Què és el cicle Hamiltonià?
Cicle o Circuit Hamiltonià en un gràfic G és un cicle que visita cada vèrtex de G exactament una vegada i torna al vèrtex inicial.
- Si el gràfic conté un cicle hamiltonià, s'anomena Gràfic hamiltonià en cas contrari ho és no hamiltonià .
- Trobar un cicle hamiltonià en un gràfic és una cosa ben coneguda Problema NP-complet , el que significa que no es coneix cap algorisme eficient per resoldre'l per a tot tipus de gràfics. Tanmateix, es pot resoldre per a tipus de gràfics petits o específics.
El problema del cicle Hamiltonià té aplicacions pràctiques en diversos camps, com ara logística, disseny de xarxes i informàtica .
Què és el camí hamiltonià?
Camí Hamiltonià en un gràfic G és un camí que visita cada vèrtex de G exactament una vegada i Camí Hamiltonià no ha de tornar al vèrtex inicial. És un camí obert.
- Similar a la Cicle Hamiltonià problema, trobar a Camí Hamiltonià en un gràfic general també ho és NP-complet i pot ser un repte. Tanmateix, sovint és un problema més fàcil que trobar un cicle hamiltonià.
- Els camins hamiltonians tenen aplicacions en diversos camps, com ara trobar rutes òptimes en xarxes de transport, disseny de circuits i investigació de teoria de grafs .
Declaració de problemes: Donat un gràfic no dirigit, la tasca és determinar si el gràfic conté un cicle hamiltonià o no. Si conté, imprimeix el camí.
Exemple:
Recomanat: si us plau, resol-ho PRÀCTICA primer, abans de passar a la solució.Entrada: gràfic[][] = {{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}}
Gràfic d'entrada[][]
canviar el nom del directori de LinuxSortida: {0, 1, 2, 4, 3, 0}.
Entrada: gràfic[][] = {{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}}
Gràfic d'entrada[][]
Sortida: La solució no existeix
Algoritme ingenu : Aquest problema es pot resoldre amb la idea següent:
Genereu totes les configuracions possibles de vèrtexs i imprimiu una configuració que compleixi les restriccions donades. Hi haurà n! configuracions (n factorials). Així, la complexitat temporal global d'aquest enfocament serà O (N!).
Utilitzant el cicle Hamiltonià Algoritme de retrocés :
Creeu una matriu de camins buida i afegiu el vèrtex 0 a ell. Afegiu altres vèrtexs, començant pel vèrtex 1 . Abans d'afegir un vèrtex, comproveu si és adjacent al vèrtex afegit anteriorment i encara no s'ha afegit. Si trobem aquest vèrtex, afegim el vèrtex com a part de la solució. Si no trobem cap vèrtex, tornem fals .
Il·lustracions:
Descobrim el cicle hamiltonià per al gràfic següent:
- Comenceu amb el node 0.
- Aplicar DFS per trobar la trajectòria hamiltoniana.
- Quan arriba el cas base (és a dir, nombre total de nodes travessats == V (vértex total) ):
- Comproveu si el node actual és un veí del node inicial.
- Com a node 2 i node 0 no són veïns els uns dels altres, així que torneu-ne.
A partir del node inicial 0 trucant a DFS
- Com que el cicle no es troba al camí {0, 3, 1, 4, 2}. Per tant, torneu des del node 2, del node 4.
- Ara, exploreu una altra opció per al node 1 (és a dir, el node 2)
- Quan arriba a la condició base, comproveu el cicle hamiltonià
- Com el node 4 no és el veí del node 0, de nou no es troba el cicle i torna.
- Retorn des del node 4, node 2, node 1.
- Ara, exploreu altres opcions per al node 3.
Cicle Hamiltonià
- En el camí hamiltonià {0,3,4,2,1,0} obtenim un cicle ja que el node 1 és el veí del node 0.
- Així que imprimiu aquest camí cíclic.
- Aquest és el nostre cicle hamiltonià.
A continuació es mostra la implementació de Backtracking per trobar el cicle Hamiltonià:
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 ?>> Sortida
Solution Exists: Following is one Hamiltonian Cycle 0 1 2 4 3 0 Solution does not exist>
Complexitat temporal: O(N!), on N és el nombre de vèrtexs.
Espai auxiliar: O(1), ja que no s'utilitza espai addicional.
Nota: El codi anterior sempre imprimeix un cicle a partir de 0 . El punt de partida no hauria d'importar, ja que el cicle es pot iniciar des de qualsevol punt. Si voleu canviar el punt de partida, hauríeu de fer dos canvis al codi anterior.
Canvia el camí[0] = 0; a camí[0] = s ; on s és el teu nou punt de partida . També canvieu el bucle per (int v = 1; v




