|
| 1 | +/** |
| 2 | + * @file |
| 3 | + * @brief Kth largest element in linear time [Wikipedia: Selection Algorithm](https://en.wikipedia.org/wiki/Selection_algorithm)[Median of Medians](https://en.wikipedia.org/wiki/Median_of_medians) |
| 4 | + * @details |
| 5 | + * Quick Select is a linear-time algorithm for finding the kth largest element in an unsorted array. |
| 6 | + * It uses the median-of-medians algorithm to guarantee a good pivot selection, achieving O(n) |
| 7 | + * average time complexity and avoiding the O(n^2) worst-case of naive quickselect. |
| 8 | + * @author [Nothinormuch](https://github.com/Nothinormuch/) |
| 9 | + */ |
| 10 | + |
| 11 | +#include<stdio.h> |
| 12 | +#include<stdlib.h> |
| 13 | +#include<time.h> |
| 14 | +#include<assert.h> |
| 15 | + |
| 16 | +/** |
| 17 | + * @brief Prints a portion of an array |
| 18 | + */ |
| 19 | +void print_arr(int * arr, int start, int stop){ |
| 20 | + printf("[%d",arr[start]); |
| 21 | + for(int i = start+1; i < stop+1; i ++){ |
| 22 | + printf(",%d",arr[i]); |
| 23 | + } |
| 24 | + printf("]"); |
| 25 | +} |
| 26 | + |
| 27 | +/** |
| 28 | + * @brief Swaps two elements in an array |
| 29 | + * @param arr array pointer |
| 30 | + * @param i Index of first element |
| 31 | + * @param j Index of second element |
| 32 | + */ |
| 33 | +void swap(int * arr, int i, int j){ |
| 34 | + int tmp = arr[i]; |
| 35 | + arr[i] = arr[j]; |
| 36 | + arr[j] = tmp; |
| 37 | +} |
| 38 | + |
| 39 | +int partition(int * arr, int start, int stop, int pivot_value); |
| 40 | +int median_of_medians(int * arr, int start, int stop); |
| 41 | +int median_of_medians_helper(int * arr, int start, int stop); |
| 42 | + |
| 43 | +/** |
| 44 | + * @brief Partitions array so elements greater than pivot are on the left |
| 45 | + * @details |
| 46 | + * Rearranges elements so that all elements greater than the pivot_value |
| 47 | + * are moved to the left side, smaller elements remain on the right. |
| 48 | + * The pivot is placed at the boundary between these two groups. |
| 49 | + * @param arr Pointer to the array |
| 50 | + * @param start Starting index of the partition range |
| 51 | + * @param stop Ending index of the partition range |
| 52 | + * @param pivot_value The value to partition around |
| 53 | + * @returns The final position of the pivot element |
| 54 | + */ |
| 55 | +int partition(int * arr, int start, int stop, int pivot_value){ |
| 56 | + int i = start; // boundary pointer between larger and smaller elements |
| 57 | + |
| 58 | + // Move all elements greater than pivot to the left |
| 59 | + for (int j = start; j <= stop; j++){ |
| 60 | + if (arr[j] > pivot_value){ |
| 61 | + swap(arr, i, j); |
| 62 | + i++; |
| 63 | + } |
| 64 | + } |
| 65 | + |
| 66 | + // Find and place the pivot at position i |
| 67 | + for (int j = i; j <= stop; j++){ |
| 68 | + if (arr[j] == pivot_value){ |
| 69 | + swap(arr, i, j); |
| 70 | + break; |
| 71 | + } |
| 72 | + } |
| 73 | + return i; |
| 74 | +} |
| 75 | + |
| 76 | +/** |
| 77 | + * @brief Finds a good pivot value using the median-of-medians algorithm |
| 78 | + * @details |
| 79 | + * Uses a divide-and-conquer strategy: divides the array into groups of 5, |
| 80 | + * finds the median of each group, then recursively finds the median of those medians. |
| 81 | + * This guarantees O(n) linear time complexity regardless of input distribution. |
| 82 | + * @param arr Pointer to the array |
| 83 | + * @param start Starting index of the range |
| 84 | + * @param stop Ending index of the range |
| 85 | + * @returns The median value (suitable for use as a pivot) |
| 86 | + */ |
| 87 | +int median_of_medians(int * arr, int start, int stop){ |
| 88 | + int len = stop - start + 1; |
| 89 | + |
| 90 | + // Base case: small arrays just get sorted and return middle element |
| 91 | + if (len <= 5){ |
| 92 | + for (int i = start; i <= stop; i++){ |
| 93 | + for (int j = start; j < stop - (i - start); j++){ |
| 94 | + if (arr[j] > arr[j + 1]){ |
| 95 | + swap(arr, j, j + 1); |
| 96 | + } |
| 97 | + } |
| 98 | + } |
| 99 | + return arr[start + len / 2]; |
| 100 | + } |
| 101 | + |
| 102 | + // Divide into groups of 5 |
| 103 | + int num_groups = (len + 4) / 5; // ceiling division |
| 104 | + int * medians = (int *)malloc(sizeof(int) * num_groups); |
| 105 | + |
| 106 | + for (int i = 0; i < num_groups; i++){ |
| 107 | + int sub_start = start + i * 5; |
| 108 | + // Last group may have fewer than 5 elements |
| 109 | + int sub_stop = (sub_start + 4 > stop) ? stop : sub_start + 4; |
| 110 | + int sub_len = sub_stop - sub_start + 1; |
| 111 | + |
| 112 | + // Sort this group |
| 113 | + for (int j = sub_start; j <= sub_stop; j++){ |
| 114 | + for (int k = sub_start; k < sub_stop - (j - sub_start); k++){ |
| 115 | + if (arr[k] > arr[k + 1]){ |
| 116 | + swap(arr, k, k + 1); |
| 117 | + } |
| 118 | + } |
| 119 | + } |
| 120 | + medians[i] = arr[sub_start + sub_len / 2]; // store median of this group |
| 121 | + } |
| 122 | + |
| 123 | + // Recursively find the median of all medians |
| 124 | + int result = median_of_medians_helper(medians, 0, num_groups - 1); |
| 125 | + free(medians); |
| 126 | + return result; |
| 127 | +} |
| 128 | + |
| 129 | +/** |
| 130 | + * @brief Recursive helper for median-of-medians algorithm |
| 131 | + * @details |
| 132 | + * This function implements the same median-of-medians logic as the parent function. |
| 133 | + * It's separated as a helper to manage recursion properly without modifying the original array unexpectedly. |
| 134 | + * @param arr Pointer to the array |
| 135 | + * @param start Starting index of the range |
| 136 | + * @param stop Ending index of the range |
| 137 | + * @returns The median value |
| 138 | + */ |
| 139 | +int median_of_medians_helper(int * arr, int start, int stop){ |
| 140 | + int len = stop - start + 1; |
| 141 | + if (len <= 5){ |
| 142 | + for (int i = start; i <= stop; i++){ |
| 143 | + for (int j = start; j < stop - (i - start); j++){ |
| 144 | + if (arr[j] > arr[j + 1]){ |
| 145 | + swap(arr, j, j + 1); |
| 146 | + } |
| 147 | + } |
| 148 | + } |
| 149 | + return arr[start + len / 2]; |
| 150 | + } |
| 151 | + |
| 152 | + int num_groups = (len + 4) / 5; |
| 153 | + int * medians = (int *)malloc(sizeof(int) * num_groups); |
| 154 | + |
| 155 | + for (int i = 0; i < num_groups; i++){ |
| 156 | + int sub_start = start + i * 5; |
| 157 | + int sub_stop = (sub_start + 4 > stop) ? stop : sub_start + 4; |
| 158 | + int sub_len = sub_stop - sub_start + 1; |
| 159 | + |
| 160 | + for (int j = sub_start; j <= sub_stop; j++){ |
| 161 | + for (int k = sub_start; k < sub_stop - (j - sub_start); k++){ |
| 162 | + if (arr[k] > arr[k + 1]){ |
| 163 | + swap(arr, k, k + 1); |
| 164 | + } |
| 165 | + } |
| 166 | + } |
| 167 | + medians[i] = arr[sub_start + sub_len / 2]; |
| 168 | + } |
| 169 | + |
| 170 | + int result = median_of_medians_helper(medians, 0, num_groups - 1); |
| 171 | + free(medians); |
| 172 | + return result; |
| 173 | +} |
| 174 | + |
| 175 | +/** |
| 176 | + * @brief Finds the kth largest element in an array |
| 177 | + * @details |
| 178 | + * Uses the median-of-medians algorithm to find a good pivot, then partitions |
| 179 | + * the array and recursively searches the appropriate half. The pivot selection |
| 180 | + * guarantees O(n) time complexity in all cases (best, average, and worst). |
| 181 | + * k is 1-based: k=1 returns the largest, k=2 returns the 2nd largest, etc. |
| 182 | + * @param arr Pointer to the array |
| 183 | + * @param k The rank to find (1 = largest, 2 = 2nd largest, ..., n = smallest) |
| 184 | + * @param start Starting index of the search range |
| 185 | + * @param stop Ending index of the search range |
| 186 | + * @returns The kth largest element, or -1 if the range is invalid |
| 187 | + */ |
| 188 | +int kth_largest(int * arr, int k, int start, int stop){ |
| 189 | + if (start > stop) return -1; |
| 190 | + |
| 191 | + // Use median-of-medians to pick a good pivot |
| 192 | + int pivot_value = median_of_medians(arr, start, stop); |
| 193 | + |
| 194 | + // Partition: larger elements go left, smaller go right of partition |
| 195 | + int pivot_index = partition(arr, start, stop, pivot_value); |
| 196 | + // Rank = how many elements are >= pivot_value |
| 197 | + int rank = pivot_index - start + 1; |
| 198 | + |
| 199 | + // Check if we found the answer |
| 200 | + if (rank == k){ |
| 201 | + return pivot_value; |
| 202 | + } |
| 203 | + // Kth largest is in left half (larger elements) |
| 204 | + else if (rank > k){ |
| 205 | + return kth_largest(arr, k, start, pivot_index - 1); |
| 206 | + } |
| 207 | + // Kth largest is in right half (smaller elements), adjust k by how many are seen |
| 208 | + else{ |
| 209 | + return kth_largest(arr, k - rank, pivot_index + 1, stop); |
| 210 | + } |
| 211 | +} |
| 212 | + |
| 213 | + |
| 214 | +/** |
| 215 | + * @brief Test cases |
| 216 | + */ |
| 217 | +static void test() { |
| 218 | + // Test 1: Simple unsorted array, find 3rd largest (17) |
| 219 | + int arr1[] = {7, 1, 15, 3, 19, 11, 5, 18, 2, 14, 9, 4, 16, 8, 12, 6, 17, 10, 13}; |
| 220 | + int result1 = kth_largest(arr1, 3, 0, 18); |
| 221 | + assert(result1 == 17); |
| 222 | + printf("Test 1 passed: 3rd largest in unsorted array is 17\n"); |
| 223 | + |
| 224 | + // Test 2: Find the largest element (k=1) |
| 225 | + int arr2[] = {5, 2, 8, 1, 9, 3}; |
| 226 | + int result2 = kth_largest(arr2, 1, 0, 5); |
| 227 | + assert(result2 == 9); |
| 228 | + printf("Test 2 passed: 1st largest (max) is 9\n"); |
| 229 | + |
| 230 | + // Test 3: Find the smallest element (k=n) |
| 231 | + int arr3[] = {5, 2, 8, 1, 9, 3}; |
| 232 | + int result3 = kth_largest(arr3, 6, 0, 5); |
| 233 | + assert(result3 == 1); |
| 234 | + printf("Test 3 passed: 6th largest (min) in 6-element array is 1\n"); |
| 235 | + |
| 236 | + // Test 4: Single element array |
| 237 | + int arr4[] = {42}; |
| 238 | + int result4 = kth_largest(arr4, 1, 0, 0); |
| 239 | + assert(result4 == 42); |
| 240 | + printf("Test 4 passed: 1st largest in single-element array is 42\n"); |
| 241 | + |
| 242 | + // Test 5: Two elements, find largest |
| 243 | + int arr5[] = {10, 20}; |
| 244 | + int result5 = kth_largest(arr5, 1, 0, 1); |
| 245 | + assert(result5 == 20); |
| 246 | + printf("Test 5 passed: 1st largest in two-element array is 20\n"); |
| 247 | + |
| 248 | + // Test 6: Two elements, find smallest |
| 249 | + int arr6[] = {10, 20}; |
| 250 | + int result6 = kth_largest(arr6, 2, 0, 1); |
| 251 | + assert(result6 == 10); |
| 252 | + printf("Test 6 passed: 2nd largest in two-element array is 10\n"); |
| 253 | + |
| 254 | + // Test 7: Array with duplicates, find 4th largest |
| 255 | + int arr7[] = {5, 3, 5, 2, 5, 1, 5}; |
| 256 | + int result7 = kth_largest(arr7, 4, 0, 6); |
| 257 | + assert(result7 == 5); // sorted desc: [5,5,5,5,3,2,1], 4th is 5 |
| 258 | + printf("Test 7 passed: 4th largest with duplicates is 5\n"); |
| 259 | + |
| 260 | + // Test 8: Already sorted (descending), find middle |
| 261 | + int arr8[] = {10, 9, 8, 7, 6, 5, 4, 3, 2, 1}; |
| 262 | + int result8 = kth_largest(arr8, 5, 0, 9); |
| 263 | + assert(result8 == 6); |
| 264 | + printf("Test 8 passed: 5th largest in sorted descending array is 6\n"); |
| 265 | + |
| 266 | + // Test 9: Already sorted (ascending), find 3rd largest |
| 267 | + int arr9[] = {1, 2, 3, 4, 5, 6, 7, 8, 9, 10}; |
| 268 | + int result9 = kth_largest(arr9, 3, 0, 9); |
| 269 | + assert(result9 == 8); |
| 270 | + printf("Test 9 passed: 3rd largest in sorted ascending array is 8\n"); |
| 271 | + |
| 272 | + // Test 10: Larger array with random values |
| 273 | + int arr10[] = {45, 23, 78, 12, 89, 34, 56, 90, 67, 21, 98, 54, 32, 11, 88, 77, 42}; |
| 274 | + int result10 = kth_largest(arr10, 5, 0, 16); |
| 275 | + assert(result10 == 78); // 5th largest: 98, 90, 89, 88, 78 |
| 276 | + printf("Test 10 passed: 5th largest in random array is 78\n"); |
| 277 | + |
| 278 | + printf("\nAll tests have successfully passed!\n"); |
| 279 | +} |
| 280 | + |
| 281 | +// Main Function |
| 282 | +int main(){ |
| 283 | + test(); |
| 284 | + return 0; |
| 285 | +} |
0 commit comments