/*
        AUTOMATICALLY GENERATED BY jTemp FROM
        /Users/jsh2/Work/openimaj/target/checkout/core/core/src/main/jtemp/org/openimaj/util/tree/Incremental#T#KDTree.jtemp
*/
/**
 * Copyright (c) 2011, The University of Southampton and the individual contributors.
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without modification,
 * are permitted provided that the following conditions are met:
 *
 *   *  Redistributions of source code must retain the above copyright notice,
 *      this list of conditions and the following disclaimer.
 *
 *   *  Redistributions in binary form must reproduce the above copyright notice,
 *      this list of conditions and the following disclaimer in the documentation
 *      and/or other materials provided with the distribution.
 *
 *   *  Neither the name of the University of Southampton nor the names of its
 *      contributors may be used to endorse or promote products derived from this
 *      software without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
 * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
 * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR
 * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
 * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON
 * ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
 * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 */
package org.openimaj.util.tree;

import java.util.ArrayList;
import java.util.Collection;
import java.util.List;
import java.util.PriorityQueue;
import java.util.Stack;

import org.openimaj.util.pair.ObjectDoublePair;
import org.openimaj.util.queue.BoundedPriorityQueue;

/**
 * Implementation of a simple incremental KDTree for <code>int[]</code>s. Includes
 * support for range search, neighbour search, and radius search. The tree created 
 * by this class will usually be rather unbalanced.
 * <p>
 * The KDTree allows fast search for points in relatively low-dimension spaces.
 * 
 * @author Jonathon Hare (jsh2@ecs.soton.ac.uk)
 */
public class IncrementalIntKDTree {
    private static class KDNode {
        int discriminateDim;
        int[] point;
        KDNode left, right;

        KDNode(int[] point, int discriminate) {
                this.point = point;
                this.left = right = null;
                this.discriminateDim = discriminate;
        }
    }
    
        KDNode _root;

        /**
         * Create an empty KDTree object
         */
        public IncrementalIntKDTree() {
                _root = null;
        }

        /**
         * Create a KDTree object and populate it with the given data.
         * 
         * @param coords
         *            the data to populate the index with.
         */
        public IncrementalIntKDTree(Collection<int[]> coords) {
                _root = null;
                insertAll(coords);
        }
        
        /**
         * Create a KDTree object and populate it with the given data.
         * 
         * @param coords
         *            the data to populate the index with.
         */
        public IncrementalIntKDTree(int[][] coords) {
                _root = null;
                insertAll(coords);
        }

        /**
         * Insert all the points from the given collection into the index.
         * 
         * @param coords
         *            The points to add.
         */
        public void insertAll(Collection<int[]> coords) {
                for (final int[] c : coords)
                        insert(c);
        }
        
        /**
         * Insert all the points from the given collection into the index.
         * 
         * @param coords
         *            The points to add.
         */
        public void insertAll(int[][] coords) {
                for (final int[] c : coords)
                        insert(c);
        }

        /**
         * Inserts a point into the tree, preserving the spatial ordering.
         * 
         * @param point
         *            Point to insert.
         */
        public void insert(int[] point) {

                if (_root == null)
                        _root = new KDNode(point, 0);
                else {
                        int discriminate;
                        KDNode curNode, tmpNode;
                        double ordinate1, ordinate2;

                        curNode = _root;

                        do {
                                tmpNode = curNode;
                                discriminate = tmpNode.discriminateDim;

                                ordinate1 = point[discriminate];
                                ordinate2 = tmpNode.point[discriminate];

                                if (ordinate1 > ordinate2)
                                        curNode = tmpNode.right;
                                else
                                        curNode = tmpNode.left;
                        } while (curNode != null);

                        if (++discriminate >= point.length)
                                discriminate = 0;

                        if (ordinate1 > ordinate2)
                                tmpNode.right = new KDNode(point, discriminate);
                        else
                                tmpNode.left = new KDNode(point, discriminate);
                }
        }

        /**
         * Determines if a point is contained within a given k-dimensional bounding
         * box.
         */
        static final boolean isContained(
                        int[] point, int[] lower, int[] upper)
        {
                double ordinate1, ordinate2, ordinate3;

                for (int i = 0; i < point.length; i++) {
                        ordinate1 = point[i];
                        ordinate2 = lower[i];
                        ordinate3 = upper[i];

                        if (ordinate1 < ordinate2 || ordinate1 > ordinate3)
                                return false;
                }

                return true;
        }

        /**
         * Searches the tree for all points contained within the bounding box
         * defined by the given upper and lower extremes
         * 
         * @param lowerExtreme
         * @param upperExtreme
         * @return the points within the given bounds
         */
        public List<int[]> rangeSearch(int[] lowerExtreme, int[] upperExtreme) {
                final ArrayList<int[]> results = new ArrayList<int[]>(1000);
                final Stack<KDNode> stack = new Stack<KDNode>();
                KDNode tmpNode;
                int discriminate;
                double ordinate1, ordinate2;

                if (_root == null)
                        return results;

                stack.push(_root);

                while (!stack.empty()) {
                        tmpNode = stack.pop();
                        discriminate = tmpNode.discriminateDim;

                        ordinate1 = tmpNode.point[discriminate];
                        ordinate2 = lowerExtreme[discriminate];

                        if (ordinate1 >= ordinate2 && tmpNode.left != null)
                                stack.push(tmpNode.left);

                        ordinate2 = upperExtreme[discriminate];

                        if (ordinate1 <= ordinate2 && tmpNode.right != null)
                                stack.push(tmpNode.right);

                        if (isContained(tmpNode.point, lowerExtreme, upperExtreme))
                                results.add(tmpNode.point);
                }

                return results;
        }

        protected static final double distance(int[] a, int[] b) {
                double s = 0;

                for (int i = 0; i < a.length; i++) {
                        final double fa = a[i];
                        final double fb = b[i];
                        s += (fa - fb) * (fa - fb);
                }
                return s;
        }

        /**
         * Find the nearest neighbour. Only one neighbour will be returned - if
         * multiple neighbours share the same location, or are equidistant, then
         * this might not be the one you expect.
         * 
         * @param query
         *            query coordinate
         * @return nearest neighbour
         */
        public ObjectDoublePair<int[]> findNearestNeighbour(int[] query) {
                final Stack<KDNode> stack = walkdown(query);
                final ObjectDoublePair<int[]> state = new ObjectDoublePair<int[]>();
                state.first = stack.peek().point;
                state.second = distance(query, state.first);

                if (state.second == 0)
                        return state;

                while (!stack.isEmpty()) {
                        final KDNode current = stack.pop();

                        checkSubtree(current, query, state);
                }

                return state;
        }

        /**
         * Find the K nearest neighbours.
         * 
         * @param query
         *            query coordinate
         * @param k
         *            the number of neighbours to find
         * @return nearest neighbours
         */
        public List<ObjectDoublePair<int[]>> findNearestNeighbours(int[] query, int k) {
                final Stack<KDNode> stack = walkdown(query);
                final BoundedPriorityQueue<ObjectDoublePair<int[]>> state = new BoundedPriorityQueue<ObjectDoublePair<int[]>>(
                                k, ObjectDoublePair.SECOND_ITEM_ASCENDING_COMPARATOR);

                final ObjectDoublePair<int[]> initialState = new ObjectDoublePair<int[]>();
                initialState.first = stack.peek().point;
                initialState.second = distance(query, initialState.first);
                state.add(initialState);

                while (!stack.isEmpty()) {
                        final KDNode current = stack.pop();

                        checkSubtreeK(current, query, state, k);
                }

                return state.toOrderedListDestructive();
        }

        /*
         * Check a subtree for a closer match
         */
        private void checkSubtree(KDNode node, int[] query, ObjectDoublePair<int[]> state) {
                if (node == null)
                        return;

                final double dist = distance(query, node.point);
                if (dist < state.second) {
                        state.first = node.point;
                        state.second = dist;
                }

                if (state.second == 0)
                        return;

                final double d = node.point[node.discriminateDim] - query[node.discriminateDim];
                if (d * d > state.second) {
                        // check subtree
                        final double ordinate1 = query[node.discriminateDim];
                        final double ordinate2 = node.point[node.discriminateDim];

                        if (ordinate1 > ordinate2)
                                checkSubtree(node.right, query, state);
                        else
                                checkSubtree(node.left, query, state);
                } else {
                        checkSubtree(node.left, query, state);
                        checkSubtree(node.right, query, state);
                }
        }

        private void checkSubtreeK(KDNode node, int[] query, PriorityQueue<ObjectDoublePair<int[]>> state, int k) {
                if (node == null)
                        return;

                final double dist = distance(query, node.point);

                boolean cont = false;
                for (final ObjectDoublePair<int[]> s : state)
                        if (s.first.equals(node.point)) {
                                cont = true;
                                break;
                        }

                if (!cont) {
                        if (state.size() < k) {
                                // collect this node
                                final ObjectDoublePair<int[]> s = new ObjectDoublePair<int[]>();
                                s.first = node.point;
                                s.second = dist;
                                state.add(s);
                        } else if (dist < state.peek().second) {
                                // replace last node
                                final ObjectDoublePair<int[]> s = state.poll();
                                s.first = node.point;
                                s.second = dist;
                                state.add(s);
                        }
                }

                final double d = node.point[node.discriminateDim] - query[node.discriminateDim];
                if (d * d > state.peek().second) {
                        // check subtree
                        final double ordinate1 = query[node.discriminateDim];
                        final double ordinate2 = node.point[node.discriminateDim];

                        if (ordinate1 > ordinate2)
                                checkSubtreeK(node.right, query, state, k);
                        else
                                checkSubtreeK(node.left, query, state, k);
                } else {
                        checkSubtreeK(node.left, query, state, k);
                        checkSubtreeK(node.right, query, state, k);
                }
        }

        /*
         * walk down the tree until we hit a leaf, and return the path taken
         */
        private Stack<KDNode> walkdown(int[] point) {
                if (_root == null)
                        return null;
                else {
                        final Stack<KDNode> stack = new Stack<KDNode>();
                        int discriminate;
                        KDNode curNode, tmpNode;
                        double ordinate1, ordinate2;

                        curNode = _root;

                        do {
                                tmpNode = curNode;
                                stack.push(tmpNode);
                                if (tmpNode.point == point)
                                        return stack;
                                discriminate = tmpNode.discriminateDim;

                                ordinate1 = point[discriminate];
                                ordinate2 = tmpNode.point[discriminate];

                                if (ordinate1 > ordinate2)
                                        curNode = tmpNode.right;
                                else
                                        curNode = tmpNode.left;
                        } while (curNode != null);

                        if (++discriminate >= point.length)
                                discriminate = 0;

                        return stack;
                }
        }
        
        /**
         * Find all the points within the given radius of the given point
         * 
         * @param centre
         *            the centre point
         * @param radius
         *            the radius
         * @return the points
         */
        public List<int[]> radiusSearch(int[] centre, int radius) {
                final int[] lower = centre.clone();
                final int[] upper = centre.clone();

                for (int i = 0; i < centre.length; i++) {
                        lower[i] -= radius;
                        upper[i] += radius;
                }

                final List<int[]> rangeList = rangeSearch(lower, upper);
                final List<int[]> radiusList = new ArrayList<int[]>(rangeList.size());
                final double radSq = radius * radius;
                for (final int[] r : rangeList) {
                        if (distance(centre, r) < radSq)
                                radiusList.add(r);
                }

                return radiusList;
        }
        
        /**
         * Find all the points within the given radius of the given point. 
         * Returns the distance to the point as well as the point itself. Distance
         * is the squared L2 distance.
         * 
         * @param centre
         *            the centre point
         * @param radius
         *            the radius
         * @return the points and distances
         */
        public List<ObjectDoublePair<int[]>> radiusDistanceSearch(int[] centre, int radius) {
                final int[] lower = centre.clone();
                final int[] upper = centre.clone();

                for (int i = 0; i < centre.length; i++) {
                        lower[i] -= radius;
                        upper[i] += radius;
                }

                final List<int[]> rangeList = rangeSearch(lower, upper);
                final List<ObjectDoublePair<int[]>> radiusList = new ArrayList<ObjectDoublePair<int[]>>(rangeList.size());
                final double radSq = radius * radius;
                for (final int[] r : rangeList) {
                        double dist = distance(centre, r);
                        if (dist < radSq)
                                radiusList.add(new ObjectDoublePair<int[]>(r, dist));
                }

                return radiusList;
        }
}