From 6ff8072e8c5a6dc1301e884f5a648a0b63bdd48a Mon Sep 17 00:00:00 2001 From: Arnur Nigmetov Date: Tue, 3 Dec 2019 20:34:28 +0100 Subject: Rename directories for bottleneck and Wasserstein --- bottleneck/include/dnn/local/kd-tree.hpp | 296 +++++++++++++++++++++++++++++++ 1 file changed, 296 insertions(+) create mode 100644 bottleneck/include/dnn/local/kd-tree.hpp (limited to 'bottleneck/include/dnn/local/kd-tree.hpp') diff --git a/bottleneck/include/dnn/local/kd-tree.hpp b/bottleneck/include/dnn/local/kd-tree.hpp new file mode 100644 index 0000000..5b84eb0 --- /dev/null +++ b/bottleneck/include/dnn/local/kd-tree.hpp @@ -0,0 +1,296 @@ +#include +#include +#include + +#include + +#include "../parallel/tbb.h" + +template +hera::bt::dnn::KDTree::KDTree(const Traits& traits, HandleContainer&& handles): + traits_(traits), + tree_(std::move(handles)), + delete_flags_(handles.size(), static_cast(0) ), + subtree_n_elems(handles.size(), static_cast(0)), + num_points_(handles.size()) +{ + init(); +} + +template +template +hera::bt::dnn::KDTree::KDTree(const Traits& traits, const Range& range): + traits_(traits) +{ + init(range); +} + +template +template +void hera::bt::dnn::KDTree::init(const Range& range) +{ + size_t sz = std::distance(std::begin(range), std::end(range)); + subtree_n_elems = std::vector(sz, 0); + delete_flags_ = std::vector(sz, 0); + num_points_ = sz; + tree_.reserve(sz); + for (PointHandle h : range) + tree_.push_back(h); + parents_.resize(sz, -1); + init(); +} + +template +void hera::bt::dnn::KDTree::init() +{ + if (tree_.empty()) + return; + +#if defined(TBB) + task_group g; + g.run(OrderTree(this, tree_.begin(), tree_.end(), -1, 0, traits())); + g.wait(); +#else + OrderTree(this, tree_.begin(), tree_.end(), -1, 0, traits()).serial(); +#endif + + for (size_t i = 0; i < tree_.size(); ++i) + indices_[tree_[i]] = i; + init_n_elems(); +} + +template +struct +hera::bt::dnn::KDTree::OrderTree +{ + OrderTree(KDTree* tree_, HCIterator b_, HCIterator e_, ssize_t p_, size_t i_, const Traits& traits_): + tree(tree_), b(b_), e(e_), p(p_), i(i_), traits(traits_) {} + + void operator()() const + { + if (e - b < 1000) + { + serial(); + return; + } + + HCIterator m = b + (e - b)/2; + ssize_t im = m - tree->tree_.begin(); + tree->parents_[im] = p; + + CoordinateComparison cmp(i, traits); + std::nth_element(b,m,e, cmp); + size_t next_i = (i + 1) % traits.dimension(); + + task_group g; + if (b < m - 1) g.run(OrderTree(tree, b, m, im, next_i, traits)); + if (e > m + 2) g.run(OrderTree(tree, m+1, e, im, next_i, traits)); + g.wait(); + } + + void serial() const + { + std::queue q; + q.push(KDTreeNode(b,e,p,i)); + while (!q.empty()) + { + HCIterator b, e; ssize_t p; size_t i; + std::tie(b,e,p,i) = q.front(); + q.pop(); + HCIterator m = b + (e - b)/2; + ssize_t im = m - tree->tree_.begin(); + tree->parents_[im] = p; + + CoordinateComparison cmp(i, traits); + std::nth_element(b,m,e, cmp); + size_t next_i = (i + 1) % traits.dimension(); + + // Replace with a size condition instead? + if (m - b > 1) + q.push(KDTreeNode(b, m, im, next_i)); + else if (b < m) + tree->parents_[im - 1] = im; + if (e - m > 2) + q.push(KDTreeNode(m+1, e, im, next_i)); + else if (e > m + 1) + tree->parents_[im + 1] = im; + } + } + + KDTree* tree; + HCIterator b, e; + ssize_t p; + size_t i; + const Traits& traits; +}; + +template +void hera::bt::dnn::KDTree::update_n_elems(ssize_t idx, const int delta) +// add delta to the number of points in node idx and update subtree_n_elems +// for all parents of the node idx +{ + //std::cout << "subtree_n_elems.size = " << subtree_n_elems.size() << std::endl; + // update the node itself + while (idx != -1) + { + //std::cout << idx << std::endl; + subtree_n_elems[idx] += delta; + idx = parents_[idx]; + } +} + +template +void hera::bt::dnn::KDTree::increase_n_elems(const ssize_t idx) +{ + update_n_elems(idx, static_cast(1)); +} + +template +void hera::bt::dnn::KDTree::decrease_n_elems(const ssize_t idx) +{ + update_n_elems(idx, static_cast(-1)); +} + +template +void hera::bt::dnn::KDTree::init_n_elems() +{ + for(size_t idx = 0; idx < tree_.size(); ++idx) { + increase_n_elems(idx); + } +} + + +template +template +void hera::bt::dnn::KDTree::search(PointHandle q, ResultsFunctor& rf) const +{ + typedef typename HandleContainer::const_iterator HCIterator; + typedef std::tuple KDTreeNode; + + if (tree_.empty()) + return; + + DistanceType D = std::numeric_limits::infinity(); + + // TODO: use tbb::scalable_allocator for the queue + std::queue nodes; + + nodes.push(KDTreeNode(tree_.begin(), tree_.end(), 0)); + + //std::cout << "started kdtree::search" << std::endl; + + while (!nodes.empty()) + { + HCIterator b, e; size_t i; + std::tie(b,e,i) = nodes.front(); + nodes.pop(); + + CoordinateComparison cmp(i, traits()); + i = (i + 1) % traits().dimension(); + + HCIterator m = b + (e - b)/2; + size_t m_idx = m - tree_.begin(); + // ignore deleted points + if ( delete_flags_[m_idx] == 0 ) { + DistanceType dist = traits().distance(q, *m); + // + weights_[m - tree_.begin()]; + //std::cout << "Supplied to functor: m : "; + //std::cout << "(" << (*(*m))[0] << ", " << (*(*m))[1] << ")"; + //std::cout << " and q : "; + //std::cout << "(" << (*q)[0] << ", " << (*q)[1] << ")" << std::endl; + //std::cout << "dist^q + weight = " << dist << std::endl; + //std::cout << "weight = " << weights_[m - tree_.begin()] << std::endl; + //std::cout << "dist = " << traits().distance(q, *m) << std::endl; + //std::cout << "dist^q = " << pow(traits().distance(q, *m), wassersteinPower) << std::endl; + + D = rf(*m, dist); + } + // we are really searching w.r.t L_\infty ball; could prune better with an L_2 ball + Coordinate diff = cmp.diff(q, *m); // diff returns signed distance + DistanceType diffToWasserPower = (diff > 0 ? 1.0 : -1.0) * fabs(diff); + + size_t lm = m + 1 + (e - (m+1))/2 - tree_.begin(); + if ( e > m + 1 and subtree_n_elems[lm] > 0 ) { + if (e > m + 1 && diffToWasserPower >= -D) { + nodes.push(KDTreeNode(m+1, e, i)); + } + } + + size_t rm = b + (m - b) / 2 - tree_.begin(); + if ( subtree_n_elems[rm] > 0 ) { + if (b < m && diffToWasserPower <= D) { + nodes.push(KDTreeNode(b, m, i)); + } + } + } + //std::cout << "exited kdtree::search" << std::endl; +} + +template +typename hera::bt::dnn::KDTree::HandleDistance hera::bt::dnn::KDTree::find(PointHandle q) const +{ + hera::bt::dnn::NNRecord nn; + search(q, nn); + return nn.result; +} + +template +typename hera::bt::dnn::KDTree::Result hera::bt::dnn::KDTree::findR(PointHandle q, DistanceType r) const +{ + hera::bt::dnn::rNNRecord rnn(r); + search(q, rnn); + //std::sort(rnn.result.begin(), rnn.result.end()); + return rnn.result; +} + +template +typename hera::bt::dnn::KDTree::Result hera::bt::dnn::KDTree::findFirstR(PointHandle q, DistanceType r) const +{ + hera::bt::dnn::firstrNNRecord rnn(r); + search(q, rnn); + return rnn.result; +} + +template +typename hera::bt::dnn::KDTree::Result hera::bt::dnn::KDTree::findK(PointHandle q, size_t k) const +{ + hera::bt::dnn::kNNRecord knn(k); + search(q, knn); + // do we need this??? + std::sort(knn.result.begin(), knn.result.end()); + return knn.result; +} + +template +struct hera::bt::dnn::KDTree::CoordinateComparison +{ + CoordinateComparison(size_t i, const Traits& traits): + i_(i), traits_(traits) {} + + bool operator()(PointHandle p1, PointHandle p2) const { return coordinate(p1) < coordinate(p2); } + Coordinate diff(PointHandle p1, PointHandle p2) const { return coordinate(p1) - coordinate(p2); } + + Coordinate coordinate(PointHandle p) const { return traits_.coordinate(p, i_); } + size_t axis() const { return i_; } + + private: + size_t i_; + const Traits& traits_; +}; + +template +void hera::bt::dnn::KDTree::delete_point(const size_t idx) +{ + // prevent double deletion + assert(delete_flags_[idx] == 0); + delete_flags_[idx] = 1; + decrease_n_elems(idx); + --num_points_; +} + +template +void hera::bt::dnn::KDTree::delete_point(PointHandle p) +{ + delete_point(indices_[p]); +} + -- cgit v1.2.3