/// <summary>Runs the Greedy search algorithm algorithm on a graph.</summary>
/// <param name="start">The node to start at.</param>
/// <param name="neighbors">Step function for all neigbors of a given node.</param>
/// <param name="heuristic">Computes the heuristic value of a given node in a graph.</param>
/// <param name="goal">Predicate for determining if we have reached the goal node.</param>
/// <returns>Stepper of the shortest path or null if no path exists.</returns>
public static Stepper <T> Greedy(T start, Neighbors neighbors, Heuristic heuristic, Goal goal)
{
// using a heap (aka priority queue) to store nodes based on their computed heuristic value
Heap <Greedy_Node> fringe = new HeapArray <Greedy_Node>(
// NOTE: I just reversed the order of left and right because smaller values are higher priority
(Greedy_Node left, Greedy_Node right) => { return(Compute <Math> .Compare(right.Priority, left.Priority)); });
// push starting node
Greedy_Node start_node = new Greedy_Node(null, start, default(Math));
fringe.Enqueue(start_node);
// run the algorithm
while (fringe.Count != 0)
{
Greedy_Node current = fringe.Dequeue();
if (goal(current.Value))
{
return(Greedy_BuildPath(current));
}
else
{
neighbors(current.Value,
(T neighbor) =>
{
Greedy_Node newNode = new Greedy_Node(current, neighbor, heuristic(neighbor));
fringe.Enqueue(newNode);
});
}
}
return(null); // goal node was not reached (no path exists)
}
/// <summary>Runs the A* search algorithm algorithm on a graph.</summary>
/// <param name="start">The node to start at.</param>
/// <param name="neighbors">Step function for all neigbors of a given node.</param>
/// <param name="heuristic">Computes the heuristic value of a given node in a graph.</param>
/// <param name="cost">Computes the cost of moving from the current node to a specific neighbor.</param>
/// <param name="goal">Predicate for determining if we have reached the goal node.</param>
/// <returns>Stepper of the shortest path or null if no path exists.</returns>
public static Stepper <T> Astar(T start, Neighbors neighbors, Heuristic heuristic, Cost cost, Goal goal)
{
// using a heap (aka priority queue) to store nodes based on their computed A* f(n) value
Heap <Astar_Node> fringe = new HeapArray <Astar_Node>(
// NOTE: Typical A* implementations prioritize smaller values
(Astar_Node left, Astar_Node right) =>
{
Comparison comparison = Compute.Compare <Math>(right.Priority, left.Priority);
return(comparison);
});
// using a map (aka dictionary) to store costs from start to current nodes
Map <Math, Astar_Node> computed_costs = new MapHashArray <Math, Astar_Node>();
// construct the f(n) for this A* execution
Astar_function function = (T node, Astar_Node previous) =>
{
Math previousCost = computed_costs.Get(previous);
Math currentCost = cost(previous.Value, node);
Math costFromStart = Compute.Add <Math>(previousCost, currentCost);
Math hueristic = heuristic(node);
return(Compute.Add <Math>(costFromStart, hueristic));
};
// push starting node
Astar_Node start_node = new Astar_Node(null, start, default(Math));
fringe.Enqueue(start_node);
computed_costs.Add(start_node, default(Math));
// run the algorithm
while (fringe.Count != 0)
{
Astar_Node current = fringe.Dequeue();
if (goal(current.Value))
{
return(Astar_BuildPath(current));
}
else
{
neighbors(current.Value,
(T neighbor) =>
{
Astar_Node newNode = new Astar_Node(current, neighbor, function(neighbor, current));
Math costValue = Compute.Add <Math>(computed_costs.Get(current), cost(current.Value, neighbor));
computed_costs.Add(newNode, costValue);
fringe.Enqueue(newNode);
});
}
}
return(null); // goal node was not reached (no path exists)
}
[TestMethod] public void Dequeue_Testing()
{
void Test <T>(T[] values, Compare <T> compare)
{
T[] clonedValues = (T[])values.Clone();
Towel.Sort.Shuffle(clonedValues);
IHeap <T> heap = new HeapArray <T>(compare);
clonedValues.Stepper(x => heap.Enqueue(x));
foreach (T value in values)
{
T dequeue = heap.Dequeue();
Assert.IsTrue(value.Equals(dequeue));
}
}
{ // int compare
const int count = 100;
int[] values = new int[count];
Stepper.Iterate(count, i => values[i] = i);
Array.Sort(values, (a, b) => - a.CompareTo(b));
Test(values, (a, b) => Compare.Wrap(a.CompareTo(b)));
}
{ // string compare
const int count = 100;
string[] values = new string[count];
Stepper.Iterate(count, i => values[i] = i.ToString());
Array.Sort(values, (a, b) => - a.CompareTo(b));
Test(values, (a, b) => Compare.Wrap(a.CompareTo(b)));
}
{ // int reverse compare
const int count = 100;
int[] values = new int[count];
Stepper.Iterate(count, i => values[i] = i);
Array.Sort(values);
Test(values, (a, b) => Compare.Wrap(-a.CompareTo(b)));
}
{ // string reverse compare
const int count = 100;
string[] values = new string[count];
Stepper.Iterate(count, i => values[i] = i.ToString());
Array.Sort(values);
Test(values, (a, b) => Compare.Wrap(-a.CompareTo(b)));
}
}
/// <summary>Runs the A* search algorithm algorithm on a graph.</summary>
/// <param name="start">The node to start at.</param>
/// <param name="neighbors">Step function for all neigbors of a given node.</param>
/// <param name="heuristic">Computes the heuristic value of a given node in a graph.</param>
/// <param name="cost">Computes the cost of moving from the current node to a specific neighbor.</param>
/// <param name="goal">Predicate for determining if we have reached the goal node.</param>
/// <returns>Stepper of the shortest path or null if no path exists.</returns>
public static Stepper <NODE> Graph <NODE, NUMERIC>(NODE start, Neighbors <NODE> neighbors, Heuristic <NODE, NUMERIC> heuristic, Cost <NODE, NUMERIC> cost, Goal <NODE> goal)
{
// using a heap (aka priority queue) to store nodes based on their computed A* f(n) value
IHeap <AstarNode <NODE, NUMERIC> > fringe = new HeapArray <AstarNode <NODE, NUMERIC> >(
// NOTE: Typical A* implementations prioritize smaller values
(a, b) => Compute.Compare(b.Priority, a.Priority));
// push starting node
fringe.Enqueue(
new AstarNode <NODE, NUMERIC>(
null,
start,
default(NUMERIC),
Constant <NUMERIC> .Zero));
// run the algorithm
while (fringe.Count != 0)
{
AstarNode <NODE, NUMERIC> current = fringe.Dequeue();
if (goal(current.Value))
{
return(BuildPath(current));
}
else
{
neighbors(current.Value,
(NODE neighbor) =>
{
NUMERIC costValue = Compute.Add(current.Cost, cost(current.Value, neighbor));
fringe.Enqueue(
new AstarNode <NODE, NUMERIC>(
current,
neighbor,
Compute.Add(heuristic(neighbor), costValue),
costValue));
});
}
}
return(null); // goal node was not reached (no path exists)
}
/// <summary>Runs the Greedy search algorithm algorithm on a graph.</summary>
/// <param name="start">The node to start at.</param>
/// <param name="neighbors">Step function for all neigbors of a given node.</param>
/// <param name="heuristic">Computes the heuristic value of a given node in a graph.</param>
/// <param name="goal">Predicate for determining if we have reached the goal node.</param>
/// <returns>Stepper of the shortest path or null if no path exists.</returns>
public static Stepper <NODE> Graph <NODE, NUMERIC>(NODE start, Neighbors <NODE> neighbors, Heuristic <NODE, NUMERIC> heuristic, Goal <NODE> goal)
{
// using a heap (aka priority queue) to store nodes based on their computed heuristic value
IHeap <GreedyNode <NODE, NUMERIC> > fringe = new HeapArray <GreedyNode <NODE, NUMERIC> >(
// NOTE: Typical graph search implementations prioritize smaller values
(a, b) => Compute.Compare(b.Priority, a.Priority));
// push starting node
fringe.Enqueue(
new GreedyNode <NODE, NUMERIC>(
null,
start,
default(NUMERIC)));
// run the algorithm
while (fringe.Count != 0)
{
GreedyNode <NODE, NUMERIC> current = fringe.Dequeue();
if (goal(current.Value))
{
return(BuildPath(current));
}
else
{
neighbors(current.Value,
(NODE neighbor) =>
{
fringe.Enqueue(
new GreedyNode <NODE, NUMERIC>(
current,
neighbor,
heuristic(neighbor)));
});
}
}
return(null); // goal node was not reached (no path exists)
}
static void Main(string[] args)
{
Random random = new Random();
int test = 10;
Console.WriteLine("You are runnning the Data Structures example.");
Console.WriteLine("======================================================");
Console.WriteLine();
#region Link (aka Tuple)
Console.WriteLine(" Link------------------------------------");
Console.WriteLine();
Console.WriteLine(" A \"Link\" is like a System.Tuple that implements");
Console.WriteLine(" Towel.DataStructures.DataStructure. A Link/Tuple is");
Console.WriteLine(" used when you have a small, known-sized set of objects");
Console.WriteLine(" that you want to bundle together without making a custom");
Console.WriteLine(" custom class.");
Console.WriteLine();
Link link = new Link <int, int, int, int, int, int>(0, 1, 2, 3, 4, 5);
Console.Write(" Traversal: ");
link.Stepper(i => Console.Write(i));
Console.WriteLine();
Console.WriteLine(" Size: " + link.Size);
Console.WriteLine();
#endregion
#region Indexed (aka Array)
Console.WriteLine(" Indexed---------------------------------");
Console.WriteLine();
Console.WriteLine(" An \"Indexed\" is just a wrapper for arrays that implements");
Console.WriteLine(" Towel.DataStructures.DataStructure. An array is used when");
Console.WriteLine(" dealing with static-sized, known-sized sets of data. Arrays");
Console.WriteLine(" can be sorted along 1 dimensions for binary searching algorithms.");
Console.WriteLine();
IIndexed <int> indexed = new IndexedArray <int>(test);
Console.Write(" Filling in (0-" + (test - 1) + ")...");
for (int i = 0; i < test; i++)
{
indexed[i] = i;
}
Console.WriteLine();
Console.Write(" Traversal: ");
indexed.Stepper(i => Console.Write(i));
Console.WriteLine();
Console.WriteLine(" Length: " + indexed.Length);
Console.WriteLine();
#endregion
#region Addable (aka List)
Console.WriteLine(" Addable---------------------------------");
Console.WriteLine();
Console.WriteLine(" An \"Addable\" is like an IList that implements");
Console.WriteLine(" Towel.DataStructures.DataStructure. \"AddableArray\" is");
Console.WriteLine(" the array implementation while \"AddableLinked\" is the");
Console.WriteLine(" the linked-list implementation. An Addable/List is used");
Console.WriteLine(" when dealing with an unknown quantity of data that you");
Console.WriteLine(" will likely have to enumerate/step through everything. The");
Console.WriteLine(" AddableArray shares the properties of an Indexed/Array in");
Console.WriteLine(" that it can be relateively quickly sorted along 1 dimensions");
Console.WriteLine(" for binary search algorithms.");
Console.WriteLine();
// AddableArray ---------------------------------------
IAddable <int> addableArray = new AddableArray <int>(test);
Console.Write(" [AddableArray] Adding (0-" + (test - 1) + ")...");
for (int i = 0; i < test; i++)
{
addableArray.Add(i);
}
Console.WriteLine();
Console.Write(" [AddableArray] Traversal: ");
addableArray.Stepper(i => Console.Write(i));
Console.WriteLine();
Console.WriteLine(" [AddableArray] Count: " + addableArray.Count);
addableArray.Clear(); // Clears the addable
Console.WriteLine();
// AddableLinked ---------------------------------------
IAddable <int> addableLinked = new AddableLinked <int>();
Console.Write(" [AddableLinked] Adding (0-" + (test - 1) + ")...");
for (int i = 0; i < test; i++)
{
addableLinked.Add(i);
}
Console.WriteLine();
Console.Write(" [AddableLinked] Traversal: ");
addableLinked.Stepper(i => Console.Write(i));
Console.WriteLine();
Console.WriteLine(" [AddableLinked] Count: " + addableLinked.Count);
addableLinked.Clear(); // Clears the addable
Console.WriteLine();
#endregion
#region FirstInLastOut (aka stack)
{
Console.WriteLine(" FirstInLastOut---------------------------------");
Console.WriteLine();
Console.WriteLine(" An \"FirstInLastOut\" is a Stack that implements");
Console.WriteLine(" Towel.DataStructures.DataStructure. \"FirstInLastOutArray\" is");
Console.WriteLine(" the array implementation while \"FirstInLastOutLinked\" is the");
Console.WriteLine(" the linked-list implementation. A FirstInLastOut/Stack is used");
Console.WriteLine(" specifically when you need the algorithm provided by the Push");
Console.WriteLine(" and Pop functions.");
Console.WriteLine();
IFirstInLastOut <int> firstInLastOutArray = new FirstInLastOutArray <int>();
Console.Write(" [FirstInLastOutArray] Pushing (0-" + (test - 1) + ")...");
for (int i = 0; i < test; i++)
{
firstInLastOutArray.Push(i);
}
Console.WriteLine();
Console.Write(" [FirstInLastOutArray] Traversal: ");
firstInLastOutArray.Stepper(i => Console.Write(i));
Console.WriteLine();
Console.WriteLine(" [FirstInLastOutArray] Pop: " + firstInLastOutArray.Pop());
Console.WriteLine(" [FirstInLastOutArray] Pop: " + firstInLastOutArray.Pop());
Console.WriteLine(" [FirstInLastOutArray] Peek: " + firstInLastOutArray.Peek());
Console.WriteLine(" [FirstInLastOutArray] Pop: " + firstInLastOutArray.Pop());
Console.WriteLine(" [FirstInLastOutArray] Count: " + firstInLastOutArray.Count);
firstInLastOutArray.Clear(); // Clears the firstInLastOut
Console.WriteLine();
IFirstInLastOut <int> firstInLastOutLinked = new FirstInLastOutLinked <int>();
Console.Write(" [FirstInLastOutLinked] Pushing (0-" + (test - 1) + ")...");
for (int i = 0; i < test; i++)
{
firstInLastOutLinked.Push(i);
}
Console.WriteLine();
Console.Write(" [FirstInLastOutLinked] Traversal: ");
firstInLastOutLinked.Stepper(i => Console.Write(i));
Console.WriteLine();
Console.WriteLine(" [FirstInLastOutLinked] Pop: " + firstInLastOutLinked.Pop());
Console.WriteLine(" [FirstInLastOutLinked] Pop: " + firstInLastOutLinked.Pop());
Console.WriteLine(" [FirstInLastOutLinked] Peek: " + firstInLastOutLinked.Peek());
Console.WriteLine(" [FirstInLastOutLinked] Pop: " + firstInLastOutLinked.Pop());
Console.WriteLine(" [FirstInLastOutLinked] Count: " + firstInLastOutLinked.Count);
firstInLastOutLinked.Clear(); // Clears the firstInLastOut
Console.WriteLine();
}
#endregion
#region FirstInFirstOut (aka Queue)
{
Console.WriteLine(" FirstInFirstOut---------------------------------");
Console.WriteLine();
Console.WriteLine(" An \"FirstInFirstOut\" is a Queue that implements");
Console.WriteLine(" Towel.DataStructures.DataStructure. \"FirstInFirstOutArray\" is");
Console.WriteLine(" the array implementation while \"FirstInFirstOutLinked\" is the");
Console.WriteLine(" the linked-list implementation. A FirstInFirstOut/Stack is used");
Console.WriteLine(" specifically when you need the algorithm provided by the Queue");
Console.WriteLine(" and Dequeue functions.");
Console.WriteLine();
IFirstInFirstOut <int> firstInFirstOutArray = new FirstInFirstOutArray <int>();
Console.Write(" [FirstInFirstOutArray] Enqueuing (0-" + (test - 1) + ")...");
for (int i = 0; i < test; i++)
{
firstInFirstOutArray.Enqueue(i);
}
Console.WriteLine();
Console.Write(" [FirstInFirstOutArray] Traversal: ");
firstInFirstOutArray.Stepper(i => Console.Write(i));
Console.WriteLine();
Console.WriteLine(" [FirstInFirstOutArray] Dequeue: " + firstInFirstOutArray.Dequeue());
Console.WriteLine(" [FirstInFirstOutArray] Dequeue: " + firstInFirstOutArray.Dequeue());
Console.WriteLine(" [FirstInFirstOutArray] Peek: " + firstInFirstOutArray.Peek());
Console.WriteLine(" [FirstInFirstOutArray] Dequeue: " + firstInFirstOutArray.Dequeue());
Console.WriteLine(" [FirstInFirstOutArray] Count: " + firstInFirstOutArray.Count);
firstInFirstOutArray.Clear(); // Clears the firstInLastOut
Console.WriteLine();
IFirstInFirstOut <int> firstInFirstOutLinked = new FirstInFirstOutLinked <int>();
Console.Write(" [FirstInFirstOutLinked] Enqueuing (0-" + (test - 1) + ")...");
for (int i = 0; i < test; i++)
{
firstInFirstOutLinked.Enqueue(i);
}
Console.WriteLine();
Console.Write(" [FirstInFirstOutLinked] Traversal: ");
firstInFirstOutLinked.Stepper(i => Console.Write(i));
Console.WriteLine();
Console.WriteLine(" [FirstInFirstOutLinked] Pop: " + firstInFirstOutLinked.Dequeue());
Console.WriteLine(" [FirstInFirstOutLinked] Pop: " + firstInFirstOutLinked.Dequeue());
Console.WriteLine(" [FirstInFirstOutLinked] Peek: " + firstInFirstOutLinked.Peek());
Console.WriteLine(" [FirstInFirstOutLinked] Pop: " + firstInFirstOutLinked.Dequeue());
Console.WriteLine(" [FirstInFirstOutLinked] Count: " + firstInFirstOutLinked.Count);
firstInFirstOutLinked.Clear(); // Clears the firstInLastOut
Console.WriteLine();
}
#endregion
#region Heap
{
Console.WriteLine(" Heap---------------------------------");
Console.WriteLine();
Console.WriteLine(" An \"Heap\" is a binary tree that stores items based on priorities.");
Console.WriteLine(" It implements Towel.DataStructures.DataStructure like the others.");
Console.WriteLine(" It uses sifting algorithms to move nodes vertically through itself.");
Console.WriteLine(" It is often the best data structure for standard priority queues.");
Console.WriteLine(" \"HeapArray\" is an implementation where the tree has been flattened");
Console.WriteLine(" into an array.");
Console.WriteLine();
Console.WriteLine(" Let's say the priority is how close a number is to \"5\".");
Console.WriteLine(" So \"Dequeue\" will give us the next closest value to \"5\".");
Comparison Priority(int a, int b)
{
int _a = Compute.AbsoluteValue(a - 5);
int _b = Compute.AbsoluteValue(b - 5);
Comparison comparison = Compare.Wrap(_b.CompareTo(_a));
return(comparison);
}
Console.WriteLine();
IHeap <int> heapArray = new HeapArray <int>(Priority);
Console.Write(" [HeapArray] Enqueuing (0-" + (test - 1) + ")...");
for (int i = 0; i < test; i++)
{
heapArray.Enqueue(i);
}
Console.WriteLine();
Console.WriteLine(" [HeapArray] Dequeue: " + heapArray.Dequeue());
Console.WriteLine(" [HeapArray] Dequeue: " + heapArray.Dequeue());
Console.WriteLine(" [HeapArray] Peek: " + heapArray.Peek());
Console.WriteLine(" [HeapArray] Dequeue: " + heapArray.Dequeue());
Console.WriteLine(" [HeapArray] Count: " + heapArray.Count);
heapArray.Clear(); // Clears the heapArray
Console.WriteLine();
}
#endregion
#region Tree
//Console.WriteLine(" Tree-----------------------------");
//Tree<int> tree_Map = new TreeMap<int>(0, Compute.Equal, Hash.Default);
//for (int i = 1; i < test; i++)
//{
// tree_Map.Add(i, i / Compute.SquareRoot(i));
//}
//Console.Write(" Children of 0 (root): ");
//tree_Map.Children(0, (int i) => { Console.Write(i + " "); });
//Console.WriteLine();
//Console.Write(" Children of " + ((int)System.Math.Sqrt(test) - 1) + " (root): ");
//tree_Map.Children(((int)System.Math.Sqrt(test) - 1), (int i) => { Console.Write(i + " "); });
//Console.WriteLine();
//Console.Write(" Traversal: ");
//tree_Map.Stepper((int i) => { Console.Write(i + " "); });
//Console.WriteLine();
//Console.WriteLine();
#endregion
#region AVL Tree
{
Console.WriteLine(" AvlTree------------------------------------------------");
Console.WriteLine();
Console.WriteLine(" An AVL Tree is a sorted binary tree.");
Console.WriteLine(" It implements Towel.DataStructures.DataStructure like the others.");
Console.WriteLine(" It allows for very fast 1D ranged queries/traversals.");
Console.WriteLine(" It is very similar to an Red Black tree, but uses a different sorting algorithm.");
Console.WriteLine();
IAvlTree <int> avlTree = new AvlTreeLinked <int>();
Console.Write(" Adding (0-" + (test - 1) + ")...");
for (int i = 0; i < test; i++)
{
avlTree.Add(i);
}
Console.WriteLine();
Console.Write(" Traversal: ");
avlTree.Stepper(i => Console.Write(i));
Console.WriteLine();
//// Note: Because the nodes in AVL Tree linked do not have
//// a parent pointer, the IEnumerable "foreach" iteration
//// is extremely slow and should be avoided. It requires
//// a stack for it's iteration.
//
//Console.Write(" Traversal Foreach: ");
//foreach (int i in avlTree)
//{
// Console.Write(i);
//}
//Console.WriteLine();
int minimum = random.Next(1, test / 2);
int maximum = random.Next(1, test / 2) + test / 2;
Console.Write(" Ranged Traversal [" + minimum + "-" + maximum + "]: ");
avlTree.Stepper(i => Console.Write(i), minimum, maximum);
Console.WriteLine();
int removal = random.Next(0, test);
Console.Write(" Remove(" + removal + "): ");
avlTree.Remove(removal);
avlTree.Stepper(i => Console.Write(i));
Console.WriteLine();
int contains = random.Next(0, test);
Console.WriteLine(" Contains(" + contains + "): " + avlTree.Contains(contains));
Console.WriteLine(" Current Least: " + avlTree.CurrentLeast);
Console.WriteLine(" Current Greatest: " + avlTree.CurrentGreatest);
Console.WriteLine(" Count: " + avlTree.Count);
avlTree.Clear(); // Clears the AVL tree
Console.WriteLine();
}
#endregion
#region Red-Black Tree
{
Console.WriteLine(" Red-Black Tree------------------------------------------------");
Console.WriteLine();
Console.WriteLine(" An Red-Black Tree is a sorted binary tree.");
Console.WriteLine(" It implements Towel.DataStructures.DataStructure like the others.");
Console.WriteLine(" It allows for very fast 1D ranged queries/traversals.");
Console.WriteLine(" It is very similar to an AVL tree, but uses a different sorting algorithm.");
Console.WriteLine();
IRedBlackTree <int> redBlackTree = new RedBlackTreeLinked <int>();
Console.Write(" Adding (0-" + (test - 1) + ")...");
for (int i = 0; i < test; i++)
{
redBlackTree.Add(i);
}
Console.WriteLine();
Console.Write(" Traversal: ");
redBlackTree.Stepper(i => Console.Write(i));
Console.WriteLine();
int minimum = random.Next(1, test / 2);
int maximum = random.Next(1, test / 2) + test / 2;
Console.Write(" Ranged Traversal [" + minimum + "-" + maximum + "]: ");
redBlackTree.Stepper(i => Console.Write(i), minimum, maximum);
Console.WriteLine();
int removal = random.Next(0, test);
Console.Write(" Remove(" + removal + "): ");
redBlackTree.Remove(removal);
redBlackTree.Stepper(i => Console.Write(i));
Console.WriteLine();
int contains = random.Next(0, test);
Console.WriteLine(" Contains(" + contains + "): " + redBlackTree.Contains(contains));
Console.WriteLine(" Current Least: " + redBlackTree.CurrentLeast);
Console.WriteLine(" Current Greatest: " + redBlackTree.CurrentGreatest);
Console.WriteLine(" Count: " + redBlackTree.Count);
redBlackTree.Clear(); // Clears the Red Black tree
Console.WriteLine();
}
#endregion
#region BTree
{
Console.WriteLine(" B Tree------------------------------------------------");
Console.WriteLine();
Console.WriteLine(" A B Tree is a sorted binary tree that allows multiple values to");
Console.WriteLine(" be stored per node. This makes it sort of a hybrid between a");
Console.WriteLine(" binary tree and an array. Because multiple values are stored ");
Console.WriteLine(" per node, it means less nodes must be traversed to completely");
Console.WriteLine(" traverse the values in the B tree.");
Console.WriteLine();
Console.WriteLine(" The generic B Tree in Towel is still in development.");
Console.WriteLine();
}
#endregion
#region Set
{
Console.WriteLine(" Set------------------------------------------------");
Console.WriteLine();
Console.WriteLine(" A Set is like an Addable/List, but it does not allow duplicates. Sets are");
Console.WriteLine(" usually implemented using hash codes. Implementations with hash codes");
Console.WriteLine(" usually have very fast \"Contains\" checks to see if a value has already");
Console.WriteLine(" been added to the set.");
Console.WriteLine();
ISet <int> setHashLinked = new SetHashLinked <int>();
Console.Write(" Adding (0-" + (test - 1) + ")...");
for (int i = 0; i < test; i++)
{
setHashLinked.Add(i);
}
Console.WriteLine();
Console.Write(" Traversal: ");
setHashLinked.Stepper(i => Console.Write(i));
Console.WriteLine();
int a = random.Next(0, test);
setHashLinked.Remove(a);
Console.Write(" Remove(" + a + "): ");
setHashLinked.Stepper(i => Console.Write(i));
Console.WriteLine();
int b = random.Next(0, test);
Console.WriteLine(" Contains(" + b + "): " + setHashLinked.Contains(b));
Console.WriteLine(" Count: " + setHashLinked.Count);
Console.WriteLine();
}
#endregion
#region Map (aka Dictionary)
{
Console.WriteLine(" Map------------------------------------------------");
Console.WriteLine();
Console.WriteLine(" A Map (aka Dictionary) is similar to a Set, but it stores two values (a ");
Console.WriteLine(" key and a value). Maps do not allow duplicate keys much like Sets don't");
Console.WriteLine(" allow duplicate values. When provided with the key, the Map uses that key");
Console.WriteLine(" to look up the value that it is associated with. Thus, it allows you to ");
Console.WriteLine(" \"map\" one object to another. As with Sets, Maps are usually implemented");
Console.WriteLine(" using hash codes.");
Console.WriteLine();
// Note: the first generic is the value, the second is the key
IMap <string, int> mapHashLinked = new MapHashLinked <string, int>();
Console.WriteLine(" Let's map each int to its word representation (ex 1 -> One).");
Console.Write(" Adding (0-" + (test - 1) + ")...");
for (int i = 0; i < test; i++)
{
mapHashLinked.Add(i, ((decimal)i).ToEnglishWords());
}
Console.WriteLine();
Console.WriteLine(" Traversal: ");
mapHashLinked.Keys(i => Console.WriteLine(" " + i + "->" + mapHashLinked[i]));
Console.WriteLine();
int a = random.Next(0, test);
mapHashLinked.Remove(a);
Console.Write(" Remove(" + a + "): ");
mapHashLinked.Keys(i => Console.Write(i));
Console.WriteLine();
int b = random.Next(0, test);
Console.WriteLine(" Contains(" + b + "): " + mapHashLinked.Contains(b));
Console.WriteLine(" Count: " + mapHashLinked.Count);
Console.WriteLine();
}
#endregion
#region OmnitreePoints
{
Console.WriteLine(" OmnitreePoints--------------------------------------");
Console.WriteLine();
Console.WriteLine(" An Omnitree is an ND SPT that allows for");
Console.WriteLine(" multidimensional sorting. Any time you need to look");
Console.WriteLine(" items up based on multiple fields/properties, then");
Console.WriteLine(" you might want to use an Omnitree. If you need to");
Console.WriteLine(" perform ranged queries on multiple dimensions, then");
Console.WriteLine(" the Omnitree is the data structure for you.");
Console.WriteLine();
Console.WriteLine(" The \"OmnitreePoints\" stores individual points (vectors),");
Console.WriteLine(" and the \"OmnitreeBounds\" stores bounded objects (spaces).");
Console.WriteLine();
IOmnitreePoints <int, double, string, decimal> omnitree =
new OmnitreePointsLinked <int, double, string, decimal>(
// This is a location delegate. (how to locate the item along each dimension)
(int index, out double a, out string b, out decimal c) =>
{
a = index;
b = index.ToString();
c = index;
});
Console.Write(" Adding (0-" + (test - 1) + ")...");
for (int i = 0; i < test; i++)
{
omnitree.Add(i);
}
Console.WriteLine();
Console.Write(" Traversal: ");
omnitree.Stepper(i => Console.Write(i));
Console.WriteLine();
int minimumXZ = random.Next(1, test / 2);
int maximumXZ = random.Next(1, test / 2) + test / 2;
string minimumY = minimumXZ.ToString();
string maximumY = maximumXZ.ToString();
Console.Write(" Spacial Traversal [" +
"(" + minimumXZ + ", \"" + minimumY + "\", " + minimumXZ + ")->" +
"(" + maximumXZ + ", \"" + maximumY + "\", " + maximumXZ + ")]: ");
omnitree.Stepper(i => Console.Write(i),
minimumXZ, maximumXZ,
minimumY, maximumY,
minimumXZ, maximumXZ);
Console.WriteLine();
// Note: this "look up" is just a very narrow spacial query that (since we know the data)
// wil only give us one result.
int lookUp = random.Next(0, test);
string lookUpToString = lookUp.ToString();
Console.Write(" Look Up (" + lookUp + ", \"" + lookUpToString + "\", " + lookUp + "): ");
omnitree.Stepper(i => Console.Write(i),
lookUp, lookUp,
lookUp.ToString(), lookUp.ToString(),
lookUp, lookUp);
Console.WriteLine();
// Ignoring dimensions on traversals example.
// If you want to ignore a column on a traversal, you can do so like this:
omnitree.Stepper(i => { /*Do Nothing*/ },
lookUp, lookUp,
Omnitree.Bound <string> .None, Omnitree.Bound <string> .None,
Omnitree.Bound <decimal> .None, Omnitree.Bound <decimal> .None);
Console.Write(" Counting Items In a Space [" +
"(" + minimumXZ + ", \"" + minimumY + "\", " + minimumXZ + ")->" +
"(" + maximumXZ + ", \"" + maximumY + "\", " + maximumXZ + ")]: ");
omnitree.CountSubSpace(
minimumXZ, maximumXZ,
minimumY, maximumY,
minimumXZ, maximumXZ);
Console.WriteLine();
int removalMinimum = random.Next(1, test / 2);
int removalMaximum = random.Next(1, test / 2) + test / 2;
string removalMinimumY = removalMinimum.ToString();
string removalMaximumY = removalMaximum.ToString();
Console.Write(" Remove (" + removalMinimum + "-" + removalMaximum + "): ");
omnitree.Remove(
removalMinimum, removalMaximum,
removalMinimumY, removalMaximumY,
removalMinimum, removalMaximum);
omnitree.Stepper(i => Console.Write(i));
Console.WriteLine();
Console.WriteLine(" Dimensions: " + omnitree.Dimensions);
Console.WriteLine(" Count: " + omnitree.Count);
omnitree.Clear(); // Clears the Omnitree
Console.WriteLine();
}
#endregion
#region OmnitreeBounds
{
Console.WriteLine(" OmnitreeBounds--------------------------------------");
Console.WriteLine();
Console.WriteLine(" An Omnitree is an ND SPT that allows for");
Console.WriteLine(" multidimensional sorting. Any time you need to look");
Console.WriteLine(" items up based on multiple fields/properties, then");
Console.WriteLine(" you might want to use an Omnitree. If you need to");
Console.WriteLine(" perform ranged queries on multiple dimensions, then");
Console.WriteLine(" the Omnitree is the data structure for you.");
Console.WriteLine();
Console.WriteLine(" The \"OmnitreePoints\" stores individual points (vectors),");
Console.WriteLine(" and the \"OmnitreeBounds\" stores bounded objects (spaces).");
Console.WriteLine();
IOmnitreeBounds <int, double, string, decimal> omnitree =
new OmnitreeBoundsLinked <int, double, string, decimal>(
// This is a location delegate. (how to locate the item along each dimension)
(int index,
out double min1, out double max1,
out string min2, out string max2,
out decimal min3, out decimal max3) =>
{
string indexToString = index.ToString();
min1 = index; max1 = index;
min2 = indexToString; max2 = indexToString;
min3 = index; max3 = index;
});
Console.Write(" Adding (0-" + (test - 1) + ")...");
for (int i = 0; i < test; i++)
{
omnitree.Add(i);
}
Console.WriteLine();
Console.Write(" Traversal: ");
omnitree.Stepper(i => Console.Write(i));
Console.WriteLine();
int minimumXZ = random.Next(1, test / 2);
int maximumXZ = random.Next(1, test / 2) + test / 2;
string minimumY = minimumXZ.ToString();
string maximumY = maximumXZ.ToString();
Console.Write(" Spacial Traversal [" +
"(" + minimumXZ + ", \"" + minimumY + "\", " + minimumXZ + ")->" +
"(" + maximumXZ + ", \"" + maximumY + "\", " + maximumXZ + ")]: ");
omnitree.StepperOverlapped(i => Console.Write(i),
minimumXZ, maximumXZ,
minimumY, maximumY,
minimumXZ, maximumXZ);
Console.WriteLine();
// Note: this "look up" is just a very narrow spacial query that (since we know the data)
// wil only give us one result.
int lookUpXZ = random.Next(0, test);
string lookUpY = lookUpXZ.ToString();
Console.Write(" Look Up (" + lookUpXZ + ", \"" + lookUpY + "\", " + lookUpXZ + "): ");
omnitree.StepperOverlapped(i => Console.Write(i),
lookUpXZ, lookUpXZ,
lookUpY, lookUpY,
lookUpXZ, lookUpXZ);
Console.WriteLine();
// Ignoring dimensions on traversals example.
// If you want to ignore a dimension on a traversal, you can do so like this:
omnitree.StepperOverlapped(i => { /*Do Nothing*/ },
lookUpXZ, lookUpXZ,
// The "None" means there is no bound, so all values are valid
Omnitree.Bound <string> .None, Omnitree.Bound <string> .None,
Omnitree.Bound <decimal> .None, Omnitree.Bound <decimal> .None);
Console.Write(" Counting Items In a Space [" +
"(" + minimumXZ + ", \"" + minimumY + "\", " + minimumXZ + ")->" +
"(" + maximumXZ + ", \"" + maximumY + "\", " + maximumXZ + ")]: " +
omnitree.CountSubSpaceOverlapped(
minimumXZ, maximumXZ,
minimumY, maximumY,
minimumXZ, maximumXZ));
Console.WriteLine();
int removalMinimumXZ = random.Next(1, test / 2);
int removalMaximumXZ = random.Next(1, test / 2) + test / 2;
string removalMinimumY = removalMinimumXZ.ToString();
string removalMaximumY = removalMaximumXZ.ToString();
Console.Write(" Remove (" + removalMinimumXZ + "-" + removalMaximumXZ + "): ");
omnitree.RemoveOverlapped(
removalMinimumXZ, removalMaximumXZ,
removalMinimumY, removalMaximumY,
removalMinimumXZ, removalMaximumXZ);
omnitree.Stepper(i => Console.Write(i));
Console.WriteLine();
Console.WriteLine(" Dimensions: " + omnitree.Dimensions);
Console.WriteLine(" Count: " + omnitree.Count);
omnitree.Clear(); // Clears the Omnitree
Console.WriteLine();
}
#endregion
#region KD Tree
{
Console.WriteLine(" KD Tree------------------------------------------------");
Console.WriteLine();
Console.WriteLine(" A KD Tree binary tree that stores points sorted along along an");
Console.WriteLine(" arbitrary number of dimensions. So it performs multidimensional");
Console.WriteLine(" sorting similar to the Omnitree (Quadtree/Octree) in Towel, but");
Console.WriteLine(" it uses a completely different algorithm and format.");
Console.WriteLine();
Console.WriteLine(" The generic KD Tree in Towel is still in development.");
Console.WriteLine();
}
#endregion
#region Graph
{
Console.WriteLine(" Graph------------------------------------------------");
Console.WriteLine();
Console.WriteLine(" A Graph is a data structure of nodes and edges. Nodes are values");
Console.WriteLine(" and edges are connections between those values. Graphs are often");
Console.WriteLine(" used to model real world data such as maps, and are often used in");
Console.WriteLine(" path finding algoritms. See the \"Algorithms\" example for path");
Console.WriteLine(" finding examples. This is just an example of how to make a graph.");
Console.WriteLine(" A \"GraphSetOmnitree\" is an implementation where nodes are stored.");
Console.WriteLine(" in a Set and edges are stored in an Omnitree (aka Quadtree).");
Console.WriteLine();
IGraph <int> graphSetOmnitree = new GraphSetOmnitree <int>();
Console.WriteLine(" Adding Nodes (0-" + (test - 1) + ")...");
for (int i = 0; i < test; i++)
{
graphSetOmnitree.Add(i);
}
int edgesPerNode = 3;
Console.WriteLine(" Adding Random Edges (0-3 per node)...");
for (int i = 0; i < test; i++)
{
// lets use a heap to randomize the edges using random priorities
IHeap <(int, int)> heap = new HeapArray <(int, int)>((x, y) => Compare.Wrap(x.Item2.CompareTo(y.Item2)));
for (int j = 0; j < test; j++)
{
if (j != i)
{
heap.Enqueue((j, random.Next()));
}
}
// dequeue some random edges from the heap and add them to the graph
int randomEdgeCount = random.Next(edgesPerNode + 1);
for (int j = 0; j < randomEdgeCount; j++)
{
graphSetOmnitree.Add(i, heap.Dequeue().Item1);
}
}
Console.Write(" Nodes (Traversal): ");
graphSetOmnitree.Stepper(i => Console.Write(i));
Console.WriteLine();
Console.WriteLine(" Edges (Traversal): ");
graphSetOmnitree.Stepper((from, to) => Console.WriteLine(" " + from + "->" + to));
Console.WriteLine();
int a = random.Next(0, test);
Console.Write(" Neighbors (" + a + "):");
graphSetOmnitree.Neighbors(a, i => Console.Write(" " + i));
Console.WriteLine();
int b = random.Next(0, test / 2);
int c = random.Next(test / 2, test);
Console.WriteLine(" Are Adjacent (" + b + ", " + c + "): " + graphSetOmnitree.Adjacent(b, c));
Console.WriteLine(" Node Count: " + graphSetOmnitree.NodeCount);
Console.WriteLine(" Edge Count: " + graphSetOmnitree.EdgeCount);
graphSetOmnitree.Clear(); // Clears the graph
Console.WriteLine();
}
#endregion
#region Trie
{
Console.WriteLine(" Trie------------------------------------------------");
Console.WriteLine();
Console.WriteLine(" A Trie is a tree where portions of the data are stored in each node");
Console.WriteLine(" such that when you traverse the tree to a leaf, you have read the contents");
Console.WriteLine(" of that leaf along the way. Because of this, a Trie allows for its values");
Console.WriteLine(" to share data, which is a form of compression. So a Trie may be used to save");
Console.WriteLine(" memory. A trie may also be a very useful tool in pattern matching, because it");
Console.WriteLine(" it allows for culling based are portions of the data.");
Console.WriteLine();
Console.WriteLine(" The generic Trie in Towel is still in development.");
Console.WriteLine();
}
#endregion
Console.WriteLine("============================================");
Console.WriteLine("Examples Complete...");
Console.ReadLine();
}
IEnumerator Search(Vector3 beginning, Vector3 end)
{
HeapArray <State> openSet = new HeapArray <State>(gridMap.LargestSize);
State goal = gridMap.RetrieveState(end);
State start = gridMap.RetrieveState(beginning);
openSet.Enqueue(start);
HashSet <State> cameFrom = new HashSet <State>();
Vector3[] intermediates = new Vector3[0];
bool foundPath = false;
while (openSet.Count > 0)
{
State present = openSet.Dequeue();
cameFrom.Add(present);
if (present == goal)
{
foundPath = true;
intermediates = ReconstructPath(start, goal);
break;
}
else
{
List <State> adjacents = gridMap.RetrieveAdjacentStates(present);
foreach (State s in adjacents)
{
if (cameFrom.Contains(s) || !(s.unblocked))
{
continue;
}
else
{
int yWeight = Mathf.Abs(present.yCoordinate - s.yCoordinate);
int xWeight = Mathf.Abs(present.xCoordinate - s.xCoordinate);
int edgeWeight = 0;
if (xWeight < yWeight)
{
edgeWeight = 10 * (yWeight - xWeight) + 14 * xWeight;
}
else
{
edgeWeight = 10 * (xWeight - yWeight) + 14 * yWeight;
}
int tentativeCost = present.gOfN + edgeWeight;
if (!(openSet.Contains(s)) || tentativeCost < s.gOfN)
{
s.parent = present;
s.gOfN = tentativeCost;
yWeight = Mathf.Abs(s.yCoordinate - goal.yCoordinate);
xWeight = Mathf.Abs(s.xCoordinate - goal.xCoordinate);
if (xWeight < yWeight)
{
s.hOfN = 10 * (yWeight - xWeight) + 14 * xWeight;
}
else
{
s.hOfN = 10 * (xWeight - yWeight) + 14 * yWeight;
}
if (!openSet.Contains(s))
{
openSet.Enqueue(s);
}
else
{
openSet.UpdateState(s);
}
}
}
}
}
}
yield return(null);
searchHandler.FinishedPresentSearch(intermediates, foundPath);
}