Skip to content
/ Graph Public

Implementation of class Graph with the different path finders

Notifications You must be signed in to change notification settings

Alpus/Graph

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

18 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Graph Library

Alexander Pushin

Implemented features:

  • Class "Graph"
  • Class "GridGraph"
  • Class "GridVisualizer"
  • Dijkstra algorithm
  • AStar
  • Heuristics
    • Manhattan distance
    • Euclid distance
    • Chebyshev distance
  • XMLGridGraphParser (only for examples in "Inputs" folder)

Decription:

This library is realization of class Graph. The main part of this lib is PathFinder class that implements different algorithms.

Graph API

Graph creation

To create graph you should use constructor. It receives 3 arguments.

  1. Number of nodes.
  • Description: Number of nodes in graph.
  • Type: std::uint64_t.
  1. List of edges:
  • Default: empty.
  • Description: List of edges that should be added in graph.
  • Type: std::tuple<uint64_t, uint64_t, double> (begin, end, cost).
  • Notes:
    • Node numbering starts with 0.
    • If there is no cost, default cost equals to 0.
  1. Is directed:
  • Default: false.
  • Description: If that field is true, constructor will add all required edges as directed.
  • Type: bool.

Example:

// Creation of directed graph with:
// 3 nodes, 2 directed edges (0 - 1 with cost 1, 0 - 2 with cost 2)
// by constructor

vector<tuple<uint64_t, uint64_t, double>> edges;
edges.push_back(make_tuple(uint64_t(0), uint64_t(1), double(1)));
edges.push_back(make_tuple(uint64_t(0), uint64_t(2), double(2)));

Graph graph(3, edges, true); // there is graph with 3 nodes and 2 edges (0 - 1, 0 - 2) with directed edges

You can add edges one-by-one using methods:

  1. Graph.addDirectedEdge()
  2. Graph.addBiDirectedEdge()

They receive 3 aruments:

  1. Start node.
  • Type: Graph::Node* const.
  1. End node.
  • Type: Graph::Node* const.
  1. Edge cost.
  • Default: 0
  • Type: double.

Example:

// Creation of directed graph with:
// 3 nodes, 2 directed edges (0 - 1 with cost 1, 0 - 2 with cost 2)
// by methods

Graph graph(3);
graph.addDirectedEdge(graph[0], graph[1], 1); // edge 0 - 1 added
graph.addDirectedEdge(graph[0], graph[2], 2); // edge 0 - 2 added

Graph information

You can get graph size by method Graph.getSize(). It will return const std::uint64_t value with number of nodes in graph.

Example:

// Get graph size
Graph graph(3);
uint32_t size = graph.getSize();  // size == 3

You can get node with required number by Graph[std::uint64_t]. It will return const Graph::Node* const with node object.

Example:

// Get node by id
const Graph::Node* node = graph[0];  // node now contains graph node with id 0

You can get abstract infinite node by Graph.getInfiniteNode(). It will return const Graph::Node* with node object. This node is unreacheable from any others nodes and has id == std::numeric_limits<uint32_t>::infinity().

Example:

// Get infinite node
const Graph::Node* node = graph.getInfiniteNode();  // node id equals to numeric_limits<uint32_t>::infinity()

Node

You can get id of node using Node.getId(). That will return const std::uint32_t value with node id.

Example:

// Get node id
uint64_t id = graph[0]->getId(); // id == 0

If you need to get edges that connected with node, use Node.getEdges() It will return const std::vector<Graph::Edge>* with node edges.

// Get node edges
const std::vector<Graph::Edge>* edges = graph[0]->getEdges(); // edges contain list of edges connected with node 0 (backward 0 - 1 with cost 1, backward 0 - 2 with cost 2).

If you need only one edge, you can use Edge[std::uint64_t]. Then you get const Graph::Edge* with needed edge.

// Get node edges
const Graph::Edge* edge = (*graph[0])[0]; // edges contain list of edges connected with node 0 (backward 0 - 1 with cost 1, backward 0 - 2 with cost 2).

Edge

You can get edge cost by method Edge.getCost(). It will return const double value with edge cost.

// Get edge cost
double cost = edge->getCost(); // cost == 1

You can get edge two ends of edge by method Edge.getFrom() and Edge.getDest(). It will return const Graph::Node* nodes.

// Get ends of edge
const Graph::Node* nodeFrom  = edge->getFrom(); // nodeFrom is Node with id 0
const Graph::Node* nodeDest  = edge->getDest(); // nodeDest is Node with id 1

You can get type of edge by method Edge.getType(). It will return const Graph::EdgeType type. EdgeType is enum class that may contain {Backward, BiDirected, Forward}.

// Get type of edge
const Graph::EdgeType type  = edge->getType(); // type == Graph::EdgeType::Backward

GridGraph

That class implement grid graph and has the same API as Graph with some differences.

Creation

Constructor of GridGraph takes 3 arguments.

  1. Width.
  • Description: grid width.
  • Type: std::uint32_t.
  1. Height:
  • Description: grid height.
  • Type: std::uint32_t.
  1. Grid:
  • Description: two-dimensional grid with costs of nodes.
  • Type: const vector<vector<double>>* const.

Example:

XMLGridGraphParser data("/home/alpus/Work/Course_work/Implementation/Inputs/DragonAge-Starcraft-somemaps/Starcraft_movingai.com_/Cauldron/3796273.xml");
GridGraph graph(data.getWidth(), data.getHeight(), data.getGrid()); // There is grid graph with values from file

In that example you noticed class XMLGridGraphParser. It is help class that can parse xml files from folder "Inputs". It just takes path to file and can return data by getters discribed below:

const vector<vector<double>>* getGrid() const;
const int32_t getWidth() const;
const int32_t getHeight() const;
const int32_t getStartWidth() const;
const int32_t getStartHeight() const;
const int32_t getEndWidth() const;
const int32_t getEndHeight() const;

GridGraph information

All methods the same with methods from Graph.

Cell

Cell is replacement of Node from graph. It has one additional method const Coord getCoord() const that returns Coord of that cell. Coord is just struct with height and width fields, method bool isValid(const GridGraph* const graph) that returns true value if cell in placed in grid, and GridGraph::Cell::Coord operator+(const GridGraph::Cell::Coord) const that makes sum of the same coordinates.

Heuristic

That class contains static functions with implementations of different heuristic functions:

static const double manhattanDistance(const GridGraph::Cell *const from, const GridGraph::Cell *const to); // Манхеттенское расстояние

static const double euclidDistance(const GridGraph::Cell *const from, const GridGraph::Cell *const to); // Евклидово расстояние

static const double chebyshevDistance(const GridGraph::Cell *const from, const GridGraph::Cell *const to); // Расстояние Чебышева

PathFinder

To use some path finder you need to call constructor of needed class.

Allowed classes:

DijkstraSearch:

  • Constructor: DijkstraSearch(Graph*).
  • No negative loops
  • Min cost path.
  • From one to all.
  • O(E log(N))

AStar:

  • Constructor: AStar(Gridraph*, heuristic_function). As a second argument you can use on of the function from Heuristic class. For example Heuristic::manhattanDistance;
  • No negative loops
  • Min cost path.
  • From one to all.
  • O(E log(N))
  • Only for graphs with heuristics

Then you need to call pathFinder.findPath() that received 2 arguments:

  1. Begin node
    • Type: const Graph::Node*
  2. Goal node
  • Type: const Graph::Node*
  • Note: You are available to send Graph.getInfiniteNode() if you want to find path to all nodes and algorithm support it.
// Find path
DijkstraSearch pathFinder(&graph);
pathFinder.findPath(graph[0], graph.getInfiniteNode()); // path from node 0 to all other nodes is founded

After that you can get cost of path to node with some id by method pathFinder.getPathCost(std::uint64_t). It will return const double* const cost of path that algorithm founded (and std::numeric_limits<double>::infinity()* if algorithm did not achieve this node).

// Get path cost
const double* cost = pathFinder.getPathCost(graph[1]); // cost == 1

And you can get order of nodes that algorithm found by method pathFinder.getFullPath(). It will return const vector<Graph::Node*> list with reached nodes in path order.

// Get path order
const vector<const Graph::Node*> path = pathFinder.getFullPath(graph[1]); // path contain nodes with id (0, 1).

GridVisualizer

That class can draw GridGraph and pathes on it using OpenCV library.

Constructor takes only 1 argument GridGraph *graph.

  1. Graph.
  • Description: grid graph object.
  • Type: GridGraph *graph.

Example:

GridVisualizer map(&graph); // There is object of visualizer

If you want to draw path that find some pathFinder, then use method 'addPath(path)'. It takes 1 argument. Constructor takes only 1 argument GridGraph *graph.

  1. Path.
  • Description: path as pathfinder returns it.
  • Type: const vector<const Graph::Node*>* const path.

Example:

const vector<const Graph::Node*> path = pathFinder.getFullPath(endCell);
map.addPath(&path); // Now visualizer contain path

The main method is drawImage(). It shows image with graph and added paths.

Example:

map.drawImage(); // You will see img on your screen

About

Implementation of class Graph with the different path finders

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages