istate) one_dm_and_grad_alpha_in_r(4,i,i_state) = n_alpha(r_i,istate) one_body_grad_2_dm_alpha_at_r(i,istate) = d\dx n_alpha(r_i,istate)^2 + d\dy n_alpha(r_i,istate)^2 + d\dz n_alpha(r_i,istate)^2 where r_i is the ith point of the grid and istate is the state number
.. c:var:: one_dm_and_grad_beta_in_r
.. code:: text
double precision, allocatable :: one_dm_and_grad_alpha_in_r (4,n_points_final_grid,N_states)
double precision, allocatable :: one_dm_and_grad_beta_in_r (4,n_points_final_grid,N_states)
double precision, allocatable :: one_body_grad_2_dm_alpha_at_r (n_points_final_grid,N_states)
double precision, allocatable :: one_body_grad_2_dm_beta_at_r (n_points_final_grid,N_states)
File: :file:`dm_in_r.irp.f`
one_dm_and_grad_alpha_in_r(1,i,i_state) = d\dx n_alpha(r_i,istate) one_dm_and_grad_alpha_in_r(2,i,i_state) = d\dy n_alpha(r_i,istate) one_dm_and_grad_alpha_in_r(3,i,i_state) = d\dz n_alpha(r_i,istate) one_dm_and_grad_alpha_in_r(4,i,i_state) = n_alpha(r_i,istate) one_body_grad_2_dm_alpha_at_r(i,istate) = d\dx n_alpha(r_i,istate)^2 + d\dy n_alpha(r_i,istate)^2 + d\dz n_alpha(r_i,istate)^2 where r_i is the ith point of the grid and istate is the state number
.. c:var:: one_dm_beta_in_r
.. code:: text
double precision, allocatable :: one_dm_alpha_in_r (n_points_integration_angular,n_points_radial_grid,nucl_num,N_states)
double precision, allocatable :: one_dm_beta_in_r (n_points_integration_angular,n_points_radial_grid,nucl_num,N_states)
File: :file:`dm_in_r.irp.f`
.. c:var:: one_e_dm_alpha_at_r
.. code:: text

87
docs/source/modules/fci.rst
View File

@ -8,12 +8,47 @@
fci
===
|CIPSI| algorithm in the full configuration interaction space.
* :command:`fci` performs the |CIPSI| calculation,
* :command:`pt2` computes the |PT2| contribution using the wave function stored in the |EZFIO|
The user point of view
----------------------
* :c:func:`fci` performs |CIPSI| calculations using a stochastic scheme for both the selection and the |PT2| contribution,
* :c:func:`pt2` computes the |PT2| contribution using the wave function stored in the |EZFIO|
database.
The main keywords/options for this module are:
* :option:`determinants n_det_max` : maximum number of Slater determinants in the CIPSI wave function. The :command:`fci` program will stop when the size of the CIPSI wave function will exceed :option:`determinants n_det_max`.
* :option:`perturbation pt2_max` : absolute value of the |PT2| to stop the CIPSI calculation. Once the |PT2| :math:`<` :option:`perturbation pt2_max`, the CIPSI calculation stops.
* :option:`determinants n_states` : number of states to consider in the CIPSI calculation.
* :option:`determinants read_wf` : if False, starts with a ROHF-like determinant, if True, starts with the current wave function(s) stored in the |EZFIO| folder.
.. note::
For a multi-state calculation, it is recommended to start with :c:func:`cis` or :c:func:`cisd`
wave functions as a guess.
* :option:`determinants s2_eig` : if True, systematically add all the determinants needed to have a pure value of :math:`S^2`. Also, if True, it tracks only the states having the good :option:`determinants expected_s2`.
.. note::
For a multi-state calculation, it is recommended to start with :c:func:`cis` or :c:func:`cisd`
wave functions as a guess.
* :option:`determinants expected_s2` : expected value of :math:`S^2` for the desired spin multiplicity.
The programmer point of view
----------------------------
This module have been created with the :ref:`cipsi` module.
.. seealso::
The documentation of the :ref:`cipsi` module.
@ -66,7 +101,45 @@ Subroutines / functions
File: :file:`fci.irp.f`
Selected Full Configuration Interaction with stochastic selection and PT2.
Selected Full Configuration Interaction with stochastic selection and PT2 .
This program performs a CIPSI-like selected CI using a stochastic scheme for both
the selection of the important Slater determinants and the computation of the PT2 correction.
This CIPSI-like algorithm will be performed for the "n_states" lowest states of the variational space (see :option:`determinants n_det_max`)
The fci program will stop when reaching at least one the two following conditions:
* number of Slater determinant > "n_det_max" (see :option:`determinants n_det_max`)
* |PT2| < "pt2_max" (see :option:`perturbation pt2_max`)
The following other options can be of interest:
If "read_wf" = False, it starts with a ROHF-like Slater determinant as a guess wave function.
If "read_wf" = True , it starts with the wave function(s) stored in the EZFIO folder as guess wave function(s).
(see :option:`determinants read_wf`)
If "s2_eig" = True, it will systematically add all necessary Slater determinants in order
to have a pure spin wave function with an :math:`S^2` value corresponding to "expected_s2".
(see :option:`determinants s2_eig` and :option:`determinants expected_s2`)
For excited states calculations, it is recommended to start with a :c:func:`cis` or :c:func:`cisd` guess wave functions
for the "n_states", and to set "s2_eig" = True.
@ -80,7 +153,13 @@ Subroutines / functions
File: :file:`pt2.irp.f`
Second order perturbative correction to the wave function contained in the EZFIO directory.
Second order perturbative correction to the wave function contained in the EZFIO directory.
This programs runs the stochastic PT2 correction on all "n_states" wave function stored in the EZFIO folder (see :option:`determinant n_states`).
The option for the PT2 correction are the "pt2_relative_error" which is the relative stochastic
error on the PT2 to reach before stopping the stochastic sampling. (see :option:`perturbation pt2_relative_error`)

67
docs/source/modules/generators_full.rst
View File

@ -33,33 +33,6 @@ Providers
.. c:var:: n_det_generators
.. code:: text
integer :: n_det_generators
File: :file:`generators.irp.f`
For Single reference wave functions, the number of generators is 1 : the Hartree-Fock determinant
.. c:var:: psi_coef_generators
.. code:: text
integer(bit_kind), allocatable :: psi_det_generators (N_int,2,psi_det_size)
double precision, allocatable :: psi_coef_generators (psi_det_size,N_states)
File: :file:`generators.irp.f`
For Single reference wave functions, the generator is the Hartree-Fock determinant
.. c:var:: psi_coef_sorted_gen
.. code:: text
@ -75,20 +48,6 @@ Providers
.. c:var:: psi_det_generators
.. code:: text
integer(bit_kind), allocatable :: psi_det_generators (N_int,2,psi_det_size)
double precision, allocatable :: psi_coef_generators (psi_det_size,N_states)
File: :file:`generators.irp.f`
For Single reference wave functions, the generator is the Hartree-Fock determinant
.. c:var:: psi_det_sorted_gen
.. code:: text
@ -117,29 +76,3 @@ Providers
For Single reference wave functions, the generator is the Hartree-Fock determinant
.. c:var:: select_max
.. code:: text
double precision, allocatable :: select_max (size_select_max)
File: :file:`generators.irp.f`
Memo to skip useless selectors
.. c:var:: size_select_max
.. code:: text
integer :: size_select_max
File: :file:`generators.irp.f`
Size of the select_max array

166
docs/source/modules/kohn_sham.rst
View File

@ -52,12 +52,162 @@ Providers
---------
.. c:var:: ao_potential_alpha_xc
.. code:: text
double precision, allocatable :: ao_potential_alpha_xc (ao_num,ao_num)
double precision, allocatable :: ao_potential_beta_xc (ao_num,ao_num)
File: :file:`pot_functionals.irp.f`
.. c:var:: ao_potential_beta_xc
.. code:: text
double precision, allocatable :: ao_potential_alpha_xc (ao_num,ao_num)
double precision, allocatable :: ao_potential_beta_xc (ao_num,ao_num)
File: :file:`pot_functionals.irp.f`
.. c:var:: e_correlation_dft
.. code:: text
double precision :: e_correlation_dft
File: :file:`pot_functionals.irp.f`
.. c:var:: e_exchange_dft
.. code:: text
double precision :: e_exchange_dft
File: :file:`pot_functionals.irp.f`
.. c:var:: fock_matrix_alpha_no_xc_ao
.. code:: text
double precision, allocatable :: fock_matrix_alpha_no_xc_ao (ao_num,ao_num)
double precision, allocatable :: fock_matrix_beta_no_xc_ao (ao_num,ao_num)
File: :file:`fock_matrix_ks.irp.f`
Mono electronic an Coulomb matrix in ao basis set
.. c:var:: fock_matrix_beta_no_xc_ao
.. code:: text
double precision, allocatable :: fock_matrix_alpha_no_xc_ao (ao_num,ao_num)
double precision, allocatable :: fock_matrix_beta_no_xc_ao (ao_num,ao_num)
File: :file:`fock_matrix_ks.irp.f`
Mono electronic an Coulomb matrix in ao basis set
.. c:var:: fock_matrix_energy
.. code:: text
double precision :: ks_energy
double precision :: two_e_energy
double precision :: one_e_energy
double precision :: fock_matrix_energy
double precision :: trace_potential_xc
File: :file:`ks_enery.irp.f`
Kohn-Sham energy containing the nuclear repulsion energy, and the various components of this quantity.
.. c:var:: ks_energy
.. code:: text
double precision :: ks_energy
double precision :: two_electron_energy
double precision :: two_e_energy
double precision :: one_e_energy
double precision :: fock_matrix_energy
double precision :: trace_potential_xc
File: :file:`ks_enery.irp.f`
Kohn-Sham energy containing the nuclear repulsion energy, and the various components of this quantity.
.. c:var:: one_e_energy
.. code:: text
double precision :: ks_energy
double precision :: two_e_energy
double precision :: one_e_energy
double precision :: fock_matrix_energy
double precision :: trace_potential_xc
File: :file:`ks_enery.irp.f`
Kohn-Sham energy containing the nuclear repulsion energy, and the various components of this quantity.
.. c:var:: trace_potential_xc
.. code:: text
double precision :: ks_energy
double precision :: two_e_energy
double precision :: one_e_energy
double precision :: fock_matrix_energy
double precision :: trace_potential_xc
File: :file:`ks_enery.irp.f`
Kohn-Sham energy containing the nuclear repulsion energy, and the various components of this quantity.
.. c:var:: two_e_energy
.. code:: text
double precision :: ks_energy
double precision :: two_e_energy
double precision :: one_e_energy
double precision :: fock_matrix_energy
double precision :: trace_potential_xc
@ -74,6 +224,20 @@ Subroutines / functions
.. c:function:: check_coherence_functional
.. code:: text
subroutine check_coherence_functional
File: :file:`ks_scf.irp.f`
.. c:function:: srs_ks_cf
.. code:: text

183
docs/source/modules/kohn_sham_rs.rst
View File

@ -60,179 +60,12 @@ Providers
---------
.. c:var:: ao_potential_alpha_xc
.. code:: text
double precision, allocatable :: ao_potential_alpha_xc (ao_num,ao_num)
double precision, allocatable :: ao_potential_beta_xc (ao_num,ao_num)
File: :file:`pot_functionals.irp.f`
.. c:var:: ao_potential_beta_xc
.. code:: text
double precision, allocatable :: ao_potential_alpha_xc (ao_num,ao_num)
double precision, allocatable :: ao_potential_beta_xc (ao_num,ao_num)
File: :file:`pot_functionals.irp.f`
.. c:var:: e_correlation_dft
.. code:: text
double precision :: e_correlation_dft
File: :file:`pot_functionals.irp.f`
.. c:var:: e_exchange_dft
.. code:: text
double precision :: e_exchange_dft
File: :file:`pot_functionals.irp.f`
.. c:var:: fock_matrix_alpha_no_xc_ao
.. code:: text
double precision, allocatable :: fock_matrix_alpha_no_xc_ao (ao_num,ao_num)
double precision, allocatable :: fock_matrix_beta_no_xc_ao (ao_num,ao_num)
File: :file:`fock_matrix_rs_ks.irp.f`
Mono electronic an Coulomb matrix in AO basis set
.. c:var:: fock_matrix_beta_no_xc_ao
.. code:: text
double precision, allocatable :: fock_matrix_alpha_no_xc_ao (ao_num,ao_num)
double precision, allocatable :: fock_matrix_beta_no_xc_ao (ao_num,ao_num)
File: :file:`fock_matrix_rs_ks.irp.f`
Mono electronic an Coulomb matrix in AO basis set
.. c:var:: fock_matrix_energy
.. code:: text
double precision :: rs_ks_energy
double precision :: two_electron_energy
double precision :: one_e_energy
double precision :: fock_matrix_energy
double precision :: trace_potential_xc
File: :file:`rs_ks_energy.irp.f`
Range-separated Kohn-Sham energy containing the nuclear repulsion energy, and the various components of this quantity.
.. c:var:: one_e_energy
.. code:: text
double precision :: rs_ks_energy
double precision :: two_electron_energy
double precision :: one_e_energy
double precision :: fock_matrix_energy
double precision :: trace_potential_xc
File: :file:`rs_ks_energy.irp.f`
Range-separated Kohn-Sham energy containing the nuclear repulsion energy, and the various components of this quantity.
.. c:var:: one_electron_energy
.. code:: text
double precision :: rs_ks_energy
double precision :: two_electron_energy
double precision :: one_electron_energy
double precision :: fock_matrix_energy
double precision :: trace_potential_xc
File: :file:`rs_ks_energy.irp.f`
Range-separated Kohn-Sham energy containing the nuclear repulsion energy, and the various components of this quantity.
.. c:var:: rs_ks_energy
.. code:: text
double precision :: rs_ks_energy
double precision :: two_electron_energy
double precision :: one_e_energy
double precision :: fock_matrix_energy
double precision :: trace_potential_xc
File: :file:`rs_ks_energy.irp.f`
Range-separated Kohn-Sham energy containing the nuclear repulsion energy, and the various components of this quantity.
.. c:var:: trace_potential_xc
.. code:: text
double precision :: rs_ks_energy
double precision :: two_electron_energy
double precision :: one_e_energy
double precision :: fock_matrix_energy
double precision :: trace_potential_xc
File: :file:`rs_ks_energy.irp.f`
Range-separated Kohn-Sham energy containing the nuclear repulsion energy, and the various components of this quantity.
.. c:var:: two_electron_energy
.. code:: text
double precision :: rs_ks_energy
double precision :: two_electron_energy
double precision :: two_e_energy
double precision :: one_e_energy
double precision :: fock_matrix_energy
double precision :: trace_potential_xc
@ -249,20 +82,6 @@ Subroutines / functions
.. c:function:: check_coherence_functional
.. code:: text
subroutine check_coherence_functional
File: :file:`rs_ks_scf.irp.f`
.. c:function:: rs_ks_scf
.. code:: text

71
docs/source/modules/single_ref_method.rst
View File

@ -12,3 +12,74 @@ Include this module for single reference methods.
Using this module, the only generator determinant is the Hartree-Fock determinant.
Providers
---------
.. c:var:: n_det_generators
.. code:: text
integer :: n_det_generators
File: :file:`generators.irp.f`
For Single reference wave functions, the number of generators is 1 : the Hartree-Fock determinant
.. c:var:: psi_coef_generators
.. code:: text
integer(bit_kind), allocatable :: psi_det_generators (N_int,2,psi_det_size)
double precision, allocatable :: psi_coef_generators (psi_det_size,N_states)
File: :file:`generators.irp.f`
For Single reference wave functions, the generator is the Hartree-Fock determinant
.. c:var:: psi_det_generators
.. code:: text
integer(bit_kind), allocatable :: psi_det_generators (N_int,2,psi_det_size)
double precision, allocatable :: psi_coef_generators (psi_det_size,N_states)
File: :file:`generators.irp.f`
For Single reference wave functions, the generator is the Hartree-Fock determinant
.. c:var:: select_max
.. code:: text
double precision, allocatable :: select_max (1)
File: :file:`generators.irp.f`
Memo to skip useless selectors
.. c:var:: size_select_max
.. code:: text
integer :: size_select_max
File: :file:`generators.irp.f`
Size of select_max

64
docs/source/modules/tools.rst
View File

@ -25,7 +25,7 @@ Subroutines / functions
File: :file:`diagonalize_h.irp.f`
program that extracts the N_states lowest states of the Hamiltonian within the set of Slater determinants stored in the EZFIO folder
program that extracts the `n_states` lowest states of the Hamiltonian within the set of Slater determinants stored in the EZFIO folder
@ -39,7 +39,17 @@ Subroutines / functions
File: :file:`fcidump.irp.f`
Produce a FCIDUMP file
Produce a regular FCIDUMP file from the MOs stored in the EZFIO folder.
To specify an active space, the class of the mos have to set in the EZFIO folder (see set_mo_class/qp_set_mo_class).
The fcidump program supports 3 types of MO_class :
* the "core" orbitals which are always doubly occupied in the calculation
* the "del" orbitals that are never occupied in the calculation
* the "act" orbitals that will be occupied by a varying number of electrons
@ -53,7 +63,11 @@ Subroutines / functions
File: :file:`four_idx_transform.irp.f`
4-index transformation of two-electron integrals from AO to MO integrals
4-index transformation of two-electron integrals from AO to MO integrals.
This program will compute the two-electron integrals on the MO basis and store it into the EZFIO folder.
This program can be useful if the AO --> MO transformation is an expensive step by itself.
@ -81,7 +95,13 @@ Subroutines / functions
File: :file:`print_e_conv.irp.f`
program that prints in a human readable format the convergence of the CIPSI algorithm
program that prints in a human readable format the convergence of the CIPSI algorithm.
for all istate, this program produces
* a file "EZFIO. :math:`istate.conv" containing the variational and var+PT2 energies as a function of N_det
* for istate > 1, a file EZFIO.` istate.delta_e.conv containing the energy difference (both var and var+PT2) with the ground state as a function of N_det
@ -95,7 +115,7 @@ Subroutines / functions
File: :file:`print_wf.irp.f`
print the wave function stored in the EZFIO folder in the intermediate normalization
print the wave function stored in the EZFIO folder in the intermediate normalization.
it also prints a lot of information regarding the excitation operators from the reference determinant
@ -129,25 +149,15 @@ Subroutines / functions
File: :file:`save_natorb.irp.f`
Save natural MOs into the EZFIO
Save natural MOs into the EZFIO
This program reads the wave function stored in the EZFIO folder,
extracts the corresponding natural orbitals and set them as the new MOs
If this is a multi-state calculation, the density matrix that produces the natural orbitals
.. c:function:: save_one_body_dm
.. code:: text
subroutine save_one_body_dm
File: :file:`save_one_body_dm.irp.f`
programs that computes the one body density on the mo basis for alpha and beta electrons from the wave function stored in the EZFIO folder, and then save it into the EZFIO folder aux_quantities.
Then, the global variable data_one_body_alpha_dm_mo and data_one_body_beta_dm_mo will automatically read the density in a further calculation.
This can be used to perform dampin on the density in RS-DFT calculation (see the density_for_dft module).
is obtained from a state-averaged of the density matrices of each state with the corresponding state_average_weight (see the doc of state_average_weight).
@ -163,9 +173,9 @@ Subroutines / functions
programs that computes the one body density on the mo basis for alpha and beta electrons from the wave function stored in the EZFIO folder, and then save it into the EZFIO folder aux_quantities.
Then, the global variable data_one_e_dm_alpha_mo and data_one_e_dm_beta_mo will automatically read the density in a further calculation.
Then, the global variable data_one_e_dm_alpha_mo and data_one_e_dm_beta_mo will automatically read this density in a further calculation.
This can be used to perform dampin on the density in RS-DFT calculation (see the density_for_dft module).
This can be used to perform damping on the density in RS-DFT calculation (see the density_for_dft module).
@ -179,7 +189,13 @@ Subroutines / functions
File: :file:`save_ortho_mos.irp.f`
Save orthonormalized MOs in the EZFIO.
Save orthonormalized MOs in the EZFIO.
This program reads the current MOs, computes the corresponding overlap matrix in the MO basis
and perform a Lowdin orthonormalization : :math:`MO_{new} = S^{-1/2} MO_{guess}`.
Thanks to the Lowdin orthonormalization, the new MOs are the most similar to the guess MOs.
@ -221,7 +237,7 @@ Subroutines / functions
File: :file:`write_integrals_erf.irp.f`
Saves the two-electron erf integrals into the EZFIO
Saves the two-electron integrals with the :math:`erf(\mu r_{12})/r_{12}` oprerator into the EZFIO folder

36
docs/source/programmers_guide/index_providers.rst
View File

@ -92,6 +92,7 @@ Index of Providers
* :c:data:`aos_dvx_beta_pbe_w`
* :c:data:`aos_grad_in_r_array`
* :c:data:`aos_grad_in_r_array_transp`
* :c:data:`aos_grad_in_r_array_transp_xyz`
* :c:data:`aos_in_r_array`
* :c:data:`aos_in_r_array_transp`
* :c:data:`aos_lapl_in_r_array`
@ -150,8 +151,6 @@ Index of Providers
* :c:data:`damping_for_rs_dft`
* :c:data:`data_energy_proj`
* :c:data:`data_energy_var`
* :c:data:`data_one_body_alpha_dm_mo`
* :c:data:`data_one_body_beta_dm_mo`
* :c:data:`data_one_e_dm_alpha_mo`
* :c:data:`data_one_e_dm_beta_mo`
* :c:data:`davidson_criterion`
@ -459,36 +458,8 @@ Index of Providers
* :c:data:`nucl_num`
* :c:data:`nucl_num_shell_aos`
* :c:data:`nuclear_repulsion`
* :c:data:`one_body_dm_alpha_ao_for_dft`
* :c:data:`one_body_dm_alpha_at_r`
* :c:data:`one_body_dm_ao_alpha`
* :c:data:`one_body_dm_ao_beta`
* :c:data:`one_body_dm_average_mo_for_dft`
* :c:data:`one_body_dm_beta_ao_for_dft`
* :c:data:`one_body_dm_beta_at_r`
* :c:data:`one_body_dm_dagger_mo_spin_index`
* :c:data:`one_body_dm_mo`
* :c:data:`one_body_dm_mo_alpha`
* :c:data:`one_body_dm_mo_alpha_average`
* :c:data:`one_body_dm_mo_alpha_for_dft`
* :c:data:`one_body_dm_mo_alpha_one_det`
* :c:data:`one_body_dm_mo_beta`
* :c:data:`one_body_dm_mo_beta_average`
* :c:data:`one_body_dm_mo_beta_for_dft`
* :c:data:`one_body_dm_mo_beta_one_det`
* :c:data:`one_body_dm_mo_diff`
* :c:data:`one_body_dm_mo_for_dft`
* :c:data:`one_body_dm_mo_spin_index`
* :c:data:`one_body_grad_2_dm_alpha_at_r`
* :c:data:`one_body_grad_2_dm_beta_at_r`
* :c:data:`one_body_single_double_dm_mo_alpha`
* :c:data:`one_body_single_double_dm_mo_beta`
* :c:data:`one_body_spin_density_ao`
* :c:data:`one_body_spin_density_mo`
* :c:data:`one_dm_alpha_in_r`
* :c:data:`one_dm_and_grad_alpha_in_r`
* :c:data:`one_dm_and_grad_beta_in_r`
* :c:data:`one_dm_beta_in_r`
* :c:data:`one_e_dm_alpha_ao_for_dft`
* :c:data:`one_e_dm_alpha_at_r`
* :c:data:`one_e_dm_alpha_in_r`
@ -516,7 +487,6 @@ Index of Providers
* :c:data:`one_e_grad_2_dm_beta_at_r`
* :c:data:`one_e_spin_density_ao`
* :c:data:`one_e_spin_density_mo`
* :c:data:`one_electron_energy`
* :c:data:`only_expected_s2`
* :c:data:`output_cpu_time_0`
* :c:data:`output_wall_time_0`
@ -725,7 +695,7 @@ Index of Providers
* :c:data:`trace_v_hxc`
* :c:data:`trace_v_xc`
* :c:data:`transpose`
* :c:data:`two_electron_energy`
* :c:data:`two_e_energy`
* :c:data:`unpaired_alpha_electrons`
* :c:data:`used_weight`
* :c:data:`var_pt2_ratio`
@ -1285,7 +1255,6 @@ Index of Subroutines/Functions
* :c:func:`save_mos_truncated`
* :c:func:`save_natorb`
* :c:func:`save_natural_mos`
* :c:func:`save_one_body_dm`
* :c:func:`save_one_e_dm`
* :c:func:`save_ortho_mos`
* :c:func:`save_ref_determinant`
@ -1326,7 +1295,6 @@ Index of Subroutines/Functions
* :c:func:`tamiser`
* :c:func:`task_done_to_taskserver`
* :c:func:`tasks_done_to_taskserver`
* :c:func:`test_det`
* :c:func:`testteethbuilding`
* :c:func:`total_memory`
* :c:func:`two_e_integrals_index`

6
docs/source/users_guide/excited_states.rst
View File

@ -1,3 +1,5 @@
.. _excited_states:
==============
Excited states
==============
@ -28,3 +30,7 @@ states have |MOs| of comparable quality. This allows for a faster convergence
of excitation energies.
.. seealso::
The documentation of the :c:func:`scf`, :c:func:`cis` and :c:func:`fci` programs.

3
docs/source/users_guide/natural_orbitals.rst
View File

@ -48,4 +48,7 @@ Hands on
Also, all the keywords to read the one- and two-electron integrals on the |MO| basis are set to "None"
in order to be sure not to read integrals not corresponding to the current set of |MOs|.
.. seealso::
The documentation of the :ref:`save_natorb` program.

4
docs/source/users_guide/printing.rst
View File

@ -26,4 +26,8 @@ interactively in :ref:`qp_edit` mode. An alternative is to use the :command:`pri
This program will, by default, print out the first :math:`10^4` determinants whatever the size of the wave function stored in the |EZFIO| folder. If you want to change the number of printed Slater determinants, just change the :option:`determinants n_det_print_wf` keyword using the :ref:`qp_edit` tool.
.. seealso::
The documentation of the :ref:`print_wf` program.

4
docs/source/users_guide/quickstart.rst
View File

@ -74,7 +74,7 @@ The expected energy is ``-92.827856698`` au.
.. seealso::
The documentation of the :ref:`hartree_fock` module.
The documentation of the :ref:`hartree_fock` module and that of the :c:func:`scf` program.
This creates the |MOs| in the |EZFIO| database that will be used to perform any other post-SCF method.
The |qp| does not handle symmetry and the |MOs| are stored by increasing order of Fock energies.
@ -137,7 +137,7 @@ The estimated |FCI| energy of HCN is ``-93.0501`` au.
.. seealso::
The documentation of the :ref:`fci` module.
The documentation of the :ref:`fci` module and that of the :c:func:`fci` program.
---------------------------

29
src/cis/README.rst
View File

@ -2,7 +2,34 @@
cis
===
This module contains a CIS program, built by setting the following rules:
This module contains a CIS program.
The user point of view
----------------------
The :command:`cis` program performs the CI of the ROHF-like + all single excitations on top of it.
This program can be very useful to :
* **Ground state calculations**: generate a guess for the ground state wave function if one is not sure that the :c:func:`scf` program gave the lowest SCF solution. In combination with :c:func:`save_natorb` it can produce new |MOs| in order to reperform an :c:func:`scf` optimization.
* **Excited states calculations**: generate guess for all the :option:`determinants n_states` wave functions, that will be used by the :c:func:`fci` program.
The main keywords/options to be used are:
* :option:`determinants n_states` : number of states to consider for the |CIS| calculation
* :option:`determinants s2_eig` : force all states to have the desired value of :math:`S^2`
* :option:`determinants expected_s2` : desired value of :math:`S^2`
The programmer point of view
----------------------------
This module have been built by setting the following rules:
* The only generator determinant is the Hartree-Fock (single-reference method)
* All generated singly excited determinants are included in the wave function (no perturbative

33
src/cis/cis.irp.f
View File

@ -5,32 +5,35 @@ program cis
!
! This program takes a reference Slater determinant of ROHF-like occupancy,
!
! and performs all single excitations on top of it, disregarding spatial symmetry and compute the "n_states" lowest eigenstates of that CI matrix. (see :option:`determinants n_states`)
! and performs all single excitations on top of it, disregarding spatial
! symmetry and compute the "n_states" lowest eigenstates of that CI matrix.
! (see :option:`determinants n_states`)
!
! This program can be useful in many cases:
!
! * GROUND STATE CALCULATION: to be sure to have the lowest scf solution, perform an scf (see the :ref:`hartree_fock` module), then a cis,
!
! save the natural orbitals (see :ref:`save_natorb`) and reperform an scf optimization from this MO guess
! * GROUND STATE CALCULATION: to be sure to have the lowest scf solution,
! perform an :c:func:`scf` (see the :ref:`hartree_fock` module), then a :c:func:`cis`,
! save the natural orbitals (see :c:func:`save_natorb`) and reperform an :c:func:`scf`
! optimization from this MO guess
!
!
!
! * EXCITED STATES CALCULATIONS: the lowest excited states are much likely to be dominanted by single-excitations.
!
! Therefore, running a cis will save the "n_states" lowest states within the CIS space
!
! in the EZFIO folder, which can afterward be used as guess wave functions for a further
!
! multi-state fci calculation if you specify "read_wf" = True before running the fci executable.
!
! Also, if you specify "s2_eig" = True, the cis will only retain states having the good value :math:`S^2` value (see expected_s2).
!
! * EXCITED STATES CALCULATIONS: the lowest excited states are much likely to
! be dominanted by single-excitations.
! Therefore, running a :c:func:`cis` will save the "n_states" lowest states within
! the CIS space
! in the EZFIO folder, which can afterward be used as guess wave functions
! for a further multi-state fci calculation if you specify "read_wf" = True
! before running the fci executable (see :option:`determinants read_wf`).
! Also, if you specify "s2_eig" = True, the cis will only retain states
! having the good value :math:`S^2` value (see :option:`determinants
! s2_eig` and :option:`determinants expected_s2`).
! If "s2_eig" = False, it will take the lowest n_states, whatever multiplicity they are.
!
!
!
! Note: if you would like to discard some orbitals, use qp_set_mo_class to specify:
! Note: if you would like to discard some orbitals, use :ref:`qp_set_mo_class` to specify:
!
! * "core" orbitals which will be always doubly occupied
!

27
src/cisd/README.rst
View File

@ -2,7 +2,32 @@
cisd
====
This module contains a CISD program, built by setting the following rules:
This module contains a CI of single and double excitations.
The user point of view
----------------------
The :command:`cisd` program performs the CI of the ROHF-like + all single and double excitations on top of it.
This program can be very useful to :
* **Ground state calculations**: generate a guess for the ground state wave function if one is not sure that the :c:func:`scf` program gave the lowest SCF solution. In combination with :c:func:`save_natorb` it can produce new |MOs| in order to reperform an :c:func:`scf` optimization.
* **Excited states calculations**: generate guess for all the :option:`determinants n_states` wave functions, that will be used by the :c:func:`fci` program.
The main keywords/options to be used are:
* :option:`determinants n_states` : number of states to consider for the |cisd| calculation
* :option:`determinants s2_eig` : force all states to have the desired value of :math:`S^2`
* :option:`determinants expected_s2` : desired value of :math:`S^2`
The programmer point of view
----------------------------
This module have been built by setting the following rules:
* The only generator determinant is the Hartree-Fock (single-reference method)
* All generated determinants are included in the wave function (no perturbative

39
src/cisd/cisd.irp.f
View File

@ -5,32 +5,37 @@ program cisd
!
! This program takes a reference Slater determinant of ROHF-like occupancy,
!
! and performs all single and double excitations on top of it, disregarding spatial symmetry and compute the "n_states" lowest eigenstates of that CI matrix.
! and performs all single and double excitations on top of it, disregarding
! spatial symmetry and compute the "n_states" lowest eigenstates of that CI
! matrix (see :option:`determinants n_states`).
!
! This program can be useful in many cases:
!
! * GROUND STATE CALCULATION: if even after a cis calculation, natural orbitals and then scf optimization, you are not sure to have the lowest scf solution,
! do the same strategy with the cisd executable to generate the natural orbitals as a guess for the scf.
! * GROUND STATE CALCULATION: if even after a :c:func:`cis` calculation, natural
! orbitals (see :c:func:`save_natorb`) and then :c:func:`scf` optimization, you are not sure to have the lowest scf
! solution,
! do the same strategy with the :c:func:`cisd` executable instead of the :c:func:`cis` exectuable to generate the natural
! orbitals as a guess for the :c:func:`scf`.
!
!
!
! * EXCITED STATES CALCULATIONS: the lowest excited states are much likely to be dominanted by single- or double-excitations.
!
! Therefore, running a cis will save the "n_states" lowest states within the CISD space
!
! in the EZFIO folder, which can afterward be used as guess wave functions for a further
!
! multi-state fci calculation if you specify "read_wf" = True before running the fci executable.
!
! Also, if you specify "s2_eig" = True, the cisd will only retain states having the good value :math:`S^2` value (see expected_s2).
!
! If "s2_eig" = False, it will take the lowest n_states, whatever multiplicity they are.
! * EXCITED STATES CALCULATIONS: the lowest excited states are much likely to
! be dominanted by single- or double-excitations.
! Therefore, running a :c:func:`cisd` will save the "n_states" lowest states within
! the CISD space
! in the EZFIO folder, which can afterward be used as guess wave functions
! for a further multi-state fci calculation if you specify "read_wf" = True
! before running the fci executable (see :option:`determinants read_wf`).
! Also, if you specify "s2_eig" = True, the cisd will only retain states
! having the good value :math:`S^2` value
! (see :option:`determinants expected_s2` and :option:`determinants s2_eig`).
! If "s2_eig" = False, it will take the lowest n_states, whatever
! multiplicity they are.
!
!
!
! Note: if you would like to discard some orbitals, use qp_set_mo_class to specify:
! Note: if you would like to discard some orbitals, use
! :ref:`qp_set_mo_class` to specify:
!
! * "core" orbitals which will be always doubly occupied
!

39
src/fci/README.rst
View File

@ -2,12 +2,47 @@
fci
===
|CIPSI| algorithm in the full configuration interaction space.
* :command:`fci` performs the |CIPSI| calculation,
* :command:`pt2` computes the |PT2| contribution using the wave function stored in the |EZFIO|
The user point of view
----------------------
* :c:func:`fci` performs |CIPSI| calculations using a stochastic scheme for both the selection and the |PT2| contribution,
* :c:func:`pt2` computes the |PT2| contribution using the wave function stored in the |EZFIO|
database.
The main keywords/options for this module are:
* :option:`determinants n_det_max` : maximum number of Slater determinants in the CIPSI wave function. The :command:`fci` program will stop when the size of the CIPSI wave function will exceed :option:`determinants n_det_max`.
* :option:`perturbation pt2_max` : absolute value of the |PT2| to stop the CIPSI calculation. Once the |PT2| :math:`<` :option:`perturbation pt2_max`, the CIPSI calculation stops.
* :option:`determinants n_states` : number of states to consider in the CIPSI calculation.
* :option:`determinants read_wf` : if False, starts with a ROHF-like determinant, if True, starts with the current wave function(s) stored in the |EZFIO| folder.
.. note::
For a multi-state calculation, it is recommended to start with :c:func:`cis` or :c:func:`cisd`
wave functions as a guess.
* :option:`determinants s2_eig` : if True, systematically add all the determinants needed to have a pure value of :math:`S^2`. Also, if True, it tracks only the states having the good :option:`determinants expected_s2`.
.. note::
For a multi-state calculation, it is recommended to start with :c:func:`cis` or :c:func:`cisd`
wave functions as a guess.
* :option:`determinants expected_s2` : expected value of :math:`S^2` for the desired spin multiplicity.
The programmer point of view
----------------------------
This module have been created with the :ref:`cipsi` module.
.. seealso::
The documentation of the :ref:`cipsi` module.

4
src/fci/fci.irp.f
View File

@ -1,7 +1,7 @@
program fci
implicit none
BEGIN_DOC
! Selected Full Configuration Interaction with stochastic selection and PT2.
! Selected Full Configuration Interaction with stochastic selection and PT2 .
!
! This program performs a CIPSI-like selected CI using a stochastic scheme for both
!
@ -33,7 +33,7 @@ program fci
!
!
!
! For excited states calculations, it is recommended to start with a cis or cisd guess wave functions
! For excited states calculations, it is recommended to start with a :c:func:`cis` or :c:func:`cisd` guess wave functions
!
! for the "n_states", and to set "s2_eig" = True.
END_DOC

6
src/tools/diagonalize_h.irp.f
View File

@ -1,7 +1,11 @@
program diagonalize_h
implicit none
BEGIN_DOC
! program that extracts the `n_states` lowest states of the Hamiltonian within the set of Slater determinants stored in the EZFIO folder
! Program that extracts the :option:`determinants n_states` lowest states of the Hamiltonian within the set of Slater determinants stored in the EZFIO folder.
!
! If :option:`determinants s2_eig` = True, it will retain only states
!
! which corresponds to the desired value of :option:`determinants expected_s2`.
END_DOC
read_wf = .True.
touch read_wf

4
src/tools/fcidump.irp.f
View File

@ -1,9 +1,9 @@
program fcidump
implicit none
BEGIN_DOC
! Produce a regular FCIDUMP file from the MOs stored in the EZFIO folder.
! Produce a regular FCIDUMP file from the |MOs| stored in the |EZFIO| folder.
!
! To specify an active space, the class of the mos have to set in the EZFIO folder (see set_mo_class/qp_set_mo_class).
! To specify an active space, the class of the mos have to set in the |EZFIO| folder (see :ref:`qp_set_mo_class`).
!
! The fcidump program supports 3 types of MO_class :
!

4
src/tools/four_idx_transform.irp.f
View File

@ -1,9 +1,9 @@
program four_idx_transform
implicit none
BEGIN_DOC
! 4-index transformation of two-electron integrals from AO to MO integrals.
! 4-index transformation of two-electron integrals from |AO| to |MO| integrals.
!
! This program will compute the two-electron integrals on the MO basis and store it into the EZFIO folder.
! This program will compute the two-electron integrals on the |MO| basis and store it into the |EZFIO| folder.
!
! This program can be useful if the AO --> MO transformation is an expensive step by itself.
END_DOC

7
src/tools/print_e_conv.irp.f
View File

@ -5,7 +5,7 @@ program print_e_conv
!
! for all istate, this program produces
!
! * a file "EZFIO.$istate.conv" containing the variational and var+PT2 energies as a function of N_det
! * a file "EZFIO.istate.conv" containing the variational and var+PT2 energies as a function of N_det
!
! * for istate > 1, a file EZFIO.$istate.delta_e.conv containing the energy difference (both var and var+PT2) with the ground state as a function of N_det
END_DOC
@ -14,8 +14,11 @@ program print_e_conv
call routine
end
subroutine routine
subroutine routine_e_conv
implicit none
BEGIN_DOC
! routine called by :c:func:`print_e_conv`
END_DOC
integer :: N_iter_tmp
integer :: i,istate
character*(128) :: output

4
src/tools/print_wf.irp.f
View File

@ -1,9 +1,9 @@
program print_wf
implicit none
BEGIN_DOC
! print the wave function stored in the EZFIO folder in the intermediate normalization.
! Print the ground state wave function stored in the |EZFIO| folder in the intermediate normalization.
!
! it also prints a lot of information regarding the excitation operators from the reference determinant
! It also prints a lot of information regarding the excitation operators from the reference determinant
!
! and a first-order perturbative analysis of the wave function.
!

7
src/tools/save_one_e_dm.irp.f
View File

@ -15,8 +15,11 @@ program save_one_e_dm
end
subroutine routine
subroutine routine_save_one_e_dm
implicit none
BEGIN_DOC
! routine called by :c:func:`save_one_e_dm`
END_DOC
call ezfio_set_aux_quantities_data_one_e_dm_alpha_mo(one_e_dm_mo_alpha)
call ezfio_set_aux_quantities_data_one_e_dm_beta_mo(one_e_dm_mo_beta)
end