more functionality for fix ave/grid
This commit is contained in:
@ -31,10 +31,18 @@
|
||||
|
||||
using namespace LAMMPS_NS;
|
||||
using namespace FixConst;
|
||||
|
||||
enum{SAMPLE,ALL};
|
||||
enum{NOSCALE,ATOM};
|
||||
enum{ONE,RUNNING,WINDOW};
|
||||
|
||||
// OFFSET avoids outside-of-box atoms being rounded to grid pts incorrectly
|
||||
// SHIFT = 0.0 assigns atoms to lower-left grid pt
|
||||
// SHIFT = 0.5 assigns atoms to nearest grid pt
|
||||
|
||||
static constexpr int OFFSET = 16384;
|
||||
static constexpr double SHIFT = 0.5;
|
||||
|
||||
/* ---------------------------------------------------------------------- */
|
||||
|
||||
FixAveGrid::FixAveGrid(LAMMPS *lmp, int narg, char **arg) :
|
||||
@ -52,11 +60,11 @@ FixAveGrid::FixAveGrid(LAMMPS *lmp, int narg, char **arg) :
|
||||
pergrid_freq = utils::inumeric(FLERR,arg[5],false,lmp);
|
||||
time_depend = 1;
|
||||
|
||||
// NOTE: test for Dxyz as well
|
||||
// NOTE: allow Dxyz as well
|
||||
|
||||
nx = utils::inumeric(FLERR,arg[6],false,lmp);
|
||||
ny = utils::inumeric(FLERR,arg[7],false,lmp);
|
||||
nz = utils::inumeric(FLERR,arg[8],false,lmp);
|
||||
nxgrid = utils::inumeric(FLERR,arg[6],false,lmp);
|
||||
nygrid = utils::inumeric(FLERR,arg[7],false,lmp);
|
||||
nzgrid = utils::inumeric(FLERR,arg[8],false,lmp);
|
||||
|
||||
// expand args if any have wildcard character "*"
|
||||
// this can reset nvalues
|
||||
@ -208,9 +216,9 @@ FixAveGrid::FixAveGrid(LAMMPS *lmp, int narg, char **arg) :
|
||||
if (pergrid_freq % nevery || nrepeat*nevery > pergrid_freq)
|
||||
error->all(FLERR,"Illegal fix ave/grid command");
|
||||
|
||||
if (nx < 1 || ny < 1 || nz < 1)
|
||||
if (nxgrid < 1 || nygrid < 1 || nzgrid < 1)
|
||||
error->all(FLERR,"Invalid fix ave/grid grid size");
|
||||
if (dimension == 2 && nz != 1)
|
||||
if (dimension == 2 && nzgrid != 1)
|
||||
error->all(FLERR,"Fix ave/grid grid Nz must be 1 for 2d simulation");
|
||||
|
||||
// error checks for ATOM mode
|
||||
@ -370,8 +378,12 @@ FixAveGrid::FixAveGrid(LAMMPS *lmp, int narg, char **arg) :
|
||||
// instantiate the Grid class and allocate per-grid memory
|
||||
// NOTE: need to extend ghost grid for ATOM mode ?
|
||||
|
||||
if (modeatom) shift = OFFSET + SHIFT;
|
||||
else shift = 0.0;
|
||||
|
||||
if (dimension == 2) {
|
||||
grid2d = new Grid2d(lmp, world, nx, ny, 0, 0.0, 0.0,
|
||||
if (modeatom)
|
||||
grid2d = new Grid2d(lmp, world, nxgrid, nygrid, 0, 0.0, shift,
|
||||
nxlo_in, nxhi_in, nylo_in, nyhi_in,
|
||||
nxlo_out, nxhi_out, nylo_out, nyhi_out);
|
||||
if (nvalues == 1)
|
||||
@ -383,7 +395,7 @@ FixAveGrid::FixAveGrid(LAMMPS *lmp, int narg, char **arg) :
|
||||
ngridout = (nxhi_out - nxlo_out + 1) * (nyhi_out - nylo_out + 1);
|
||||
|
||||
} else {
|
||||
grid3d = new Grid3d(lmp, world, nx, ny, nz, 0, 0.0, 0.0,
|
||||
grid3d = new Grid3d(lmp, world, nxgrid, nygrid, nzgrid, 0, 0.0, shift,
|
||||
nxlo_in, nxhi_in, nylo_in, nyhi_in, nzlo_in, nzhi_in,
|
||||
nxlo_out, nxhi_out, nylo_out, nyhi_out,
|
||||
nzlo_out, nzhi_out);
|
||||
@ -399,15 +411,14 @@ FixAveGrid::FixAveGrid(LAMMPS *lmp, int narg, char **arg) :
|
||||
(nzhi_out - nzlo_out + 1);
|
||||
}
|
||||
|
||||
// zero the array since dump may access it on timestep 0
|
||||
// zero the array since a variable may access it before first run
|
||||
// zero the grid since dump may access it on timestep 0
|
||||
|
||||
/*
|
||||
int nlocal = atom->nlocal;
|
||||
for (int i = 0; i < nlocal; i++)
|
||||
for (int m = 0; m < nvalues; m++)
|
||||
array[i][m] = 0.0;
|
||||
*/
|
||||
zero_grid();
|
||||
|
||||
// bin indices for ATOM mode
|
||||
|
||||
maxatom = 0;
|
||||
bin = nullptr;
|
||||
|
||||
// nvalid = next step on which end_of_step does something
|
||||
// add nvalid to all computes that store invocation times
|
||||
@ -438,6 +449,7 @@ FixAveGrid::~FixAveGrid()
|
||||
memory->destroy2d_offset(array2d,nylo_out,nxlo_out);
|
||||
memory->destroy3d_offset(vec3d,nzlo_out,nylo_out,nxlo_out);
|
||||
memory->destroy4d_offset_last(array3d,nzlo_out,nylo_out,nxlo_out);
|
||||
memory->destroy(bin);
|
||||
}
|
||||
|
||||
/* ---------------------------------------------------------------------- */
|
||||
@ -480,8 +492,8 @@ void FixAveGrid::init()
|
||||
// check that grid sizes for all fields match grid size for this fix
|
||||
|
||||
if (modegrid) {
|
||||
Compute *icompute;
|
||||
Fix *ifix;
|
||||
Compute *compute;
|
||||
Fix *fix;
|
||||
Grid2d *grid2d;
|
||||
Grid3d *grid3d;
|
||||
|
||||
@ -490,26 +502,26 @@ void FixAveGrid::init()
|
||||
for (int m = 0; m < nvalues; m++) {
|
||||
if (dimension == 2) {
|
||||
if (which[m] == ArgInfo::COMPUTE) {
|
||||
icompute = modify->compute[value2index[m]];
|
||||
grid2d = (Grid2d *) icompute->get_grid_by_index(value2grid[m]);
|
||||
compute = modify->compute[value2index[m]];
|
||||
grid2d = (Grid2d *) compute->get_grid_by_index(value2grid[m]);
|
||||
} else {
|
||||
ifix = modify->fix[value2index[m]];
|
||||
grid2d = (Grid2d *) ifix->get_grid_by_index(value2grid[m]);
|
||||
fix = modify->fix[value2index[m]];
|
||||
grid2d = (Grid2d *) fix->get_grid_by_index(value2grid[m]);
|
||||
}
|
||||
grid2d->get_size(nxtmp,nytmp);
|
||||
if (nxtmp != nx || nytmp != ny)
|
||||
if (nxtmp != nxgrid || nytmp != nygrid)
|
||||
error->all(FLERR,"Fix ave/grid value grid sizes do not match");
|
||||
|
||||
} else {
|
||||
if (which[m] == ArgInfo::COMPUTE) {
|
||||
icompute = modify->compute[value2index[m]];
|
||||
grid3d = (Grid3d *) icompute->get_grid_by_index(value2grid[m]);
|
||||
compute = modify->compute[value2index[m]];
|
||||
grid3d = (Grid3d *) compute->get_grid_by_index(value2grid[m]);
|
||||
} else {
|
||||
ifix = modify->fix[value2index[m]];
|
||||
grid3d = (Grid3d *) ifix->get_grid_by_index(value2grid[m]);
|
||||
fix = modify->fix[value2index[m]];
|
||||
grid3d = (Grid3d *) fix->get_grid_by_index(value2grid[m]);
|
||||
}
|
||||
grid3d->get_size(nxtmp,nytmp,nztmp);
|
||||
if (nxtmp != nx || nytmp != ny || nztmp != nz)
|
||||
if (nxtmp != nxgrid || nytmp != nygrid || nztmp != nzgrid)
|
||||
error->all(FLERR,"Fix ave/grid value grid sizes do not match");
|
||||
}
|
||||
}
|
||||
@ -537,7 +549,7 @@ void FixAveGrid::setup(int /*vflag*/)
|
||||
|
||||
void FixAveGrid::end_of_step()
|
||||
{
|
||||
int i,j,m,n;
|
||||
int i,j,m,n,ix,iy,iz;
|
||||
|
||||
// skip if not step which requires doing something
|
||||
|
||||
@ -545,93 +557,332 @@ void FixAveGrid::end_of_step()
|
||||
if (ntimestep != nvalid) return;
|
||||
nvalid_last = nvalid;
|
||||
|
||||
// zero if first step
|
||||
|
||||
/*
|
||||
if (irepeat == 0)
|
||||
for (i = 0; i < nlocal; i++)
|
||||
for (m = 0; m < nvalues; m++)
|
||||
array[i][m] = 0.0;
|
||||
*/
|
||||
// zero owned and ghost grid points if first step
|
||||
|
||||
if (irepeat == 0) zero_grid();
|
||||
|
||||
// ATOM mode
|
||||
// accumulate per-atom attributes,computes,fixes,variables to local grid
|
||||
// compute/fix/variable may invoke computes so wrap with clear/add
|
||||
|
||||
/*
|
||||
|
||||
if (modeatom) {
|
||||
modify->clearstep_compute();
|
||||
modify->clearstep_compute();
|
||||
|
||||
int *mask = atom->mask;
|
||||
// bin[i][dim] = indices of bin each atom is in
|
||||
// not set if group mask does not match
|
||||
// NOTE: error check if atom out of grid bounds?
|
||||
|
||||
for (m = 0; m < nvalues; m++) {
|
||||
n = value2index[m];
|
||||
j = argindex[m];
|
||||
double *boxlo = domain->boxlo;
|
||||
double dxinv = nxgrid/domain->xprd;
|
||||
double dyinv = nygrid/domain->yprd;
|
||||
double dzinv = nzgrid/domain->zprd;
|
||||
|
||||
if (which[m] == ArgInfo::X) {
|
||||
double **x = atom->x;
|
||||
for (i = 0; i < nlocal; i++)
|
||||
if (mask[i] & groupbit) array[i][m] += x[i][j];
|
||||
double **x = atom->x;
|
||||
int *mask = atom->mask;
|
||||
int nlocal = atom->nlocal;
|
||||
|
||||
} else if (which[m] == ArgInfo::V) {
|
||||
double **v = atom->v;
|
||||
for (i = 0; i < nlocal; i++)
|
||||
if (mask[i] & groupbit) array[i][m] += v[i][j];
|
||||
if (nlocal > maxatom) {
|
||||
memory->destroy(bin);
|
||||
maxatom = atom->nmax;
|
||||
memory->create(bin,maxatom,dimension,"fix_ave/grid:bin");
|
||||
}
|
||||
|
||||
} else if (which[m] == ArgInfo::F) {
|
||||
double **f = atom->f;
|
||||
for (i = 0; i < nlocal; i++)
|
||||
if (mask[i] & groupbit) array[i][m] += f[i][j];
|
||||
|
||||
// invoke compute if not previously invoked
|
||||
|
||||
} else if (which[m] == ArgInfo::COMPUTE) {
|
||||
Compute *compute = modify->compute[n];
|
||||
if (!(compute->invoked_flag & Compute::INVOKED_PERATOM)) {
|
||||
compute->compute_peratom();
|
||||
compute->invoked_flag |= Compute::INVOKED_PERATOM;
|
||||
if (dimension == 2) {
|
||||
for (i = 0; i < nlocal; i++) {
|
||||
if (!(mask[i] & groupbit)) continue;
|
||||
ix = static_cast<int> ((x[i][0]-boxlo[0])*dxinv + shift) - OFFSET;
|
||||
iy = static_cast<int> ((x[i][1]-boxlo[1])*dyinv + shift) - OFFSET;
|
||||
bin[i][0] = iy;
|
||||
bin[i][1] = ix;
|
||||
}
|
||||
|
||||
if (j == 0) {
|
||||
double *compute_vector = compute->vector_atom;
|
||||
for (i = 0; i < nlocal; i++)
|
||||
if (mask[i] & groupbit) array[i][m] += compute_vector[i];
|
||||
} else {
|
||||
int jm1 = j - 1;
|
||||
double **compute_array = compute->array_atom;
|
||||
for (i = 0; i < nlocal; i++)
|
||||
if (mask[i] & groupbit) array[i][m] += compute_array[i][jm1];
|
||||
} else {
|
||||
for (i = 0; i < nlocal; i++) {
|
||||
if (!(mask[i] & groupbit)) continue;
|
||||
ix = static_cast<int> ((x[i][0]-boxlo[0])*dxinv + shift) - OFFSET;
|
||||
iy = static_cast<int> ((x[i][1]-boxlo[1])*dyinv + shift) - OFFSET;
|
||||
iz = static_cast<int> ((x[i][2]-boxlo[2])*dzinv + shift) - OFFSET;
|
||||
bin[i][0] = iz;
|
||||
bin[i][1] = iy;
|
||||
bin[i][2] = ix;
|
||||
}
|
||||
}
|
||||
|
||||
// access fix fields, guaranteed to be ready
|
||||
for (m = 0; m < nvalues; m++) {
|
||||
n = value2index[m];
|
||||
j = argindex[m];
|
||||
|
||||
if (which[m] == ArgInfo::X) {
|
||||
if (dimension == 2) {
|
||||
if (nvalues == 1) {
|
||||
for (i = 0; i < nlocal; i++) {
|
||||
if (mask[i] & groupbit)
|
||||
vec2d[bin[i][0]][bin[i][1]] += x[i][j];
|
||||
}
|
||||
} else
|
||||
for (i = 0; i < nlocal; i++) {
|
||||
if (mask[i] & groupbit)
|
||||
array2d[bin[i][0]][bin[i][1]][m] += x[i][j];
|
||||
}
|
||||
} else {
|
||||
if (nvalues == 1) {
|
||||
for (i = 0; i < nlocal; i++) {
|
||||
if (mask[i] & groupbit)
|
||||
vec3d[bin[i][0]][bin[i][1]][bin[i][2]] += x[i][j];
|
||||
}
|
||||
} else
|
||||
for (i = 0; i < nlocal; i++) {
|
||||
if (mask[i] & groupbit)
|
||||
array3d[bin[i][0]][bin[i][1]][bin[i][2]][m] += x[i][j];
|
||||
}
|
||||
}
|
||||
|
||||
} else if (which[m] == ArgInfo::V) {
|
||||
double **v = atom->v;
|
||||
|
||||
} else if (which[m] == ArgInfo::FIX) {
|
||||
if (j == 0) {
|
||||
double *fix_vector = modify->fix[n]->vector_atom;
|
||||
for (i = 0; i < nlocal; i++)
|
||||
if (mask[i] & groupbit) array[i][m] += fix_vector[i];
|
||||
} else {
|
||||
int jm1 = j - 1;
|
||||
double **fix_array = modify->fix[n]->array_atom;
|
||||
for (i = 0; i < nlocal; i++)
|
||||
if (mask[i] & groupbit) array[i][m] += fix_array[i][jm1];
|
||||
if (dimension == 2) {
|
||||
if (nvalues == 1) {
|
||||
for (i = 0; i < nlocal; i++) {
|
||||
if (mask[i] & groupbit)
|
||||
vec2d[bin[i][0]][bin[i][1]] += v[i][j];
|
||||
}
|
||||
} else
|
||||
for (i = 0; i < nlocal; i++) {
|
||||
if (mask[i] & groupbit)
|
||||
array2d[bin[i][0]][bin[i][1]][m] += v[i][j];
|
||||
}
|
||||
} else {
|
||||
if (nvalues == 1) {
|
||||
for (i = 0; i < nlocal; i++) {
|
||||
if (mask[i] & groupbit)
|
||||
vec3d[bin[i][0]][bin[i][1]][bin[i][2]] += v[i][j];
|
||||
}
|
||||
} else
|
||||
for (i = 0; i < nlocal; i++) {
|
||||
if (mask[i] & groupbit)
|
||||
array3d[bin[i][0]][bin[i][1]][bin[i][2]][m] += v[i][j];
|
||||
}
|
||||
}
|
||||
|
||||
} else if (which[m] == ArgInfo::F) {
|
||||
double **f = atom->f;
|
||||
|
||||
if (dimension == 2) {
|
||||
if (nvalues == 1) {
|
||||
for (i = 0; i < nlocal; i++) {
|
||||
if (mask[i] & groupbit)
|
||||
vec2d[bin[i][0]][bin[i][1]] += f[i][j];
|
||||
}
|
||||
} else
|
||||
for (i = 0; i < nlocal; i++) {
|
||||
if (mask[i] & groupbit)
|
||||
array2d[bin[i][0]][bin[i][1]][m] += f[i][j];
|
||||
}
|
||||
} else {
|
||||
if (nvalues == 1) {
|
||||
for (i = 0; i < nlocal; i++) {
|
||||
if (mask[i] & groupbit)
|
||||
vec3d[bin[i][0]][bin[i][1]][bin[i][2]] += f[i][j];
|
||||
}
|
||||
} else
|
||||
for (i = 0; i < nlocal; i++) {
|
||||
if (mask[i] & groupbit)
|
||||
array3d[bin[i][0]][bin[i][1]][bin[i][2]][m] += f[i][j];
|
||||
}
|
||||
}
|
||||
|
||||
// per-atom compute or fix
|
||||
// invoke compute if not previously invoked
|
||||
|
||||
} else if (which[m] == ArgInfo::COMPUTE || which[m] == ArgInfo::FIX) {
|
||||
double *ovector,**oarray;
|
||||
|
||||
if (which[m] == ArgInfo::COMPUTE) {
|
||||
Compute *compute = modify->compute[n];
|
||||
if (!(compute->invoked_flag & Compute::INVOKED_PERATOM)) {
|
||||
compute->compute_peratom();
|
||||
compute->invoked_flag |= Compute::INVOKED_PERATOM;
|
||||
}
|
||||
if (j == 0) ovector = compute->vector_atom;
|
||||
else oarray = compute->array_atom;
|
||||
} else {
|
||||
Fix *fix = modify->fix[n];
|
||||
if (j == 0) ovector = fix->vector_atom;
|
||||
else oarray = fix->array_atom;
|
||||
}
|
||||
|
||||
if (dimension == 2) {
|
||||
if (nvalues == 1) {
|
||||
if (j == 0) {
|
||||
for (i = 0; i < nlocal; i++) {
|
||||
if (mask[i] & groupbit)
|
||||
vec2d[bin[i][0]][bin[i][1]] += ovector[i];
|
||||
}
|
||||
} else {
|
||||
int jm1 = j = 1;
|
||||
for (i = 0; i < nlocal; i++) {
|
||||
if (mask[i] & groupbit)
|
||||
vec2d[bin[i][0]][bin[i][1]] += oarray[i][jm1];
|
||||
}
|
||||
}
|
||||
} else {
|
||||
if (j == 0) {
|
||||
for (i = 0; i < nlocal; i++) {
|
||||
if (mask[i] & groupbit)
|
||||
array2d[bin[i][0]][bin[i][1]][m] += ovector[i];
|
||||
}
|
||||
} else {
|
||||
int jm1 = j - 1;
|
||||
for (i = 0; i < nlocal; i++) {
|
||||
if (mask[i] & groupbit)
|
||||
array2d[bin[i][0]][bin[i][1]][m] += oarray[i][jm1];
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
} else {
|
||||
if (nvalues == 1) {
|
||||
if (j == 0) {
|
||||
for (i = 0; i < nlocal; i++) {
|
||||
if (mask[i] & groupbit)
|
||||
vec3d[bin[i][0]][bin[i][1]][bin[i][2]] += ovector[i];
|
||||
}
|
||||
} else {
|
||||
int jm1 = j - 1;
|
||||
for (i = 0; i < nlocal; i++) {
|
||||
if (mask[i] & groupbit)
|
||||
vec3d[bin[i][0]][bin[i][1]][bin[i][2]] += oarray[i][jm1];
|
||||
}
|
||||
}
|
||||
} else {
|
||||
if (j == 0) {
|
||||
for (i = 0; i < nlocal; i++) {
|
||||
if (mask[i] & groupbit)
|
||||
array3d[bin[i][0]][bin[i][1]][bin[i][2]][m] += ovector[i];
|
||||
}
|
||||
} else {
|
||||
int jm1 = j - 1;
|
||||
for (i = 0; i < nlocal; i++) {
|
||||
if (mask[i] & groupbit)
|
||||
array3d[bin[i][0]][bin[i][1]][bin[i][2]][m] += oarray[i][jm1];
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
// evaluate atom-style variable
|
||||
// final argument = 1 sums result to array
|
||||
|
||||
} else if (which[m] == ArgInfo::VARIABLE) {
|
||||
/*
|
||||
if (array) input->variable->compute_atom(n,igroup,&array[0][m],nvalues,1);
|
||||
else input->variable->compute_atom(n,igroup,nullptr,nvalues,1);
|
||||
*/
|
||||
}
|
||||
|
||||
// evaluate atom-style variable
|
||||
// final argument = 1 sums result to array
|
||||
|
||||
} else if (which[m] == ArgInfo::VARIABLE) {
|
||||
if (array) input->variable->compute_atom(n,igroup,&array[0][m],nvalues,1);
|
||||
else input->variable->compute_atom(n,igroup,nullptr,nvalues,1);
|
||||
}
|
||||
}
|
||||
}
|
||||
*/
|
||||
|
||||
// GRID mode
|
||||
// accumulate results of computes & fixes to local grid
|
||||
|
||||
if (modegrid) {
|
||||
|
||||
for (m = 0; m < nvalues; m++) {
|
||||
n = value2index[m];
|
||||
j = argindex[m];
|
||||
int idata = value2data[m];
|
||||
|
||||
Compute *compute;
|
||||
Fix *fix;
|
||||
|
||||
if (which[m] == ArgInfo::COMPUTE) {
|
||||
compute = modify->compute[n];
|
||||
if (!(compute->invoked_flag & Compute::INVOKED_PERGRID)) {
|
||||
compute->compute_pergrid();
|
||||
compute->invoked_flag |= Compute::INVOKED_PERGRID;
|
||||
}
|
||||
} else if (which[m] == ArgInfo::FIX) fix = modify->fix[n];
|
||||
|
||||
if (dimension == 2) {
|
||||
double **ovec2d,***oarray2d;
|
||||
if (which[m] == ArgInfo::COMPUTE) {
|
||||
if (j == 0)
|
||||
ovec2d = (double **) compute->get_griddata_by_index(idata);
|
||||
else
|
||||
oarray2d = (double ***) compute->get_griddata_by_index(idata);
|
||||
} else {
|
||||
if (j == 0)
|
||||
ovec2d = (double **) fix->get_griddata_by_index(idata);
|
||||
else
|
||||
oarray2d = (double ***) fix->get_griddata_by_index(idata);
|
||||
}
|
||||
|
||||
if (nvalues == 1) {
|
||||
if (j == 0) {
|
||||
for (iy = nylo_in; iy <= nyhi_in; iy++)
|
||||
for (ix = nxlo_in; ix <= nxhi_in; ix++)
|
||||
vec2d[iy][ix] += ovec2d[iy][ix];
|
||||
} else {
|
||||
int jm1 = j - 1;
|
||||
for (iy = nylo_in; iy <= nyhi_in; iy++)
|
||||
for (ix = nxlo_in; ix <= nxhi_in; ix++)
|
||||
vec2d[iy][ix] += oarray2d[iy][ix][jm1];
|
||||
}
|
||||
} else {
|
||||
if (j == 0) {
|
||||
for (iy = nylo_in; iy <= nyhi_in; iy++)
|
||||
for (ix = nxlo_in; ix <= nxhi_in; ix++)
|
||||
array2d[iy][ix][m] += ovec2d[iy][ix];
|
||||
} else {
|
||||
int jm1 = j - 1;
|
||||
for (iy = nylo_in; iy <= nyhi_in; iy++)
|
||||
for (ix = nxlo_in; ix <= nxhi_in; ix++)
|
||||
array2d[iy][ix][m] += oarray2d[iy][ix][jm1];
|
||||
}
|
||||
}
|
||||
|
||||
} else {
|
||||
double ***ovec3d,****oarray3d;
|
||||
if (which[m] == ArgInfo::COMPUTE) {
|
||||
if (j == 0)
|
||||
ovec3d = (double ***) compute->get_griddata_by_index(idata);
|
||||
else
|
||||
oarray3d = (double ****) compute->get_griddata_by_index(idata);
|
||||
} else {
|
||||
if (j == 0)
|
||||
ovec3d = (double ***) fix->get_griddata_by_index(idata);
|
||||
else
|
||||
oarray3d = (double ****) fix->get_griddata_by_index(idata);
|
||||
}
|
||||
|
||||
if (nvalues == 1) {
|
||||
if (j == 0) {
|
||||
for (iz = nzlo_in; iz <= nzhi_in; iz++)
|
||||
for (iy = nylo_in; iy <= nyhi_in; iy++)
|
||||
for (ix = nxlo_in; ix <= nxhi_in; ix++)
|
||||
vec3d[iz][iy][ix] += ovec3d[iz][iy][ix];
|
||||
} else {
|
||||
int jm1 = j - 1;
|
||||
for (iz = nzlo_in; iz <= nzhi_in; iz++)
|
||||
for (iy = nylo_in; iy <= nyhi_in; iy++)
|
||||
for (ix = nxlo_in; ix <= nxhi_in; ix++)
|
||||
vec3d[iz][iy][ix] += oarray3d[iz][iy][ix][jm1];
|
||||
}
|
||||
} else {
|
||||
if (j == 0) {
|
||||
for (iz = nzlo_in; iz <= nzhi_in; iz++)
|
||||
for (iy = nylo_in; iy <= nyhi_in; iy++)
|
||||
for (ix = nxlo_in; ix <= nxhi_in; ix++)
|
||||
array3d[iz][iy][ix][m] += ovec3d[iz][iy][ix];
|
||||
} else {
|
||||
int jm1 = j - 1;
|
||||
for (iz = nzlo_in; iz <= nzhi_in; iz++)
|
||||
for (iy = nylo_in; iy <= nyhi_in; iy++)
|
||||
for (ix = nxlo_in; ix <= nxhi_in; ix++)
|
||||
array3d[iz][iy][ix][m] += oarray3d[iz][iy][ix][jm1];
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
// done if irepeat < nrepeat
|
||||
@ -640,28 +891,74 @@ void FixAveGrid::end_of_step()
|
||||
irepeat++;
|
||||
if (irepeat < nrepeat) {
|
||||
nvalid += nevery;
|
||||
modify->addstep_compute(nvalid);
|
||||
if (modeatom) modify->addstep_compute(nvalid);
|
||||
return;
|
||||
}
|
||||
|
||||
irepeat = 0;
|
||||
nvalid = ntimestep+peratom_freq - ((bigint)nrepeat-1)*nevery;
|
||||
modify->addstep_compute(nvalid);
|
||||
nvalid = ntimestep+peratom_freq - ((bigint) nrepeat-1)*nevery;
|
||||
if (modeatom) modify->addstep_compute(nvalid);
|
||||
|
||||
// if (array == nullptr) return;
|
||||
|
||||
|
||||
//NOTE: need to do comm for atom mode ?
|
||||
// just return if this proc owns no grid points
|
||||
|
||||
if (ngridout == 0) return;
|
||||
|
||||
|
||||
// NOTE: need to do comm for atom mode ?
|
||||
|
||||
|
||||
|
||||
// average the final result for the Nfreq timestep
|
||||
// just loop over owned grid points
|
||||
|
||||
double invrepeat = 1.0/nrepeat;
|
||||
|
||||
/*
|
||||
double repeat = nrepeat;
|
||||
for (i = 0; i < nlocal; i++)
|
||||
for (m = 0; m < nvalues; m++)
|
||||
array[i][m] /= repeat;
|
||||
*/
|
||||
if (dimension == 2) {
|
||||
if (nvalues == 1) {
|
||||
for (iy = nylo_in; iy <= nyhi_in; iy++)
|
||||
for (ix = nxlo_in; ix <= nxhi_in; ix++)
|
||||
vec2d[iy][ix] *= invrepeat;
|
||||
} else {
|
||||
for (iy = nylo_in; iy <= nyhi_in; iy++)
|
||||
for (ix = nxlo_in; ix <= nxhi_in; ix++)
|
||||
for (m = 0; m <= nvalues; m++)
|
||||
array2d[iy][ix][m] *= invrepeat;
|
||||
}
|
||||
} else {
|
||||
if (nvalues == 1) {
|
||||
for (iz = nzlo_in; iz <= nzhi_in; iz++)
|
||||
for (iy = nylo_in; iy <= nyhi_in; iy++)
|
||||
for (ix = nxlo_in; ix <= nxhi_in; ix++)
|
||||
vec3d[iz][iy][ix] *= invrepeat;
|
||||
} else {
|
||||
for (iz = nzlo_in; iz <= nzhi_in; iz++)
|
||||
for (iy = nylo_in; iy <= nyhi_in; iy++)
|
||||
for (ix = nxlo_in; ix <= nxhi_in; ix++)
|
||||
for (m = 0; m <= nvalues; m++)
|
||||
array3d[iz][iy][ix][m] *= invrepeat;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
/* ----------------------------------------------------------------------
|
||||
zero grid values incluing ghost cells
|
||||
------------------------------------------------------------------------- */
|
||||
|
||||
void FixAveGrid::zero_grid()
|
||||
{
|
||||
if (dimension == 2) {
|
||||
if (nvalues == 1) {
|
||||
if (ngridout) memcpy(&vec2d[0][0],0,ngridout*sizeof(double));
|
||||
} else {
|
||||
if (ngridout) memcpy(&array2d[0][0][0],0,ngridout*nvalues*sizeof(double));
|
||||
}
|
||||
} else {
|
||||
if (nvalues == 1) {
|
||||
if (ngridout) memcpy(&vec3d[0][0][0],0,ngridout*sizeof(double));
|
||||
} else {
|
||||
if (ngridout) memcpy(&array3d[0][0][0][0],0,ngridout*nvalues*sizeof(double));
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
/* ----------------------------------------------------------------------
|
||||
|
||||
Reference in New Issue
Block a user