diff --git a/DIRECTORY.md b/DIRECTORY.md index e8b2604d0..7c2e6956b 100644 --- a/DIRECTORY.md +++ b/DIRECTORY.md @@ -46,6 +46,8 @@ * [Main Cll](https://github.com/TheAlgorithms/C-Plus-Plus/blob/master/data_structures/cll/main_cll.cpp) * [Disjoint Set](https://github.com/TheAlgorithms/C-Plus-Plus/blob/master/data_structures/disjoint_set.cpp) * [Doubly Linked List](https://github.com/TheAlgorithms/C-Plus-Plus/blob/master/data_structures/doubly_linked_list.cpp) + * [Dsu Path Compression](https://github.com/TheAlgorithms/C-Plus-Plus/blob/master/data_structures/dsu_path_compression.cpp) + * [Dsu Union Rank](https://github.com/TheAlgorithms/C-Plus-Plus/blob/master/data_structures/dsu_union_rank.cpp) * [Linked List](https://github.com/TheAlgorithms/C-Plus-Plus/blob/master/data_structures/linked_list.cpp) * [Linkedlist Implentation Usingarray](https://github.com/TheAlgorithms/C-Plus-Plus/blob/master/data_structures/linkedlist_implentation_usingarray.cpp) * [List Array](https://github.com/TheAlgorithms/C-Plus-Plus/blob/master/data_structures/list_array.cpp) diff --git a/data_structures/dsu_path_compression.cpp b/data_structures/dsu_path_compression.cpp new file mode 100644 index 000000000..a5c0aec33 --- /dev/null +++ b/data_structures/dsu_path_compression.cpp @@ -0,0 +1,213 @@ +/** + * @file + * @brief [DSU (Disjoint + * sets)](https://en.wikipedia.org/wiki/Disjoint-set-data_structure) + * @details + * It is a very powerful data structure that keeps track of different + * clusters(sets) of elements, these sets are disjoint(doesnot have a common + * element). Disjoint sets uses cases : for finding connected components in a + * graph, used in Kruskal's algorithm for finding Minimum Spanning tree. + * Operations that can be performed: + * 1) UnionSet(i,j): add(element i and j to the set) + * 2) findSet(i): returns the representative of the set to which i belogngs to. + * 3) get_max(i),get_min(i) : returns the maximum and minimum + * Below is the class-based approach which uses the heuristic of path + * compression. Using path compression in findSet(i),we are able to get to the + * representative of i in O(1) time. + * @author [AayushVyasKIIT](https://github.com/AayushVyasKIIT) + * @see dsu_union_rank.cpp + */ + +#include /// for assert +#include /// for IO operations +#include /// for std::vector + +using std::cout; +using std::endl; +using std::vector; + +/** + * @brief Disjoint sets union data structure, class based representation. + * @param n number of elements + */ +class dsu { + private: + vector p; ///< keeps track of the parent of ith element + vector depth; ///< tracks the depth(rank) of i in the tree + vector setSize; ///< size of each chunk(set) + vector maxElement; ///< maximum of each set to which i belongs to + vector minElement; ///< minimum of each set to which i belongs to + public: + /** + * @brief contructor for initialising all data members. + * @param n number of elements + */ + explicit dsu(uint64_t n) { + p.assign(n, 0); + /// initially, all of them are their own parents + for (uint64_t i = 0; i < n; i++) { + p[i] = i; + } + /// initially all have depth are equals to zero + depth.assign(n, 0); + maxElement.assign(n, 0); + minElement.assign(n, 0); + for (uint64_t i = 0; i < n; i++) { + depth[i] = 0; + maxElement[i] = i; + minElement[i] = i; + } + setSize.assign(n, 0); + /// initially set size will be equals to one + for (uint64_t i = 0; i < n; i++) { + setSize[i] = 1; + } + } + + /** + * @brief Method to find the representative of the set to which i belongs + * to, T(n) = O(1) + * @param i element of some set + * @returns representative of the set to which i belongs to. + */ + uint64_t findSet(uint64_t i) { + /// using path compression + if (p[i] == i) { + return i; + } + return (p[i] = findSet(p[i])); + } + /** + * @brief Method that combines two disjoint sets to which i and j belongs to + * and make a single set having a common representative. + * @param i element of some set + * @param j element of some set + * @returns void + */ + void UnionSet(uint64_t i, uint64_t j) { + /// check if both belongs to the same set or not + if (isSame(i, j)) { + return; + } + + // we find the representative of the i and j + uint64_t x = findSet(i); + uint64_t y = findSet(j); + + /// always keeping the min as x + /// shallow tree + if (depth[x] > depth[y]) { + std::swap(x, y); + } + /// making the shallower root's parent the deeper root + p[x] = y; + + /// if same depth, then increase one's depth + if (depth[x] == depth[y]) { + depth[y]++; + } + /// total size of the resultant set + setSize[y] += setSize[x]; + /// changing the maximum elements + maxElement[y] = std::max(maxElement[x], maxElement[y]); + minElement[y] = std::min(minElement[x], minElement[y]); + } + /** + * @brief A utility function which check whether i and j belongs to + * same set or not + * @param i element of some set + * @param j element of some set + * @returns `true` if element `i` and `j` ARE in the same set + * @returns `false` if element `i` and `j` are NOT in same set + */ + bool isSame(uint64_t i, uint64_t j) { + if (findSet(i) == findSet(j)) { + return true; + } + return false; + } + /** + * @brief prints the minimum, maximum and size of the set to which i belongs + * to + * @param i element of some set + * @returns void + */ + vector get(uint64_t i) { + vector ans; + ans.push_back(get_min(i)); + ans.push_back(get_max(i)); + ans.push_back(size(i)); + return ans; + } + /** + * @brief A utility function that returns the size of the set to which i + * belongs to + * @param i element of some set + * @returns size of the set to which i belongs to + */ + uint64_t size(uint64_t i) { return setSize[findSet(i)]; } + /** + * @brief A utility function that returns the max element of the set to + * which i belongs to + * @param i element of some set + * @returns maximum of the set to which i belongs to + */ + uint64_t get_max(uint64_t i) { return maxElement[findSet(i)]; } + /** + * @brief A utility function that returns the min element of the set to + * which i belongs to + * @param i element of some set + * @returns minimum of the set to which i belongs to + */ + uint64_t get_min(uint64_t i) { return minElement[findSet(i)]; } +}; + +/** + * @brief Self-test implementations, 1st test + * @returns void + */ +static void test1() { + // the minimum, maximum, and size of the set + uint64_t n = 10; ///< number of items + dsu d(n + 1); ///< object of class disjoint sets + // set 1 + d.UnionSet(1, 2); // performs union operation on 1 and 2 + d.UnionSet(1, 4); // performs union operation on 1 and 4 + vector ans = {1, 4, 3}; + for (uint64_t i = 0; i < ans.size(); i++) { + assert(d.get(4).at(i) == ans[i]); // makes sure algorithm works fine + } + cout << "1st test passed!" << endl; +} +/** + * @brief Self-implementations, 2nd test + * @returns void + */ +static void test2() { + // the minimum, maximum, and size of the set + uint64_t n = 10; ///< number of items + dsu d(n + 1); ///< object of class disjoint sets + // set 1 + d.UnionSet(3, 5); + d.UnionSet(5, 6); + d.UnionSet(5, 7); + vector ans = {3, 7, 4}; + for (uint64_t i = 0; i < ans.size(); i++) { + assert(d.get(3).at(i) == ans[i]); // makes sure algorithm works fine + } + cout << "2nd test passed!" << endl; +} + +/** + * @brief Main function + * @returns 0 on exit + * */ +int main() { + uint64_t n = 10; ///< number of items + dsu d(n + 1); ///< object of class disjoint sets + + test1(); // run 1st test case + test2(); // run 2nd test case + + return 0; +} diff --git a/data_structures/dsu_union_rank.cpp b/data_structures/dsu_union_rank.cpp new file mode 100644 index 000000000..8936d6d69 --- /dev/null +++ b/data_structures/dsu_union_rank.cpp @@ -0,0 +1,187 @@ +/** + * @file + * @brief [DSU (Disjoint + * sets)](https://en.wikipedia.org/wiki/Disjoint-set-data_structure) + * @details + * dsu : It is a very powerful data structure which keeps track of different + * clusters(sets) of elements, these sets are disjoint(doesnot have a common + * element). Disjoint sets uses cases : for finding connected components in a + * graph, used in Kruskal's algorithm for finding Minimum Spanning tree. + * Operations that can be performed: + * 1) UnionSet(i,j): add(element i and j to the set) + * 2) findSet(i): returns the representative of the set to which i belogngs to. + * 3) getParents(i): prints the parent of i and so on and so forth. + * Below is the class-based approach which uses the heuristic of union-ranks. + * Using union-rank in findSet(i),we are able to get to the representative of i + * in slightly delayed O(logN) time but it allows us to keep tracks of the + * parent of i. + * @author [AayushVyasKIIT](https://github.com/AayushVyasKIIT) + * @see dsu_path_compression.cpp + */ + +#include /// for assert +#include /// for IO operations +#include /// for std::vector + +using std::cout; +using std::endl; +using std::vector; + +/** + * @brief Disjoint sets union data structure, class based representation. + * @param n number of elements + */ +class dsu { + private: + vector p; ///< keeps track of the parent of ith element + vector depth; ///< tracks the depth(rank) of i in the tree + vector setSize; ///< size of each chunk(set) + public: + /** + * @brief constructor for initialising all data members + * @param n number of elements + */ + explicit dsu(uint64_t n) { + p.assign(n, 0); + /// initially all of them are their own parents + depth.assign(n, 0); + setSize.assign(n, 0); + for (uint64_t i = 0; i < n; i++) { + p[i] = i; + depth[i] = 0; + setSize[i] = 1; + } + } + /** + * @brief Method to find the representative of the set to which i belongs + * to, T(n) = O(logN) + * @param i element of some set + * @returns representative of the set to which i belongs to + */ + uint64_t findSet(uint64_t i) { + /// using union-rank + while (i != p[i]) { + i = p[i]; + } + return i; + } + /** + * @brief Method that combines two disjoint sets to which i and j belongs to + * and make a single set having a common representative. + * @param i element of some set + * @param j element of some set + * @returns void + */ + void unionSet(uint64_t i, uint64_t j) { + /// checks if both belongs to same set or not + if (isSame(i, j)) { + return; + } + /// we find representative of the i and j + uint64_t x = findSet(i); + uint64_t y = findSet(j); + + /// always keeping the min as x + /// in order to create a shallow tree + if (depth[x] > depth[y]) { + std::swap(x, y); + } + /// making the shallower tree, root parent of the deeper root + p[x] = y; + + /// if same depth, then increase one's depth + if (depth[x] == depth[y]) { + depth[y]++; + } + /// total size of the resultant set + setSize[y] += setSize[x]; + } + /** + * @brief A utility function which check whether i and j belongs to same set + * or not + * @param i element of some set + * @param j element of some set + * @returns `true` if element i and j are in same set + * @returns `false` if element i and j are not in same set + */ + bool isSame(uint64_t i, uint64_t j) { + if (findSet(i) == findSet(j)) { + return true; + } + return false; + } + /** + * @brief Method to print all the parents of i, or the path from i to + * representative. + * @param i element of some set + * @returns void + */ + vector getParents(uint64_t i) { + vector ans; + while (p[i] != i) { + ans.push_back(i); + i = p[i]; + } + ans.push_back(i); + return ans; + } +}; +/** + * @brief Self-implementations, 1st test + * @returns void + */ +static void test1() { + /* checks the parents in the resultant structures */ + uint64_t n = 10; ///< number of elements + dsu d(n + 1); ///< object of class disjoint sets + d.unionSet(2, 1); ///< performs union operation on 1 and 2 + d.unionSet(1, 4); + d.unionSet(8, 1); + d.unionSet(3, 5); + d.unionSet(5, 6); + d.unionSet(5, 7); + d.unionSet(9, 10); + d.unionSet(2, 10); + // keeping track of the changes using parent pointers + vector ans = {7, 5}; + for (uint64_t i = 0; i < ans.size(); i++) { + assert(d.getParents(7).at(i) == + ans[i]); // makes sure algorithm works fine + } + cout << "1st test passed!" << endl; +} +/** + * @brief Self-implementations, 2nd test + * @returns void + */ +static void test2() { + // checks the parents in the resultant structures + uint64_t n = 10; ///< number of elements + dsu d(n + 1); ///< object of class disjoint sets + d.unionSet(2, 1); /// performs union operation on 1 and 2 + d.unionSet(1, 4); + d.unionSet(8, 1); + d.unionSet(3, 5); + d.unionSet(5, 6); + d.unionSet(5, 7); + d.unionSet(9, 10); + d.unionSet(2, 10); + + /// keeping track of the changes using parent pointers + vector ans = {2, 1, 10}; + for (uint64_t i = 0; i < ans.size(); i++) { + assert(d.getParents(2).at(i) == + ans[i]); /// makes sure algorithm works fine + } + cout << "2nd test passed!" << endl; +} +/** + * @brief Main function + * @returns 0 on exit + */ +int main() { + test1(); // run 1st test case + test2(); // run 2nd test case + + return 0; +}