Add Wigglesort (#3032)

Co-authored-by: Antonia Strack <antonia.strack@thm.mni.de>
Co-authored-by: Andrii Siriak <siryaka@gmail.com>
This commit is contained in:
BranAndSceolan
2022-04-25 15:54:33 +02:00
committed by GitHub
parent 7c31c5bada
commit 99ff5163e3
2 changed files with 157 additions and 0 deletions

View File

@ -0,0 +1,82 @@
package com.thealgorithms.sorts;
import java.util.Arrays;
import static com.thealgorithms.maths.Ceil.ceil;
import static com.thealgorithms.maths.Floor.floor;
import static com.thealgorithms.searches.QuickSelect.select;
/**
* A wiggle sort implementation based on John L.s' answer in
* https://cs.stackexchange.com/questions/125372/how-to-wiggle-sort-an-array-in-linear-time-complexity
* Also have a look at: https://cs.stackexchange.com/questions/125372/how-to-wiggle-sort-an-array-in-linear-time-complexity?noredirect=1&lq=1
* Not all arrays are wiggle-sortable. This algorithm will find some obviously not wiggle-sortable arrays and throw an error,
* but there are some exceptions that won't be caught, for example [1, 2, 2].
*/
public class WiggleSort implements SortAlgorithm {
@Override
public <T extends Comparable<T>> T[] sort(T[] unsorted) {
return wiggleSort(unsorted);
}
private int mapIndex(int index, int n) {
return ((2 * index + 1) % (n | 1));
}
/**
* Modified Dutch National Flag Sort. See also: sorts/DutchNationalFlagSort
*
* @param sortThis array to sort into group "greater", "equal" and "smaller" than median
* @param median defines the groups
* @param <T> extends interface Comparable
*/
private <T extends Comparable<T>> void triColorSort(T[] sortThis, T median) {
int n = sortThis.length;
int i = 0;
int j = 0;
int k = n - 1;
while (j <= k) {
if (0 < sortThis[mapIndex(j, n)].compareTo(median)) {
SortUtils.swap(sortThis, mapIndex(j, n), mapIndex(i, n));
i++;
j++;
} else if (0 > sortThis[mapIndex(j, n)].compareTo(median)) {
SortUtils.swap(sortThis, mapIndex(j, n), mapIndex(k, n));
k--;
} else {
j++;
}
}
}
private <T extends Comparable<T>> T[] wiggleSort(T[] sortThis) {
// find the median using quickSelect (if the result isn't in the array, use the next greater value)
T median;
median = select(Arrays.<T>asList(sortThis), (int) floor(sortThis.length / 2.0));
int numMedians = 0;
for (T sortThi : sortThis) {
if (0 == sortThi.compareTo(median)) {
numMedians++;
}
}
// added condition preventing off-by-one errors for odd arrays.
// https://cs.stackexchange.com/questions/150886/how-to-find-wiggle-sortable-arrays-did-i-misunderstand-john-l-s-answer?noredirect=1&lq=1
if (sortThis.length % 2 == 1 && numMedians == ceil(sortThis.length / 2.0)) {
T smallestValue = select(Arrays.asList(sortThis), 0);
if (!(0 == smallestValue.compareTo(median))) {
throw new IllegalArgumentException("For odd Arrays if the median appears ceil(n/2) times, " +
"the median has to be the smallest values in the array.");
}
}
if (numMedians > ceil(sortThis.length / 2.0)) {
throw new IllegalArgumentException("No more than half the number of values may be the same.");
}
triColorSort(sortThis, median);
return sortThis;
}
}