File: kdtree.cc

package info (click to toggle)
eckit 2.0.1-2
  • links: PTS, VCS
  • area: main
  • in suites: sid
  • size: 11,044 kB
  • sloc: cpp: 111,103; ansic: 2,826; yacc: 590; lex: 361; python: 302; sh: 162; makefile: 53
file content (302 lines) | stat: -rw-r--r-- 10,197 bytes parent folder | download | duplicates (3)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
/*
 * (C) Copyright 1996- ECMWF.
 *
 * This software is licensed under the terms of the Apache Licence Version 2.0
 * which can be obtained at http://www.apache.org/licenses/LICENSE-2.0.
 *
 * In applying this licence, ECMWF does not waive the privileges and immunities
 * granted to it by virtue of its status as an intergovernmental organisation nor
 * does it submit to any jurisdiction.
 */


#include <vector>

#include "eckit/container/KDTree.h"
#include "eckit/geo/PointXY.h"
#include "eckit/testing/Test.h"


namespace eckit::geo::test {


//----------------------------------------------------------------------------------------------------------------------

struct TestTreeTrait {
    using Point   = geo::PointXY;
    using Payload = double;
};

//----------------------------------------------------------------------------------------------------------------------

/// \brief Class used to test whether any point in a kd-tree lies in the interior of an
/// axis-aligned box.
template <typename TreeTrait>
struct PointInBoxInteriorFinder {
    using KDTree = KDTreeX<TreeTrait>;
    using Point  = typename KDTree::Point;
    using Alloc  = typename KDTree::Alloc;
    using Node   = typename KDTree::Node;

    /// \brief Returns true if any point in \p tree lies in the interior of the specified
    /// axis-aligned box.
    ///
    /// \param tree
    ///   Tree to search.
    /// \param lbound
    ///   Lower-left corner of the axis-aligned box.
    /// \param ubound
    ///   Upper-right corner of the axis-aligned box.
    static bool isAnyPointInBoxInterior(const KDTree& tree, const Point& lbound, const Point& ubound) {
        if (!tree.root_) {
            return false;
        }

        auto& alloc = tree.alloc_;
        auto* root  = alloc.convert(tree.root_, static_cast<Node*>(nullptr));
        ASSERT(root != nullptr);

        return isAnyPointInBoxInterior(root, alloc, lbound, ubound);
    }

private:

    /// \brief Returns true if the point stored in \p node or any of its descendants lies in the
    /// interior of the axis-aligned box with bottom-left and top-right corners at
    /// \p lbound and \p ubound.
    static bool isAnyPointInBoxInterior(const Node* node, Alloc& alloc, const Point& lbound, const Point& ubound) {
        if (node == nullptr) {
            return false;
        }

        const auto& point = node->value().point();

        if (isPointInBoxInterior(point, lbound, ubound)) {
            return true;
        }

        const size_t axis = node->axis();

        return (lbound.x(axis) < point.x(axis) && isAnyPointInBoxInterior(node->left(alloc), alloc, lbound, ubound)) ||
               (ubound.x(axis) > point.x(axis) && isAnyPointInBoxInterior(node->right(alloc), alloc, lbound, ubound));
    }

    /// \brief Returns true if \p point is in the interior of the axis-aligned box
    /// with bottom-left and top-right corners at \p lbound and \p ubound.
    static bool isPointInBoxInterior(const Point& point, const Point& lbound, const Point& ubound) {
        for (size_t d = 0; d < Point::DIMS; ++d) {
            if (point.x(d) <= lbound.x(d) || point.x(d) >= ubound.x(d)) {
                return false;
            }
        }
        return true;
    }
};

//----------------------------------------------------------------------------------------------------------------------

/// \brief Returns true if any point in \p tree is in the interior of the axis-aligned box
/// with bottom-left and top-right corners at \p lbound and \p ubound.
template <typename TreeTraits>
bool isAnyPointInBoxInterior(const KDTreeX<TreeTraits>& tree, const typename KDTreeX<TreeTraits>::Point& lbound,
                             const typename KDTreeX<TreeTraits>::Point& ubound) {
    return PointInBoxInteriorFinder<TreeTraits>::isAnyPointInBoxInterior(tree, lbound, ubound);
}

//----------------------------------------------------------------------------------------------------------------------

#define EXPECT_POINT_EQUAL(a, b)                       \
    for (size_t i = 0; i < Point::dimensions(); ++i) { \
        EXPECT(a.x(i) == b.x(i));                      \
    }


CASE("test_eckit_container_kdtree_constructor") {
    using Tree  = KDTreeMemory<TestTreeTrait>;
    using Point = Tree::PointType;

    // build k-d tree (offline)
    Tree kd;
    std::vector<Tree::Value> points;
    for (size_t i = 0; i < 10; ++i) {
        for (size_t j = 0; j < 10; ++j) {
            points.emplace_back(Point{static_cast<double>(i), static_cast<double>(j)}, 99.9);
        }
    }
    kd.build(points.begin(), points.end());

    // size
    EXPECT_EQUAL(kd.size(), points.size());

    // pick a point
    auto ref = points[points.size() / 2].point();

    SECTION("test single closest point") {
        // a point similar to an existing one
        EXPECT_POINT_EQUAL(ref, kd.nearestNeighbour(ref + Point{0.1, 0.1}).point());

        // exact match to a point
        EXPECT_POINT_EQUAL(ref, kd.nearestNeighbour(ref).point());

        // off the scale, i.e. not within a group of points (+)
        EXPECT_POINT_EQUAL(points.back().point(),
                           kd.nearestNeighbour(points.back().point() + Point{1000., 0.}).point());

        // off the scale, i.e. not within a group of points (-)
        EXPECT_POINT_EQUAL(points.front().point(),
                           kd.nearestNeighbour(points.front().point() + Point{-1000., 0.}).point());
    }

    SECTION("test N nearest") {
        // move this point so it lies between four equally, make sure we differ by 0.5 along each axis
        auto test = ref + Point{0.5, 0.5};

        for (auto& near : kd.kNearestNeighbours(test, 4)) {
            auto diff = near.point() - test;
            for (size_t i = 0; i < Point::dimensions(); ++i) {
                EXPECT(Point::distance(Point{0., 0.}, diff, i) == 0.5);
            }
        }
    }

    SECTION("test a custom visitor") {
        // Test a custom visitor. The purpose of doing that in this test is to ensure that the public
        // interface of KDTree is sufficient to write a custom class traversing the tree.
        auto a      = Point{0.25, 0.25};
        auto lbound = ref - a;
        auto ubound = ref + a;
        EXPECT(isAnyPointInBoxInterior(kd, lbound, ubound));

        auto b = Point{0.5, 0.5};
        lbound = lbound + b;
        ubound = ubound + b;
        EXPECT_NOT(isAnyPointInBoxInterior(kd, lbound, ubound));
    }
}

CASE("test_eckit_container_kdtree_insert") {
    using Tree  = KDTreeMemory<TestTreeTrait>;
    using Point = Tree::PointType;

    // build k-d tree (online)
    Tree kd;
    std::vector<Tree::Value> points;
    for (size_t i = 0; i < 10; ++i) {
        for (size_t j = 0; j < 10; ++j) {
            points.emplace_back(Point{static_cast<double>(i), static_cast<double>(j)}, 99.9);
            kd.insert(points.back());
        }
    }

    // size
    EXPECT_EQUAL(kd.size(), points.size());

    // pick a point
    auto ref = points[points.size() / 2].point();

    SECTION("test single closest point") {
        // a point similar to an existing one
        EXPECT_POINT_EQUAL(ref, kd.nearestNeighbour(ref + Point{0.1, 0.1}).point());

        // exact match to a point
        EXPECT_POINT_EQUAL(ref, kd.nearestNeighbour(ref).point());

        // off the scale, i.e. not within a group of points (+)
        EXPECT_POINT_EQUAL(points.back().point(),
                           kd.nearestNeighbour(points.back().point() + Point{1000., 0.}).point());

        // off the scale, i.e. not within a group of points (-)
        EXPECT_POINT_EQUAL(points.front().point(),
                           kd.nearestNeighbour(points.front().point() + Point{-1000., 0.}).point());
    }

    SECTION("test N nearest") {
        // move this point so it lies between four equally, make sure we differ by 0.5 along each axis
        auto test = ref + Point{0.5, 0.5};

        for (auto& near : kd.kNearestNeighbours(test, 4)) {
            auto diff = near.point() - test;
            for (size_t i = 0; i < Point::dimensions(); ++i) {
                EXPECT(Point::distance(Point{0., 0.}, diff, i) == 0.5);
            }
        }
    }
}

CASE("test_kdtree_mapped") {
    using Tree  = KDTreeMapped<TestTreeTrait>;
    using Point = Tree::PointType;

    std::vector<Tree::Value> points;
    for (size_t i = 0; i < 10; ++i) {
        for (size_t j = 0; j < 10; ++j) {
            points.emplace_back(Point{static_cast<double>(i), static_cast<double>(j)}, 99.9);
        }
    }

    // pick a point
    auto ref = points[points.size() / 2].point();

    auto passTest = [&](Tree& kd, const Point& p) -> bool {
        // perturb it a little
        // we should find the same point
        auto nr = kd.nearestNeighbour(p + Point{0.1, 0.1}).point();
        for (size_t i = 0; i < Point::dimensions(); ++i) {
            if (nr.x(i) != p.x(i)) {
                return false;
            }
        }
        return true;
    };

    PathName path("test_kdtree_mapped.kdtree");

    // Write file with k-d tree
    {
        if (path.exists()) {
            path.unlink();
        }

        Tree kd(path, points.size(), 0);
        EXPECT(kd.empty());

        kd.build(points);

        EXPECT_EQUAL(kd.size(), points.size());
        EXPECT(passTest(kd, ref));
    }

    // Load file with k-d tree
    {
        Tree kd(path, 0, 0);

        // Cannot insert point as the tree is readonly
        EXPECT_THROWS_AS(kd.insert(points.front()), AssertionFailed);

        // Cannot build with points as the tree is readonly
        EXPECT_THROWS_AS(kd.build(points), AssertionFailed);

        EXPECT_EQUAL(kd.size(), points.size());
        EXPECT(passTest(kd, ref));
    }
}

CASE("test_kdtree_iterate_empty") {
    using Tree = KDTreeMemory<TestTreeTrait>;

    size_t count = 0;
    Tree kd;
    for (auto& item : kd) {
        count++;
    }
    EXPECT_EQUAL(count, 0);
    EXPECT(kd.empty());
}

//----------------------------------------------------------------------------------------------------------------------

}  // namespace eckit::geo::test

int main(int argc, char** argv) {
    return eckit::testing::run_tests(argc, argv);
}