Skip to content

GitLab

Commit c92b7906 authored by Uwe Schulzweida's avatar Uwe Schulzweida
Browse files

TEST_BBOX

parent f81e9422
Hide whitespace changes
Inline Side-by-side
src/grid_search.cc
View file @ c92b7906
......@@ -137,7 +137,7 @@ struct gridsearch *gridsearch_create_reg2d(bool is_cyclic, size_t dims[2], const
}
static
kdTree_t *gs_create_kdtree(size_t n, const double *restrict lons, const double *restrict lats)
kdTree_t *gs_create_kdtree(size_t n, const double *restrict lons, const double *restrict lats, struct gridsearch *gs)
{
struct kd_point *pointlist = (struct kd_point *) Malloc(n*sizeof(struct kd_point));
// see example_cartesian.c
......@@ -145,13 +145,12 @@ kdTree_t *gs_create_kdtree(size_t n, const double *restrict lons, const double *
kdata_t min[3], max[3];
min[0] = min[1] = min[2] = 1e9;
max[0] = max[1] = max[2] = -1e9;
kdata_t *restrict point;
#if defined(HAVE_OPENMP4)
#pragma omp simd
#pragma omp parallel for reduction(min: min) reduction(max: max)
#endif
for ( size_t i = 0; i < n; i++ )
{
point = pointlist[i].point;
kdata_t *restrict point = pointlist[i].point;
LLtoXYZ_kd(lons[i], lats[i], point);
for ( unsigned j = 0; j < 3; ++j )
{
......@@ -161,6 +160,16 @@ kdTree_t *gs_create_kdtree(size_t n, const double *restrict lons, const double *
pointlist[i].index = i;
}
#ifdef TEST_BBOX
for ( unsigned j = 0; j < 3; ++j )
{
gs->min[j] = min[j];
gs->max[j] = max[j];
}
#endif
// printf("min %g %g %g max %g %g %g\n", min[0], min[1], min[2], max[0], max[1], max[2]);
kdTree_t *kdt = kd_buildTree(pointlist, n, min, max, 3, ompNumThreads);
if ( pointlist ) Free(pointlist);
if ( kdt == NULL ) cdoAbort("kd_buildTree failed!");
......@@ -178,13 +187,13 @@ nfTree_t *gs_create_nanoflann(size_t n, const double *restrict lons, const doubl
min[0] = min[1] = min[2] = 1e9;
max[0] = max[1] = max[2] = -1e9;
// Generating Point Cloud
float point[3];
pointcloud->pts.resize(n);
#if defined(HAVE_OPENMP4)
#pragma omp simd
#pragma omp parallel for reduction(min: min) reduction(max: max)
#endif
for ( size_t i = 0; i < n; i++ )
{
float point[3];
LLtoXYZ_f(lons[i], lats[i], point);
pointcloud->pts[i].x = point[0];
pointcloud->pts[i].y = point[1];
......@@ -214,7 +223,7 @@ kdTree_t *gs_create_kdsph(size_t n, const double *restrict lons, const double *r
max[0] = max[1] = -1e9;
kdata_t *restrict point;
#if defined(HAVE_OPENMP4)
#pragma omp simd
//#pragma omp simd
#endif
for ( size_t i = 0; i < n; i++ )
{
......@@ -480,7 +489,7 @@ struct gridsearch *gridsearch_create(size_t n, const double *restrict lons, cons
gs->n = n;
if ( n == 0 ) return gs;
gs->kdt = gs_create_kdtree(n, lons, lats);
gs->kdt = gs_create_kdtree(n, lons, lats, gs);
gs->search_radius = cdo_default_search_radius();
......@@ -496,7 +505,7 @@ struct gridsearch *gridsearch_create_nn(size_t n, const double *restrict lons, c
gs->n = n;
if ( n == 0 ) return gs;
if ( gs->method_nn == GS_KDTREE ) gs->kdt = gs_create_kdtree(n, lons, lats);
if ( gs->method_nn == GS_KDTREE ) gs->kdt = gs_create_kdtree(n, lons, lats, gs);
else if ( gs->method_nn == GS_NANOFLANN ) gs->nft = gs_create_nanoflann(n, lons, lats, gs);
else if ( gs->method_nn == GS_KDSPH ) gs->kdt = gs_create_kdsph(n, lons, lats);
else if ( gs->method_nn == GS_NEARPT3 ) gs->near = gs_create_nearpt3(n, lons, lats);
......@@ -549,7 +558,7 @@ double gs_set_range(double *prange)
}
static
size_t gs_nearest_kdtree(kdTree_t *kdt, double lon, double lat, double *prange)
size_t gs_nearest_kdtree(kdTree_t *kdt, double lon, double lat, double *prange, struct gridsearch *gs)
{
size_t index = GS_NOT_FOUND;
if ( kdt == NULL ) return index;
......@@ -557,10 +566,15 @@ size_t gs_nearest_kdtree(kdTree_t *kdt, double lon, double lat, double *prange)
float range0 = gs_set_range(prange);
kdata_t range = KDATA_SCALE(range0);
kdata_t point[3];
LLtoXYZ_kd(lon, lat, point);
kdata_t query_pt[3];
LLtoXYZ_kd(lon, lat, query_pt);
#ifdef TEST_BBOX
for ( unsigned j = 0; j < 3; ++j )
if ( query_pt[j] < gs->min[j] || query_pt[j] > gs->max[j] ) return index;
#endif
kdNode *node = kd_nearest(kdt->node, point, &range, 3);
kdNode *node = kd_nearest(kdt->node, query_pt, &range, 3);
float frange = KDATA_INVSCALE(range);
if ( !(frange < range0) ) node = NULL;
......@@ -612,11 +626,11 @@ size_t gs_nearest_kdsph(kdTree_t *kdt, double lon, double lat, double *prange)
float range0 = gs_set_range(prange);
kdata_t range = KDATA_SCALE(range0);
kdata_t point[2];
point[0] = lon;
point[1] = lat;
kdata_t query_pt[2];
query_pt[0] = lon;
query_pt[1] = lat;
kdNode *node = kd_nearest(kdt->node, point, &range, 3);
kdNode *node = kd_nearest(kdt->node, query_pt, &range, 3);
float frange = KDATA_INVSCALE(range);
if ( !(frange < range0) ) node = NULL;
......@@ -636,22 +650,22 @@ size_t gs_nearest_nearpt3(struct gsNear *near, double lon, double lat, double *p
#if defined(ENABLE_NEARPT3)
float range0 = gs_set_range(prange);
float point[3];
LLtoXYZ_f(lon, lat, point);
float query_pt[3];
LLtoXYZ_f(lon, lat, query_pt);
Coord_T q[3];
q[0] = NPT3SCALE(point[0]);
q[1] = NPT3SCALE(point[1]);
q[2] = NPT3SCALE(point[2]);
q[0] = NPT3SCALE(query_pt[0]);
q[1] = NPT3SCALE(query_pt[1]);
q[2] = NPT3SCALE(query_pt[2]);
int closestpt = nearpt3_query(near->nearpt3, q);
if ( closestpt >= 0 )
{
float point0[3];
LLtoXYZ_f(near->plons[closestpt], near->plats[closestpt], point0);
float query_pt0[3];
LLtoXYZ_f(near->plons[closestpt], near->plats[closestpt], query_pt0);
float range = distance(point, point0);
float range = distance(query_pt, query_pt0);
if ( range < range0 )
{
index = (size_t) closestpt;
......@@ -675,8 +689,8 @@ size_t gs_nearest_full(struct gsFull *full, double lon, double lat, double *pra
float range0 = gs_set_range(prange);
float point[3];
LLtoXYZ_f(lon, lat, point);
float query_pt[3];
LLtoXYZ_f(lon, lat, query_pt);
size_t n = full->n;
float **pts = full->pts;
......@@ -684,7 +698,7 @@ size_t gs_nearest_full(struct gsFull *full, double lon, double lat, double *pra
float dist = FLT_MAX;
for ( size_t i = 0; i < n; i++ )
{
float d = distance(point, pts[i]);
float d = distance(query_pt, pts[i]);
if ( closestpt >=n || d < dist || (d<=dist && i < closestpt) )
{
dist = d;
......@@ -712,7 +726,7 @@ size_t gridsearch_nearest(struct gridsearch *gs, double lon, double lat, double
if ( gs )
{
// clang-format off
if ( gs->method_nn == GS_KDTREE ) index = gs_nearest_kdtree(gs->kdt, lon, lat, prange);
if ( gs->method_nn == GS_KDTREE ) index = gs_nearest_kdtree(gs->kdt, lon, lat, prange, gs);
else if ( gs->method_nn == GS_NANOFLANN ) index = gs_nearest_nanoflann(gs->nft, lon, lat, prange);
else if ( gs->method_nn == GS_KDSPH ) index = gs_nearest_kdsph(gs->kdt, lon, lat, prange);
else if ( gs->method_nn == GS_NEARPT3 ) index = gs_nearest_nearpt3(gs->near, lon, lat, prange);
......@@ -729,16 +743,16 @@ struct pqueue *gridsearch_qnearest(struct gridsearch *gs, double lon, double lat
{
if ( gs->kdt == NULL ) return NULL;
kdata_t point[3];
kdata_t query_pt[3];
float range0 = gs_set_range(prange);
kdata_t range = KDATA_SCALE(range0);
struct pqueue *result = NULL;
LLtoXYZ_kd(lon, lat, point);
LLtoXYZ_kd(lon, lat, query_pt);
if ( gs )
{
result = kd_qnearest(gs->kdt->node, point, &range, nnn, 3);
result = kd_qnearest(gs->kdt->node, query_pt, &range, nnn, 3);
// printf("range %g %g %g %p\n", lon, lat, range, node);
float frange = KDATA_INVSCALE(range);
......
src/grid_search.h
View file @ c92b7906
......@@ -6,18 +6,15 @@
#include "nearpt3c.h"
#include "nanoflann.hpp"
#define GS_NOT_FOUND SIZE_MAX
//#define TEST_BBOX 1
#define GS_NOT_FOUND SIZE_MAX
enum T_GRIDSEARCH_METHOD_NN {GS_FULL=1, GS_NANOFLANN, GS_KDTREE, GS_KDSPH, GS_NEARPT3};
template <typename T>
struct PointCloud
{
struct Point
{
T x,y,z;
};
struct Point { T x,y,z; };
std::vector<Point> pts;
// Must return the number of data points
......@@ -82,6 +79,9 @@ struct gridsearch {
double *coslon, *sinlon; // cosine, sine of grid lons (for distance)
double lonmin, lonmax, latmin, latmax;
#ifdef TEST_BBOX
float min[3], max[3];
#endif
};
struct gsknn {
......
src/nanoflann.hpp
View file @ c92b7906
......@@ -46,7 +46,7 @@
#ifndef NANOFLANN_HPP_
#define NANOFLANN_HPP_
#define NANOFLANN_FIRST_MATCH 1
#define NANOFLANN_FIRST_MATCH 1
#include <vector>
#include <cassert>
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment

Imprint | Privacy Policy