Electrons
Table of Contents
1. Context
The electron module stores both the fundamental electronic configuration (number and positions of electrons) and computed quantities (distances, potential energies) in the context. The walker mechanism allows multiple electronic configurations to be handled simultaneously, which is essential for quantum Monte Carlo calculations.
1.1. Data structure
The following data is stored in the context:
| Variable | Type | Description |
|---|---|---|
uninitialized |
int32_t |
Keeps bit set for uninitialized data |
num |
int64_t |
Total number of electrons |
up_num |
int64_t |
Number of up-spin electrons |
down_num |
int64_t |
Number of down-spin electrons |
provided |
bool |
If true, electron is valid |
walker |
qmckl_point |
Current set of walkers |
walker_old |
qmckl_point |
Previous set of walkers |
Computed data:
| Variable | Type | Description |
|---|---|---|
ee_distance |
double[walker.num][num][num] |
Electron-electron distances |
ee_distance_date |
uint64_t |
Last modification date of the electron-electron distances |
en_distance |
double[num][nucl_num] |
Electron-nucleus distances |
en_distance_date |
uint64_t |
Last modification date of the electron-electron distances |
ee_potential |
double[walker.num] |
Electron-electron potential energy |
ee_potential_date |
uint64_t |
Last modification date of the electron-electron potential |
en_potential |
double[walker.num] |
Electron-nucleus potential energy |
en_potential_date |
int64_t |
Date when the electron-nucleus potential energy was computed |
1.2. Data structures
The electron data is organized using two levels of structures. The qmckl_walker
structure represents a collection of electronic configurations (walkers) being
sampled in the Monte Carlo process. Each walker contains multiple electrons at
specific positions.
The qmckl_electron_struct contains all electron-related data, including the
current and previous walker states (for computing acceptance ratios), and
cached computed quantities like inter-particle distances and potential energies.
The date fields enable automatic invalidation of cached results when electron
positions change.
The uninitialized integer contains one bit set to one for each
initialization function which has not been called. It becomes equal
to zero after all initialization functions have been called. The
struct is then initialized and provided == true.
Some values are initialized by default, and are not concerned by
this mechanism.
bool qmckl_electron_provided (const qmckl_context context);
1.3. Initialization functions
To set the data relative to the electrons in the context, the
following functions need to be called. When the data structure is
initialized, the internal coord_new and coord_old arrays are
both not allocated.
qmckl_exit_code qmckl_set_electron_num (qmckl_context context, const int64_t up_num, const int64_t down_num); qmckl_exit_code qmckl_set_electron_coord (qmckl_context context, const char transp, const int64_t walk_num, const double* coord, const int64_t size_max);
To set the number of electrons, we give the number of up-spin and down-spin electrons to the context and we set the number of walkers.
The following function sets the electron coordinates of all the walkers. When this is done, the pointers to the old and new sets of coordinates are swapped, and the new coordinates are overwritten. This can be done only when the data relative to electrons have been set.
size_max should be equal equal or geater than elec_num *
walker.num * 3, to be symmetric with qmckl_get_electron_coord.
Important: changing the electron coordinates increments the date in the context.
1.4. Access functions
Access functions return QMCKL_SUCCESS when the data has been
successfully retrieved. It returnes QMCKL_INVALID_CONTEXT when
the context is not a valid context, and QMCKL_NOT_PROVIDED when
the data has not been provided. If the function returns
successfully, the variable pointed by the pointer given in argument
contains the requested data. Otherwise, this variable is untouched.
1.4.1. Number of electrons
qmckl_exit_code qmckl_get_electron_num (const qmckl_context context, int64_t* const num); qmckl_exit_code qmckl_get_electron_up_num (const qmckl_context context, int64_t* const up_num); qmckl_exit_code qmckl_get_electron_down_num (const qmckl_context context, int64_t* const down_num);
1.4.2. Number of walkers
A walker is a set of electron coordinates that are arguments of
the wave function. walk_num is the number of walkers.
qmckl_exit_code qmckl_get_electron_walk_num (const qmckl_context context, int64_t* const walk_num);
1.4.3. Electron coordinates
Returns the current electron coordinates. The pointer is assumed
to point on a memory block of size size_max ≥ 3 * elec_num * walker.num.
The order of the indices is:
| Normal | Transposed | |
|---|---|---|
| C | [walker.num*elec_num][3] |
[3][walker.num*elec_num] |
| Fortran | (3,walker.num*elec_num) |
(walker.num*elec_num, 3) |
qmckl_exit_code qmckl_get_electron_coord (const qmckl_context context, const char transp, double* const coord, const int64_t size_max);
As the walker attribute is equal to points, returning the
current electron coordinates is equivalent to returning the
current points.
2. Computation
The computed data is stored in the context so that it can be reused by different kernels. To ensure that the data is valid, for each computed data the date of the context is stored when it is computed. To know if some data needs to be recomputed, we check if the date of the dependencies are more recent than the date of the data to compute. If it is the case, then the data is recomputed and the current date is stored.
2.1. Electron-electron distances
2.1.1. Get
qmckl_exit_code qmckl_get_electron_ee_distance(qmckl_context context, double* const distance, const int64_t size_max);
2.1.2. Compute
| Variable | Type | In/Out | Description |
|---|---|---|---|
context |
qmckl_context |
in | Global state |
elec_num |
int64_t |
in | Number of electrons |
walk_num |
int64_t |
in | Number of walkers |
coord |
double[3][walk_num][elec_num] |
in | Electron coordinates |
ee_distance |
double[walk_num][elec_num][elec_num] |
out | Electron-electron distances |
function qmckl_compute_ee_distance(context, elec_num, walk_num, coord, ee_distance) & result(info) bind(C) use qmckl_constants use qmckl, only : qmckl_distance implicit none integer (qmckl_context) , intent(in) , value :: context integer (c_int64_t) , intent(in) , value :: elec_num integer (c_int64_t) , intent(in) , value :: walk_num real (c_double ) , intent(in) :: coord(elec_num,3,walk_num) real (c_double ) , intent(out) :: ee_distance(elec_num,elec_num,walk_num) integer (qmckl_exit_code) :: info integer*8 :: k, i, j double precision :: x, y, z info = QMCKL_SUCCESS if (context == QMCKL_NULL_CONTEXT) then info = QMCKL_INVALID_CONTEXT return endif if (elec_num <= 0) then info = QMCKL_INVALID_ARG_2 return endif if (walk_num <= 0) then info = QMCKL_INVALID_ARG_3 return endif do k=1,walk_num info = qmckl_distance(context, 'T', 'T', elec_num, elec_num, & coord(1,k,1), elec_num * walk_num, & coord(1,k,1), elec_num * walk_num, & ee_distance(1,1,k), elec_num) if (info /= QMCKL_SUCCESS) then exit endif end do end function qmckl_compute_ee_distance
2.2. Electron-electron potential
ee_potential is given by
\[ \mathcal{V}_{ee} = \sum_{i=1}^{N_e}\sum_{j>i}^{N_e}\frac{1}{r_{ij}} \]
where \(\mathcal{V}_{ee}\) is the ee potential and \(r_{ij}\) the ee
distance.
2.2.1. Get
qmckl_exit_code qmckl_get_electron_ee_potential(qmckl_context context, double* const ee_potential);
2.2.2. Compute
| Variable | Type | In/Out | Description |
|---|---|---|---|
context |
qmckl_context |
in | Global state |
elec_num |
int64_t |
in | Number of electrons |
walk_num |
int64_t |
in | Number of walkers |
ee_distance |
double[walk_num][elec_num][elec_num] |
in | Electron-electron distances |
ee_potential |
double[walk_num] |
out | Electron-electron potential |
function qmckl_compute_ee_potential(context, elec_num, walk_num, & ee_distance, ee_potential) result(info) bind(C) use qmckl_constants implicit none integer(qmckl_context), intent(in), value :: context integer (c_int64_t) , intent(in) , value :: elec_num integer (c_int64_t) , intent(in) , value :: walk_num real (c_double ) , intent(in) :: ee_distance(elec_num,elec_num,walk_num) real (c_double ) , intent(out) :: ee_potential(walk_num) integer (qmckl_exit_code) :: info integer*8 :: nw, i, j info = QMCKL_SUCCESS if (context == QMCKL_NULL_CONTEXT) then info = QMCKL_INVALID_CONTEXT return endif if (elec_num <= 0) then info = QMCKL_INVALID_ARG_2 return endif if (walk_num <= 0) then info = QMCKL_INVALID_ARG_3 return endif ee_potential = 0.0d0 do nw=1,walk_num do j=2,elec_num do i=1,j-1 if (dabs(ee_distance(i,j,nw)) > 1e-5) then ee_potential(nw) = ee_potential(nw) + 1.0d0/(ee_distance(i,j,nw)) endif end do end do end do end function qmckl_compute_ee_potential
2.3. Electron-nucleus distances
2.3.1. Get
qmckl_exit_code qmckl_get_electron_en_distance(qmckl_context context, double* const distance, const int64_t size_max);
2.3.2. Compute
| Variable | Type | In/Out | Description |
|---|---|---|---|
context |
qmckl_context |
in | Global state |
point_num |
int64_t |
in | Number of points |
nucl_num |
int64_t |
in | Number of nuclei |
elec_coord |
double[3][point_num] |
in | Electron coordinates |
nucl_coord |
double[3][nucl_num] |
in | Nuclear coordinates |
en_distance |
double[point_num][nucl_num] |
out | Electron-nucleus distances |
function qmckl_compute_en_distance(context, & point_num, nucl_num, elec_coord, nucl_coord, en_distance) & result(info) bind(C) use qmckl_constants use qmckl, only : qmckl_distance implicit none integer (qmckl_context), intent(in), value :: context integer (c_int64_t) , intent(in), value :: point_num integer (c_int64_t) , intent(in), value :: nucl_num real (c_double ) , intent(in) :: elec_coord(point_num,3) real (c_double ) , intent(in) :: nucl_coord(nucl_num,3) real (c_double ) , intent(out) :: en_distance(nucl_num,point_num) integer(qmckl_exit_code) :: info info = QMCKL_SUCCESS if (context == QMCKL_NULL_CONTEXT) then info = QMCKL_INVALID_CONTEXT return endif if (point_num <= 0) then info = QMCKL_INVALID_ARG_2 return endif if (nucl_num <= 0) then info = QMCKL_INVALID_ARG_3 return endif info = qmckl_distance(context, 'T', 'T', nucl_num, point_num, & nucl_coord, nucl_num, & elec_coord, point_num, & en_distance, nucl_num) end function qmckl_compute_en_distance
2.4. Electron-nucleus potential
en_potential computes the en potential energy
\[ \mathcal{V}_{en} = -\sum_{i=1}^{N_e}\sum_{A=1}^{N_n}\frac{Z_A}{r_{iA}} \]
where \(\mathcal{V}_{en}\) is the en potential, \(r_{iA}\) the en
distance and \(Z_A\) is the nuclear charge.
2.4.1. Get
qmckl_exit_code qmckl_get_electron_en_potential(qmckl_context context, double* const en_potential);
2.4.2. Compute
| Variable | Type | In/Out | Description |
|---|---|---|---|
context |
qmckl_context |
in | Global state |
elec_num |
int64_t |
in | Number of electrons |
nucl_num |
int64_t |
in | Number of nuclei |
walk_num |
int64_t |
in | Number of walkers |
charge |
double[nucl_num] |
in | charge of nucleus |
en_distance |
double[walk_num][elec_num][nucl_num] |
in | Electron-electron distances |
en_potential |
double[walk_num] |
out | Electron-electron potential |
function qmckl_compute_en_potential(context, elec_num, nucl_num, walk_num, & charge, en_distance, en_potential) & result(info) bind(C) use qmckl implicit none integer (qmckl_context), intent(in), value :: context integer (c_int64_t) , intent(in) , value :: elec_num integer (c_int64_t) , intent(in) , value :: nucl_num integer (c_int64_t) , intent(in) , value :: walk_num real (c_double ) , intent(in) :: charge(nucl_num) real (c_double ) , intent(in) :: en_distance(nucl_num,elec_num,walk_num) real (c_double ) , intent(out) :: en_potential(walk_num) integer(qmckl_exit_code) :: info integer*8 :: nw, i, j info = QMCKL_SUCCESS if (context == QMCKL_NULL_CONTEXT) then info = QMCKL_INVALID_CONTEXT return endif if (elec_num <= 0) then info = QMCKL_INVALID_ARG_2 return endif if (walk_num <= 0) then info = QMCKL_INVALID_ARG_3 return endif en_potential = 0.0d0 do nw=1,walk_num do i=1,elec_num do j=1,nucl_num if (dabs(en_distance(j,i,nw)) > 1.d-6) then en_potential(nw) = en_potential(nw) - charge(j)/(en_distance(j,i,nw)) endif end do end do end do end function qmckl_compute_en_potential
qmckl_exit_code qmckl_compute_en_potential ( const qmckl_context context, const int64_t elec_num, const int64_t nucl_num, const int64_t walk_num, const double* charge, const double* en_distance, double* const en_potential );