Qu’est-ce que le cycle hamiltonien ?
Cycle ou circuit hamiltonien dans un graphique g est un cycle qui visite chaque sommet de g exactement une fois et revient au sommet de départ.
- Si le graphe contient un cycle hamiltonien, on l'appelle Graphique hamiltonien sinon c'est non-Hamiltonien .
- Trouver un cycle hamiltonien dans un graphique est une méthode bien connue Problème NP-complet , ce qui signifie qu’il n’existe aucun algorithme efficace connu pour le résoudre pour tous les types de graphiques. Cependant, il peut être résolu pour des types de graphiques petits ou spécifiques.
Le problème du cycle hamiltonien a des applications pratiques dans divers domaines, tels que logistique, conception de réseaux et informatique .
Qu’est-ce que le chemin hamiltonien ?
Chemin hamiltonien dans un graphique g est un chemin qui visite chaque sommet de G exactement une fois et Chemin hamiltonien il n’est pas nécessaire de revenir au sommet de départ. C'est un chemin ouvert.
- Semblable à la Cycle hamiltonien problème, trouver un Chemin hamiltonien dans un graphique général est également NP-complet et peut être un défi. Cependant, il s’agit souvent d’un problème plus simple que de trouver un cycle hamiltonien.
- Les chemins hamiltoniens ont des applications dans divers domaines, tels que trouver des itinéraires optimaux dans les réseaux de transport, la conception de circuits et la recherche sur la théorie des graphes .
Énoncé des problèmes : Étant donné un graphe non orienté, la tâche consiste à déterminer si le graphe contient ou non un cycle hamiltonien. S'il contient, imprime le chemin.
Exemple:
Recommandé : veuillez le résoudre sur PRATIQUE d'abord, avant de passer à la solution.Saisir: graph[][] = {{0, 1, 0, 1, 0},{1, 0, 1, 1, 1},{0, 1, 0, 0, 1},{1, 1, 0, 0, 1},{0, 1, 1, 1, 0}}
Graphique d'entrée[][]
circuit additionneur completSortir: {0, 1, 2, 4, 3, 0}.
Saisir: graph[][] = {{0, 1, 0, 1, 0},{1, 0, 1, 1, 1},{0, 1, 0, 0, 1},{1, 1, 0, 0, 0},{0, 1, 1, 0, 0}}
Graphique d'entrée[][]
Sortir: La solution n'existe pas
Algorithme naïf : Ce problème peut être résolu en utilisant l'idée ci-dessous :
Générez toutes les configurations possibles de sommets et imprimez une configuration qui satisfait les contraintes données. Il y aura n! (n factorielles) configurations. La complexité temporelle globale de cette approche sera donc SUR!).
Cycle hamiltonien utilisant Algorithme de retour en arrière :
Créez un tableau de chemins vide et ajoutez un sommet 0 à cela. Ajouter d'autres sommets, en commençant par le sommet 1 . Avant d'ajouter un sommet, vérifiez s'il est adjacent au sommet précédemment ajouté et s'il n'est pas déjà ajouté. Si nous trouvons un tel sommet, nous l’ajoutons dans le cadre de la solution. Si nous ne trouvons pas de sommet alors nous revenons FAUX .
Illustrations :
Découvrons le cycle hamiltonien pour le graphique suivant :
- Commencez par le nœud 0 .
- Appliquez DFS pour trouver le chemin hamiltonien.
- Lorsque le scénario de référence est atteint (c.-à-d. nombre total de nœuds traversés == V (sommet total) ) :
- Vérifiez si le nœud actuel est un voisin du nœud de départ.
- En tant que nœud 2 et nœud 0 ne sont pas voisins les uns des autres, alors revenez-en.
À partir du nœud de démarrage 0 appelant DFS
- Comme le cycle n'est pas trouvé dans le chemin {0, 3, 1, 4, 2}. Donc, revenez du nœud 2, nœud 4.
- Maintenant, explorez une autre option pour le nœud 1 (c'est-à-dire le nœud 2)
- Lorsqu'il atteint à nouveau la condition de base, vérifiez le cycle hamiltonien
- Comme le nœud 4 n'est pas le voisin du nœud 0, encore une fois le cycle n'est pas trouvé puis retourne.
- Retour du nœud 4, nœud 2, nœud 1.
- Maintenant, explorez d'autres options pour le nœud 3.
Cycle hamiltonien
- Dans le chemin hamiltonien {0,3,4,2,1,0} nous obtenons le cycle car le nœud 1 est le voisin du nœud 0.
- Alors imprimez ce chemin cyclique .
- C'est notre cycle hamiltonien.
Vous trouverez ci-dessous l'implémentation du backtracking pour trouver le cycle hamiltonien :
C++ /* C++ program for solution of Hamiltonian Cycle problem using backtracking */ #include using namespace std; // Number of vertices in the graph #define V 5 void printSolution(int path[]); /* A utility function to check if the vertex v can be added at index 'pos' in the Hamiltonian Cycle constructed so far (stored in 'path[]') */ bool isSafe(int v, bool graph[V][V], int path[], int pos) { /* Check if this vertex is an adjacent vertex of the previously added vertex. */ if (graph [path[pos - 1]][ v ] == 0) return false; /* Check if the vertex has already been included. This step can be optimized by creating an array of size V */ for (int i = 0; i < pos; i++) if (path[i] == v) return false; return true; } /* A recursive utility function to solve hamiltonian cycle problem */ bool hamCycleUtil(bool graph[V][V], int path[], int pos) { /* base case: If all vertices are included in Hamiltonian Cycle */ if (pos == V) { // And if there is an edge from the // last included vertex to the first vertex if (graph[path[pos - 1]][path[0]] == 1) return true; else return false; } // Try different vertices as a next candidate // in Hamiltonian Cycle. We don't try for 0 as // we included 0 as starting point in hamCycle() for (int v = 1; v < V; v++) { /* Check if this vertex can be added // to Hamiltonian Cycle */ if (isSafe(v, graph, path, pos)) { path[pos] = v; /* recur to construct rest of the path */ if (hamCycleUtil (graph, path, pos + 1) == true) return true; /* If adding vertex v doesn't lead to a solution, then remove it */ path[pos] = -1; } } /* If no vertex can be added to Hamiltonian Cycle constructed so far, then return false */ return false; } /* This function solves the Hamiltonian Cycle problem using Backtracking. It mainly uses hamCycleUtil() to solve the problem. It returns false if there is no Hamiltonian Cycle possible, otherwise return true and prints the path. Please note that there may be more than one solutions, this function prints one of the feasible solutions. */ bool hamCycle(bool graph[V][V]) { int *path = new int[V]; for (int i = 0; i < V; i++) path[i] = -1; /* Let us put vertex 0 as the first vertex in the path. If there is a Hamiltonian Cycle, then the path can be started from any point of the cycle as the graph is undirected */ path[0] = 0; if (hamCycleUtil(graph, path, 1) == false ) { cout << '
Solution does not exist'; return false; } printSolution(path); return true; } /* A utility function to print solution */ void printSolution(int path[]) { cout << 'Solution Exists:' ' Following is one Hamiltonian Cycle
'; for (int i = 0; i < V; i++) cout << path[i] << ' '; // Let us print the first vertex again // to show the complete cycle cout << path[0] << ' '; cout << endl; } // Driver Code int main() { /* Let us create the following graph (0)--(1)--(2) | / | | / | | / | (3)-------(4) */ bool graph1[V][V] = {{0, 1, 0, 1, 0}, {1, 0, 1, 1, 1}, {0, 1, 0, 0, 1}, {1, 1, 0, 0, 1}, {0, 1, 1, 1, 0}}; // Print the solution hamCycle(graph1); /* Let us create the following graph (0)--(1)--(2) | / | | / | | / | (3) (4) */ bool graph2[V][V] = {{0, 1, 0, 1, 0}, {1, 0, 1, 1, 1}, {0, 1, 0, 0, 1}, {1, 1, 0, 0, 0}, {0, 1, 1, 0, 0}}; // Print the solution hamCycle(graph2); return 0; } // This is code is contributed by rathbhupendra> C++ #include using namespace std; int main() { cout << 'GFG!'; return 0; }> C /* C program for solution of Hamiltonian Cycle problem using backtracking */ #include // Number of vertices in the graph #define V 5 void printSolution(int path[]); /* A utility function to check if the vertex v can be added at index 'pos' in the Hamiltonian Cycle constructed so far (stored in 'path[]') */ int isSafe(int v, int graph[V][V], int path[], int pos) { /* Check if this vertex is an adjacent vertex of the previously added vertex. */ if (graph [ path[pos-1] ][ v ] == 0) return 0; /* Check if the vertex has already been included. This step can be optimized by creating an array of size V */ for (int i = 0; i < pos; i++) if (path[i] == v) return 0; return 1; } /* A recursive utility function to solve hamiltonian cycle problem */ int hamCycleUtil(int graph[V][V], int path[], int pos) { /* base case: If all vertices are included in Hamiltonian Cycle */ if (pos == V) { // And if there is an edge from the last included vertex to the // first vertex if ( graph[ path[pos-1] ][ path[0] ] == 1 ) return 1; else return 0; } // Try different vertices as a next candidate in Hamiltonian Cycle. // We don't try for 0 as we included 0 as starting point in hamCycle() for (int v = 1; v < V; v++) { /* Check if this vertex can be added to Hamiltonian Cycle */ if (isSafe(v, graph, path, pos)) { path[pos] = v; /* recur to construct rest of the path */ if (hamCycleUtil (graph, path, pos+1) == 1) return 1; /* If adding vertex v doesn't lead to a solution, then remove it */ path[pos] = -1; } } /* If no vertex can be added to Hamiltonian Cycle constructed so far, then return false */ return 0; } /* This function solves the Hamiltonian Cycle problem using Backtracking. It mainly uses hamCycleUtil() to solve the problem. It returns false if there is no Hamiltonian Cycle possible, otherwise return true and prints the path. Please note that there may be more than one solutions, this function prints one of the feasible solutions. */ int hamCycle(int graph[V][V]) { int path[V]; for (int i = 0; i < V; i++) path[i] = -1; /* Let us put vertex 0 as the first vertex in the path. If there is a Hamiltonian Cycle, then the path can be started from any point of the cycle as the graph is undirected */ path[0] = 0; if ( hamCycleUtil(graph, path, 1) == 0 ) { printf('
Solution does not exist'); return 0; } printSolution(path); return 1; } /* A utility function to print solution */ void printSolution(int path[]) { printf ('Solution Exists:' ' Following is one Hamiltonian Cycle
'); for (int i = 0; i < V; i++) printf(' %d ', path[i]); // Let us print the first vertex again to show the complete cycle printf(' %d ', path[0]); printf('
'); } // driver program to test above function int main() { /* Let us create the following graph (0)--(1)--(2) | / | | / | | / | (3)-------(4) */ int graph1[V][V] = {{0, 1, 0, 1, 0}, {1, 0, 1, 1, 1}, {0, 1, 0, 0, 1}, {1, 1, 0, 0, 1}, {0, 1, 1, 1, 0}, }; // Print the solution hamCycle(graph1); /* Let us create the following graph (0)--(1)--(2) | / | | / | | / | (3) (4) */ int graph2[V][V] = {{0, 1, 0, 1, 0}, {1, 0, 1, 1, 1}, {0, 1, 0, 0, 1}, {1, 1, 0, 0, 0}, {0, 1, 1, 0, 0}, }; // Print the solution hamCycle(graph2); return 0; }> Java /* Java program for solution of Hamiltonian Cycle problem using backtracking */ class HamiltonianCycle { final int V = 5; int path[]; /* A utility function to check if the vertex v can be added at index 'pos'in the Hamiltonian Cycle constructed so far (stored in 'path[]') */ boolean isSafe(int v, int graph[][], int path[], int pos) { /* Check if this vertex is an adjacent vertex of the previously added vertex. */ if (graph[path[pos - 1]][v] == 0) return false; /* Check if the vertex has already been included. This step can be optimized by creating an array of size V */ for (int i = 0; i < pos; i++) if (path[i] == v) return false; return true; } /* A recursive utility function to solve hamiltonian cycle problem */ boolean hamCycleUtil(int graph[][], int path[], int pos) { /* base case: If all vertices are included in Hamiltonian Cycle */ if (pos == V) { // And if there is an edge from the last included // vertex to the first vertex if (graph[path[pos - 1]][path[0]] == 1) return true; else return false; } // Try different vertices as a next candidate in // Hamiltonian Cycle. We don't try for 0 as we // included 0 as starting point in hamCycle() for (int v = 1; v < V; v++) { /* Check if this vertex can be added to Hamiltonian Cycle */ if (isSafe(v, graph, path, pos)) { path[pos] = v; /* recur to construct rest of the path */ if (hamCycleUtil(graph, path, pos + 1) == true) return true; /* If adding vertex v doesn't lead to a solution, then remove it */ path[pos] = -1; } } /* If no vertex can be added to Hamiltonian Cycle constructed so far, then return false */ return false; } /* This function solves the Hamiltonian Cycle problem using Backtracking. It mainly uses hamCycleUtil() to solve the problem. It returns false if there is no Hamiltonian Cycle possible, otherwise return true and prints the path. Please note that there may be more than one solutions, this function prints one of the feasible solutions. */ int hamCycle(int graph[][]) { path = new int[V]; for (int i = 0; i < V; i++) path[i] = -1; /* Let us put vertex 0 as the first vertex in the path. If there is a Hamiltonian Cycle, then the path can be started from any point of the cycle as the graph is undirected */ path[0] = 0; if (hamCycleUtil(graph, path, 1) == false) { System.out.println('
Solution does not exist'); return 0; } printSolution(path); return 1; } /* A utility function to print solution */ void printSolution(int path[]) { System.out.println('Solution Exists: Following' + ' is one Hamiltonian Cycle'); for (int i = 0; i < V; i++) System.out.print(' ' + path[i] + ' '); // Let us print the first vertex again to show the // complete cycle System.out.println(' ' + path[0] + ' '); } // driver program to test above function public static void main(String args[]) { HamiltonianCycle hamiltonian = new HamiltonianCycle(); /* Let us create the following graph (0)--(1)--(2) | / | | / | | / | (3)-------(4) */ int graph1[][] = {{0, 1, 0, 1, 0}, {1, 0, 1, 1, 1}, {0, 1, 0, 0, 1}, {1, 1, 0, 0, 1}, {0, 1, 1, 1, 0}, }; // Print the solution hamiltonian.hamCycle(graph1); /* Let us create the following graph (0)--(1)--(2) | / | | / | | / | (3) (4) */ int graph2[][] = {{0, 1, 0, 1, 0}, {1, 0, 1, 1, 1}, {0, 1, 0, 0, 1}, {1, 1, 0, 0, 0}, {0, 1, 1, 0, 0}, }; // Print the solution hamiltonian.hamCycle(graph2); } } // This code is contributed by Abhishek Shankhadhar> Python # Python program for solution of # hamiltonian cycle problem class Graph(): def __init__(self, vertices): self.graph = [[0 for column in range(vertices)] for row in range(vertices)] self.V = vertices ''' Check if this vertex is an adjacent vertex of the previously added vertex and is not included in the path earlier ''' def isSafe(self, v, pos, path): # Check if current vertex and last vertex # in path are adjacent if self.graph[ path[pos-1] ][v] == 0: return False # Check if current vertex not already in path for vertex in path: if vertex == v: return False return True # A recursive utility function to solve # hamiltonian cycle problem def hamCycleUtil(self, path, pos): # base case: if all vertices are # included in the path if pos == self.V: # Last vertex must be adjacent to the # first vertex in path to make a cycle if self.graph[ path[pos-1] ][ path[0] ] == 1: return True else: return False # Try different vertices as a next candidate # in Hamiltonian Cycle. We don't try for 0 as # we included 0 as starting point in hamCycle() for v in range(1,self.V): if self.isSafe(v, pos, path) == True: path[pos] = v if self.hamCycleUtil(path, pos+1) == True: return True # Remove current vertex if it doesn't # lead to a solution path[pos] = -1 return False def hamCycle(self): path = [-1] * self.V ''' Let us put vertex 0 as the first vertex in the path. If there is a Hamiltonian Cycle, then the path can be started from any point of the cycle as the graph is undirected ''' path[0] = 0 if self.hamCycleUtil(path,1) == False: print ('Solution does not exist
') return False self.printSolution(path) return True def printSolution(self, path): print ('Solution Exists: Following', 'is one Hamiltonian Cycle') for vertex in path: print (vertex ) # Driver Code ''' Let us create the following graph (0)--(1)--(2) | / | | / | | / | (3)-------(4) ''' g1 = Graph(5) g1.graph = [ [0, 1, 0, 1, 0], [1, 0, 1, 1, 1], [0, 1, 0, 0, 1,],[1, 1, 0, 0, 1], [0, 1, 1, 1, 0], ] # Print the solution g1.hamCycle(); ''' Let us create the following graph (0)--(1)--(2) | / | | / | | / | (3) (4) ''' g2 = Graph(5) g2.graph = [ [0, 1, 0, 1, 0], [1, 0, 1, 1, 1], [0, 1, 0, 0, 1,], [1, 1, 0, 0, 0], [0, 1, 1, 0, 0], ] # Print the solution g2.hamCycle(); # This code is contributed by Divyanshu Mehta> C# // C# program for solution of Hamiltonian // Cycle problem using backtracking using System; public class HamiltonianCycle { readonly int V = 5; int []path; /* A utility function to check if the vertex v can be added at index 'pos'in the Hamiltonian Cycle constructed so far (stored in 'path[]') */ bool isSafe(int v, int [,]graph, int []path, int pos) { /* Check if this vertex is an adjacent vertex of the previously added vertex. */ if (graph[path[pos - 1], v] == 0) return false; /* Check if the vertex has already been included. This step can be optimized by creating an array of size V */ for (int i = 0; i < pos; i++) if (path[i] == v) return false; return true; } /* A recursive utility function to solve hamiltonian cycle problem */ bool hamCycleUtil(int [,]graph, int []path, int pos) { /* base case: If all vertices are included in Hamiltonian Cycle */ if (pos == V) { // And if there is an edge from the last included // vertex to the first vertex if (graph[path[pos - 1],path[0]] == 1) return true; else return false; } // Try different vertices as a next candidate in // Hamiltonian Cycle. We don't try for 0 as we // included 0 as starting point in hamCycle() for (int v = 1; v < V; v++) { /* Check if this vertex can be added to Hamiltonian Cycle */ if (isSafe(v, graph, path, pos)) { path[pos] = v; /* recur to construct rest of the path */ if (hamCycleUtil(graph, path, pos + 1) == true) return true; /* If adding vertex v doesn't lead to a solution, then remove it */ path[pos] = -1; } } /* If no vertex can be added to Hamiltonian Cycle constructed so far, then return false */ return false; } /* This function solves the Hamiltonian Cycle problem using Backtracking. It mainly uses hamCycleUtil() to solve the problem. It returns false if there is no Hamiltonian Cycle possible, otherwise return true and prints the path. Please note that there may be more than one solutions, this function prints one of the feasible solutions. */ int hamCycle(int [,]graph) { path = new int[V]; for (int i = 0; i < V; i++) path[i] = -1; /* Let us put vertex 0 as the first vertex in the path. If there is a Hamiltonian Cycle, then the path can be started from any point of the cycle as the graph is undirected */ path[0] = 0; if (hamCycleUtil(graph, path, 1) == false) { Console.WriteLine('
Solution does not exist'); return 0; } printSolution(path); return 1; } /* A utility function to print solution */ void printSolution(int []path) { Console.WriteLine('Solution Exists: Following' + ' is one Hamiltonian Cycle'); for (int i = 0; i < V; i++) Console.Write(' ' + path[i] + ' '); // Let us print the first vertex again // to show the complete cycle Console.WriteLine(' ' + path[0] + ' '); } // Driver code public static void Main(String []args) { HamiltonianCycle hamiltonian = new HamiltonianCycle(); /* Let us create the following graph (0)--(1)--(2) | / | | / | | / | (3)-------(4) */ int [,]graph1= {{0, 1, 0, 1, 0}, {1, 0, 1, 1, 1}, {0, 1, 0, 0, 1}, {1, 1, 0, 0, 1}, {0, 1, 1, 1, 0}, }; // Print the solution hamiltonian.hamCycle(graph1); /* Let us create the following graph (0)--(1)--(2) | / | | / | | / | (3) (4) */ int [,]graph2 = {{0, 1, 0, 1, 0}, {1, 0, 1, 1, 1}, {0, 1, 0, 0, 1}, {1, 1, 0, 0, 0}, {0, 1, 1, 0, 0}, }; // Print the solution hamiltonian.hamCycle(graph2); } } // This code contributed by Rajput-Ji> Javascript >
PHP // PHP program for solution of // Hamiltonian Cycle problem // using backtracking $V = 5; /* A utility function to check if the vertex v can be added at index 'pos' in the Hamiltonian Cycle constructed so far (stored in 'path[]') */ function isSafe($v, $graph, &$path, $pos) { /* Check if this vertex is an adjacent vertex of the previously added vertex. */ if ($graph[$path[$pos - 1]][$v] == 0) return false; /* Check if the vertex has already been included. This step can be optimized by creating an array of size V */ for ($i = 0; $i < $pos; $i++) if ($path[$i] == $v) return false; return true; } /* A recursive utility function to solve hamiltonian cycle problem */ function hamCycleUtil($graph, &$path, $pos) { global $V; /* base case: If all vertices are included in Hamiltonian Cycle */ if ($pos == $V) { // And if there is an edge from the // last included vertex to the first vertex if ($graph[$path[$pos - 1]][$path[0]] == 1) return true; else return false; } // Try different vertices as a next candidate in // Hamiltonian Cycle. We don't try for 0 as we // included 0 as starting point hamCycle() for ($v = 1; $v < $V; $v++) { /* Check if this vertex can be added to Hamiltonian Cycle */ if (isSafe($v, $graph, $path, $pos)) { $path[$pos] = $v; /* recur to construct rest of the path */ if (hamCycleUtil($graph, $path, $pos + 1) == true) return true; /* If adding vertex v doesn't lead to a solution, then remove it */ $path[$pos] = -1; } } /* If no vertex can be added to Hamiltonian Cycle constructed so far, then return false */ return false; } /* This function solves the Hamiltonian Cycle problem using Backtracking. It mainly uses hamCycleUtil() to solve the problem. It returns false if there is no Hamiltonian Cycle possible, otherwise return true and prints the path. Please note that there may be more than one solutions, this function prints one of the feasible solutions. */ function hamCycle($graph) { global $V; $path = array_fill(0, $V, 0); for ($i = 0; $i < $V; $i++) $path[$i] = -1; /* Let us put vertex 0 as the first vertex in the path. If there is a Hamiltonian Cycle, then the path can be started from any point of the cycle as the graph is undirected */ $path[0] = 0; if (hamCycleUtil($graph, $path, 1) == false) { echo('
Solution does not exist'); return 0; } printSolution($path); return 1; } /* A utility function to print solution */ function printSolution($path) { global $V; echo('Solution Exists: Following is '. 'one Hamiltonian Cycle
'); for ($i = 0; $i < $V; $i++) echo(' '.$path[$i].' '); // Let us print the first vertex again to show the // complete cycle echo(' '.$path[0].'
'); } // Driver Code /* Let us create the following graph (0)--(1)--(2) | / | | / | | / | (3)-------(4) */ $graph1 = array(array(0, 1, 0, 1, 0), array(1, 0, 1, 1, 1), array(0, 1, 0, 0, 1), array(1, 1, 0, 0, 1), array(0, 1, 1, 1, 0), ); // Print the solution hamCycle($graph1); /* Let us create the following graph (0)--(1)--(2) | / | | / | | / | (3) (4) */ $graph2 = array(array(0, 1, 0, 1, 0), array(1, 0, 1, 1, 1), array(0, 1, 0, 0, 1), array(1, 1, 0, 0, 0), array(0, 1, 1, 0, 0)); // Print the solution hamCycle($graph2); // This code is contributed by mits ?>> Sortir
Solution Exists: Following is one Hamiltonian Cycle 0 1 2 4 3 0 Solution does not exist>
Complexité temporelle : O(N!), où N est le nombre de sommets.
Espace auxiliaire : O(1), puisqu'aucun espace supplémentaire n'est utilisé.
Note: Le code ci-dessus imprime toujours un cycle commençant à 0 . Le point de départ n’a pas d’importance puisque le cycle peut démarrer à partir de n’importe quel point. Si vous souhaitez modifier le point de départ, vous devez apporter deux modifications au code ci-dessus.
Changer le chemin[0] = 0 ; à chemin[0] = s ; où s est ton nouveau point de départ . Changez également la boucle pour (int v = 1; v




