Ejemplo n.º 1
0
        /// <summary>
        /// Used by delete(). Ensures that all nodes from the passed node up to the root have the minimum number of entries.
        /// Note that the parent and parentEntry stacks are expected to contain the nodeIds of all parents up to the root.
        /// </summary>
        /// <param name="rTree"></param>
        internal void condenseTree(RTree rTree)
        {
            // CT1 [Initialize] Set n=l. Set the list of eliminated nodes to be empty.
            NodeBase         n               = this;
            NodeInternal     parent          = null;
            int              parentEntry     = 0;
            Stack <NodeBase> eliminatedNodes = new Stack <NodeBase>();

            // CT2 [Find parent entry] If N is the root, go to CT6. Otherwise
            // let P be the parent of N, and let En be N's entry in P
            while (n.level != rTree.treeHeight)
            {
                parent      = rTree.parents.Pop() as NodeInternal;
                parentEntry = rTree.parentsEntry.Pop();

                // CT3 [Eliminiate under-full node] If N has too few entries,
                // delete En from P and add N to the list of eliminated nodes
                if (n.entryCount < rTree.minNodeEntries)
                {
                    parent.deleteEntry(parentEntry);
                    eliminatedNodes.Push(n);
                }
                else
                {
                    // CT4 [Adjust covering rectangle] If N has not been eliminated,
                    // adjust EnI to tightly contain all entries in N
                    if (n.minimumBoundingRectangle.MinX != parent.entries[parentEntry].Value.MinX || n.minimumBoundingRectangle.MinY != parent.entries[parentEntry].Value.MinY ||
                        n.minimumBoundingRectangle.MaxX != parent.entries[parentEntry].Value.MaxX || n.minimumBoundingRectangle.MaxY != parent.entries[parentEntry].Value.MaxY)
                    {
                        Rectangle d = parent.entries[parentEntry].Value;
                        parent.entries[parentEntry] = n.minimumBoundingRectangle;
                        parent.recalculateMBRIfInfluencedBy(ref d);
                    }
                }
                // CT5 [Move up one level in tree] Set N=P and repeat from CT2
                n = parent;
            }

            // CT6 [Reinsert orphaned entries] Reinsert all entries of nodes in set Q. Entries from eliminated leaf nodes are reinserted in tree leaves as in
            // Insert(), but entries from higher level nodes must be placed higher in the tree, so that leaves of their dependent subtrees will be on the same
            // level as leaves of the main tree
            while (eliminatedNodes.Count > 0)
            {
                NodeBase e = eliminatedNodes.Pop();
                for (int j = 0; j < e.entryCount; j++)
                {
                    if (e.level == 1)
                    {
                        rTree.AddInternal(e.entries[j].Value);
                    }
                    else
                    {
                        NodeInternal nInternal = e as NodeInternal;
                        rTree.AddInternal(e.entries[j].Value, e.level, nInternal.childNodes[j]);
                    }
                }
            }
        }
Ejemplo n.º 2
0
        internal NodeInternal adjustTree(RTree rTree, NodeInternal nn)
        {
            // AT1 [Initialize] Set N=L. If L was split previously, set NN to be the resulting second node.

            // AT2 [Check if done] If N is the root, stop
            NodeInternal n = this;

            while (n.level != rTree.treeHeight)
            {
                // AT3 [Adjust covering rectangle in parent entry] Let P be the parent node of N, and let En be N's entry in P. Adjust EnI so that it tightly encloses all entry rectangles in N.
                NodeInternal parent = rTree.parents.Pop() as NodeInternal;
                int          entry  = rTree.parentsEntry.Pop();

                if (parent.childNodes[entry] != n)
                {
                    throw new UnexpectedException("Error: entry " + entry + " in node " + parent + " should point to node " + n + "; actually points to node " + parent.childNodes[entry]);
                }

                Rectangle r = (Rectangle)parent.entries[entry];
                if (r.MinX != n.minimumBoundingRectangle.MinX || r.MinY != n.minimumBoundingRectangle.MinY ||
                    r.MaxX != n.minimumBoundingRectangle.MaxX || r.MaxY != n.minimumBoundingRectangle.MaxY)
                {
                    r = n.minimumBoundingRectangle;
                    Update();
                    parent.entries[entry] = r;
                    parent.recalculateMBR();
                }

                // AT4 [Propagate node split upward] If N has a partner NN resulting from an earlier split, create a new entry Enn with Ennp pointing to NN and
                // Enni enclosing all rectangles in NN. Add Enn to P if there is room. Otherwise, invoke splitNode to produce P and PP containing Enn and all P's old entries.
                NodeInternal newNode = null;
                if (nn != null)
                {
                    if (parent.entryCount < rTree.maxNodeEntries)
                    {
                        parent.addEntry(ref nn.minimumBoundingRectangle, nn);
                    }
                    else
                    {
                        newNode = parent.splitNode(rTree, ref nn.minimumBoundingRectangle, nn);
                    }
                }

                // AT5 [Move up to next level] Set N = P and set NN = PP if a split occurred. Repeat from AT2
                n  = parent;
                nn = newNode;

                parent  = null;
                newNode = null;
            }

            return(nn);
        }
Ejemplo n.º 3
0
 internal override double Nearest(Point p, RTree rTree, double furthestDistanceSq, PriorityQueueRTree nearestRectangles)
 {
     for (int i = 0; i < entryCount; i++)
     {
         double tempDistanceSq = entries[i].Value.distanceSq(p.x, p.y);
         // a rectangle nearer than actualNearest
         if (tempDistanceSq <= furthestDistanceSq)
         {
             NodeBase node = childNodes[i]; // search the child node
             furthestDistanceSq = node.Nearest(p, rTree, furthestDistanceSq, nearestRectangles);
         }
     }
     return(furthestDistanceSq);
 }
Ejemplo n.º 4
0
 internal override double Nearest(Point p, RTree rTree, double furthestDistanceSq, PriorityQueueRTree nearestRectangles)
 {
     for (int i = 0; i < entryCount; i++)
     {
         double tempDistanceSq = entries[i].Value.distanceSq(p.x, p.y);
         if (tempDistanceSq < furthestDistanceSq)
         {
             furthestDistanceSq = tempDistanceSq;
             nearestRectangles.Clear();
         }
         if (tempDistanceSq <= furthestDistanceSq)
         {
             nearestRectangles.Insert(entries[i].Value, tempDistanceSq);
         }
     }
     return(furthestDistanceSq);
 }
Ejemplo n.º 5
0
        internal override void reorganize(RTree rtree)
        {
            int countdownIndex = rtree.maxNodeEntries - 1;

            for (int index = 0; index < entryCount; index++)
            {
                if (entries[index] == null)
                {
                    while (entries[countdownIndex] == null && countdownIndex > index)
                    {
                        countdownIndex--;
                    }
                    entries[index]          = entries[countdownIndex];
                    entries[countdownIndex] = null;
                }
            }
        }
Ejemplo n.º 6
0
        private int pickNext(RTree rTree, NodeLeaf newNode)
        {
            double maxDifference = double.NegativeInfinity;
            int    next          = 0;
            int    nextGroup     = 0;

            maxDifference = double.NegativeInfinity;
#if RtreeCheck
            Console.WriteLine("pickNext()");
#endif
            for (int i = 0; i < rTree.maxNodeEntries; i++)
            {
                if (rTree.entryStatus[i] == ((byte)RTree.EntryStatus.unassigned))
                {
                    if (entries[i] == null)
                    {
                        throw new UnexpectedException("Error: Node " + this + ", entry " + i + " is null");
                    }
                    Rectangle entryR          = entries[i].Value;
                    double    nIncrease       = minimumBoundingRectangle.Enlargement(ref entryR);
                    double    newNodeIncrease = newNode.minimumBoundingRectangle.Enlargement(ref entryR);
                    double    difference      = Math.Abs(nIncrease - newNodeIncrease);
                    if (difference > maxDifference)
                    {
                        next = i;
                        if (nIncrease < newNodeIncrease)
                        {
                            nextGroup = 0;
                        }
                        else if (newNodeIncrease < nIncrease)
                        {
                            nextGroup = 1;
                        }
                        else if (minimumBoundingRectangle.Area < newNode.minimumBoundingRectangle.Area)
                        {
                            nextGroup = 0;
                        }
                        else if (newNode.minimumBoundingRectangle.Area < minimumBoundingRectangle.Area)
                        {
                            nextGroup = 1;
                        }
                        else if (newNode.entryCount < rTree.maxNodeEntries / 2)
                        {
                            nextGroup = 0;
                        }
                        else
                        {
                            nextGroup = 1;
                        }
                        maxDifference = difference;
                    }
#if RtreeCheck
                    Console.WriteLine("Entry " + i + " group0 increase = " + nIncrease + ", group1 increase = " + newNodeIncrease + ", diff = " + difference + ", MaxDiff = " + maxDifference + " (entry " + next + ")");
#endif
                }
            }
            rTree.entryStatus[next] = ((byte)RTree.EntryStatus.assigned);
            if (nextGroup == 0)
            {
                Update();
                Rectangle r = entries[next].Value;
                if (r.MinX < minimumBoundingRectangle.MinX)
                {
                    minimumBoundingRectangle.MinX = r.MinX;
                }
                if (r.MinY < minimumBoundingRectangle.MinY)
                {
                    minimumBoundingRectangle.MinY = r.MinY;
                }
                if (r.MaxX > minimumBoundingRectangle.MaxX)
                {
                    minimumBoundingRectangle.MaxX = r.MaxX;
                }
                if (r.MaxY > minimumBoundingRectangle.MaxY)
                {
                    minimumBoundingRectangle.MaxY = r.MaxY;
                }
                entryCount++;
            }
            else
            {
                // move to new node.
                Rectangle r = entries[next].Value;
                newNode.addEntry(ref r);
                entries[next] = null;
            }
            return(next);
        }
Ejemplo n.º 7
0
        internal NodeLeaf splitNode(RTree rTree, Rectangle r)
        {
            // [Pick first entry for each group] Apply algorithm pickSeeds to
            // choose two entries to be the first elements of the groups. Assign
            // each to a group.

            // debug code

            /*double initialArea = 0;
             *   if (log.isDebugEnabled())
             *   {
             *     double unionMinX = Math.Min(n.mbrMinX, newRectMinX);
             *     double unionMinY = Math.Min(n.mbrMinY, newRectMinY);
             *     double unionMaxX = Math.Max(n.mbrMaxX, newRectMaxX);
             *     double unionMaxY = Math.Max(n.mbrMaxY, newRectMaxY);
             *
             *     initialArea = (unionMaxX - unionMinX) * (unionMaxY - unionMinY);
             *   }*/

            System.Array.Copy(rTree.initialEntryStatus, 0, rTree.entryStatus, 0, rTree.maxNodeEntries);
            Update();
            NodeLeaf newNode = null;

            newNode = new NodeLeaf(level, rTree.maxNodeEntries);

            pickSeeds(rTree, ref r, newNode); // this also sets the entryCount to 1

            // [Check if done] If all entries have been assigned, stop. If one group has so few entries that all the rest must be assigned to it in
            // order for it to have the minimum number m, assign them and stop.
            while (entryCount + newNode.entryCount < rTree.maxNodeEntries + 1)
            {
                if (rTree.maxNodeEntries + 1 - newNode.entryCount == rTree.minNodeEntries)
                {
                    // assign all remaining entries to original node
                    for (int i = 0; i < rTree.maxNodeEntries; i++)
                    {
                        if (rTree.entryStatus[i] == ((byte)RTree.EntryStatus.unassigned))
                        {
                            rTree.entryStatus[i] = ((byte)RTree.EntryStatus.assigned);

                            if (entries[i].Value.MinX < minimumBoundingRectangle.MinX)
                            {
                                minimumBoundingRectangle.MinX = entries[i].Value.MinX;
                            }
                            if (entries[i].Value.MinY < minimumBoundingRectangle.MinY)
                            {
                                minimumBoundingRectangle.MinY = entries[i].Value.MinY;
                            }
                            if (entries[i].Value.MaxX > minimumBoundingRectangle.MaxX)
                            {
                                minimumBoundingRectangle.MaxX = entries[i].Value.MaxX;
                            }
                            if (entries[i].Value.MaxY > minimumBoundingRectangle.MaxY)
                            {
                                minimumBoundingRectangle.MaxY = entries[i].Value.MaxY;
                            }
                            entryCount++;
                        }
                    }
                    break;
                }
                if (rTree.maxNodeEntries + 1 - entryCount == rTree.minNodeEntries)
                {
                    // assign all remaining entries to new node
                    for (int i = 0; i < rTree.maxNodeEntries; i++)
                    {
                        if (rTree.entryStatus[i] == ((byte)RTree.EntryStatus.unassigned))
                        {
                            rTree.entryStatus[i] = ((byte)RTree.EntryStatus.assigned);
                            Rectangle entryR = entries[i].Value;
                            newNode.addEntry(ref entryR);
                            entries[i] = null;
                        }
                    }
                    break;
                }

                // [Select entry to assign] Invoke algorithm pickNext to choose the next entry to assign. Add it to the group whose covering rectangle
                // will have to be enlarged least to accommodate it. Resolve ties by adding the entry to the group with smaller area, then to the
                // the one with fewer entries, then to either. Repeat from S2
                pickNext(rTree, newNode);
            }

            reorganize(rTree);

            // check that the MBR stored for each node is correct.
#if RtreeCheck
            if (!this.minimumBoundingRectangle.Equals(calculateMBR()))
            {
                throw new UnexpectedException("Error: splitNode old node MBR wrong");
            }
            if (!newNode.minimumBoundingRectangle.Equals(newNode.calculateMBR()))
            {
                throw new UnexpectedException("Error: splitNode new node MBR wrong");
            }
#endif

#if RtreeCheck
            double newArea            = minimumBoundingRectangle.Area + newNode.minimumBoundingRectangle.Area;
            double percentageIncrease = (100 * (newArea - initialArea)) / initialArea;
            Console.WriteLine("Node " + this + " split. New area increased by " + percentageIncrease + "%");
#endif

            return(newNode);
        }
Ejemplo n.º 8
0
        private void pickSeeds(RTree rTree, ref Rectangle r, NodeLeaf newNode)
        {
            // Find extreme rectangles along all dimension. Along each dimension, find the entry whose rectangle has the highest low side, and the one
            // with the lowest high side. Record the separation.
            double maxNormalizedSeparation = -1; // initialize to -1 so that even overlapping rectangles will be considered for the seeds
            int    highestLowIndex         = -1;
            int    lowestHighIndex         = -1;

            Update();
            // for the purposes of picking seeds, take the MBR of the node to include the new rectangle aswell.
            if (r.MinX < minimumBoundingRectangle.MinX)
            {
                minimumBoundingRectangle.MinX = r.MinX;
            }
            if (r.MinY < minimumBoundingRectangle.MinY)
            {
                minimumBoundingRectangle.MinY = r.MinY;
            }
            if (r.MaxX > minimumBoundingRectangle.MaxX)
            {
                minimumBoundingRectangle.MaxX = r.MaxX;
            }
            if (r.MaxY > minimumBoundingRectangle.MaxY)
            {
                minimumBoundingRectangle.MaxY = r.MaxY;
            }

            double mbrLenX = minimumBoundingRectangle.MaxX - minimumBoundingRectangle.MinX;
            double mbrLenY = minimumBoundingRectangle.MaxY - minimumBoundingRectangle.MinY;

#if RtreeCheck
            Console.WriteLine("pickSeeds(): Node = " + this);
#endif

            double tempHighestLow      = r.MinX;
            int    tempHighestLowIndex = -1; // -1 indicates the new rectangle is the seed

            double tempLowestHigh      = r.MaxX;
            int    tempLowestHighIndex = -1; // -1 indicates the new rectangle is the seed

            for (int i = 0; i < entryCount; i++)
            {
                double tempLow = entries[i].Value.MinX;
                if (tempLow >= tempHighestLow)
                {
                    tempHighestLow      = tempLow;
                    tempHighestLowIndex = i;
                } // ensure that the same index cannot be both lowestHigh and highestLow
                else
                {
                    double tempHigh = entries[i].Value.MaxX;
                    if (tempHigh <= tempLowestHigh)
                    {
                        tempLowestHigh      = tempHigh;
                        tempLowestHighIndex = i;
                    }
                }

                // PS2 [Adjust for shape of the rectangle cluster] Normalize the separations by dividing by the widths of the entire set along the corresponding dimension
                double normalizedSeparation = mbrLenX == 0 ? 1 : (tempHighestLow - tempLowestHigh) / mbrLenX;
                if (normalizedSeparation > 1 || normalizedSeparation < -1)
                {
                    Console.WriteLine("Invalid normalized separation X");
                }

#if RtreeCheck
                Console.WriteLine("Entry " + i + ", dimension X: HighestLow = " + tempHighestLow + " (index " + tempHighestLowIndex + ")" + ", LowestHigh = " + tempLowestHigh + " (index " + tempLowestHighIndex + ", NormalizedSeparation = " + normalizedSeparation);
#endif
                // PS3 [Select the most extreme pair] Choose the pair with the greatest normalized separation along any dimension.
                // Note that if negative it means the rectangles overlapped. However still include overlapping rectangles if that is the only choice available.
                if (normalizedSeparation >= maxNormalizedSeparation)
                {
                    highestLowIndex         = tempHighestLowIndex;
                    lowestHighIndex         = tempLowestHighIndex;
                    maxNormalizedSeparation = normalizedSeparation;
                }
            }

            // Repeat for the Y dimension
            tempHighestLow      = r.MinY;
            tempHighestLowIndex = -1; // -1 indicates the new rectangle is the seed

            tempLowestHigh      = r.MaxY;
            tempLowestHighIndex = -1; // -1 indicates the new rectangle is the seed

            for (int i = 0; i < entryCount; i++)
            {
                double tempLow = entries[i].Value.MinY;
                if (tempLow >= tempHighestLow)
                {
                    tempHighestLow      = tempLow;
                    tempHighestLowIndex = i;
                } // ensure that the same index cannot be both lowestHigh and highestLow
                else
                {
                    double tempHigh = entries[i].Value.MaxY;
                    if (tempHigh <= tempLowestHigh)
                    {
                        tempLowestHigh      = tempHigh;
                        tempLowestHighIndex = i;
                    }
                }

                // PS2 [Adjust for shape of the rectangle cluster] Normalize the separations by dividing by the widths of the entire set along the corresponding dimension
                double normalizedSeparation = mbrLenY == 0 ? 1 : (tempHighestLow - tempLowestHigh) / mbrLenY;
                if (normalizedSeparation > 1 || normalizedSeparation < -1)
                {
                    throw new UnexpectedException("Invalid normalized separation Y");
                }
#if RtreeCheck
                Console.WriteLine("Entry " + i + ", dimension Y: HighestLow = " + tempHighestLow + " (index " + tempHighestLowIndex + ")" + ", LowestHigh = " + tempLowestHigh + " (index " + tempLowestHighIndex + ", NormalizedSeparation = " + normalizedSeparation);
#endif
                // PS3 [Select the most extreme pair] Choose the pair with the greatest normalized separation along any dimension.
                // Note that if negative it means the rectangles overlapped. However still include overlapping rectangles if that is the only choice available.
                if (normalizedSeparation >= maxNormalizedSeparation)
                {
                    highestLowIndex         = tempHighestLowIndex;
                    lowestHighIndex         = tempLowestHighIndex;
                    maxNormalizedSeparation = normalizedSeparation;
                }
            }

            // At this point it is possible that the new rectangle is both highestLow and lowestHigh. This can happen if all rectangles in the node overlap the new rectangle.
            // Resolve this by declaring that the highestLowIndex is the lowest Y and, the lowestHighIndex is the largest X (but always a different rectangle)
            if (highestLowIndex == lowestHighIndex)
            {
                highestLowIndex = -1;
                double tempMinY = r.MinY;
                lowestHighIndex = 0;
                double tempMaxX = entries[0].Value.MaxX;

                for (int i = 1; i < entryCount; i++)
                {
                    if (entries[i].Value.MinY < tempMinY)
                    {
                        tempMinY        = entries[i].Value.MinY;
                        highestLowIndex = i;
                    }
                    else if (entries[i].Value.MaxX > tempMaxX)
                    {
                        tempMaxX        = entries[i].Value.MaxX;
                        lowestHighIndex = i;
                    }
                }
            }

            // highestLowIndex is the seed for the new node.
            if (highestLowIndex == -1)
            {
                newNode.addEntry(ref r);
            }
            else
            {
                Rectangle entryRectangle = entries[highestLowIndex].Value;
                newNode.addEntry(ref entryRectangle);
                entries[highestLowIndex] = r; // move the new rectangle into the space vacated by the seed for the new node
            }

            // lowestHighIndex is the seed for the original node.
            if (lowestHighIndex == -1)
            {
                lowestHighIndex = highestLowIndex;
            }

            rTree.entryStatus[lowestHighIndex] = ((byte)RTree.EntryStatus.assigned);
            entryCount = 1;
            minimumBoundingRectangle = entries[lowestHighIndex].Value;
        }
Ejemplo n.º 9
0
 internal abstract void reorganize(RTree rtree);
Ejemplo n.º 10
0
 internal abstract double Nearest(Point p, RTree rTree, double furthestDistanceSq, PriorityQueueRTree nearestRectangles);