Java算法之堆排序代码示例

public class HeapSort {
  private int[] numbers;
  private int length;
  public HeapSort(int[] numbers) {
    this.numbers = numbers;
    this.length = numbers.length;
  }
  /**
   * 调整二叉树
   * 如果父节点编号为x,那么左子节点的编号是2x,右子节点的编号是2x+1
   * 节点编号从1开始,对应数组中的索引是编号-1
   * @param nodeId 节点编号,从1开始
   */
  public void adjust(int nodeId) {
    int swapId;
    int flag = 0; //是否需要继续向下调整
    while(nodeId * 2 <= this.length && flag == 0) {
      //首先判断它和左子节点的关系,并用swapId记录值较小的节点编号(最大堆是记录较大的)
      int index = nodeId - 1; //节点对应数组中数字的索引
      int leftChild = nodeId * 2 - 1; //左子节点对应数组中数字的索引
      int rightChild = nodeId * 2; //右子节点对应数组中数字的索引
      if(numbers[index] < numbers[leftChild]) {
        swapId = nodeId * 2;
      } else {
        swapId = nodeId;
      }
      //如果有右子节点,再与右子节点比较
      if(nodeId * 2 + 1 <= this.length) {
        if(numbers[swapId - 1] < numbers[rightChild])
          swapId = nodeId * 2 + 1;
      }
      //如果最小的节点编号不是自己,说明子节点中有比父节点更小的
      if(swapId != nodeId) {
        swap(swapId,nodeId);
        nodeId = swapId;
      } else {
        flag = 1;
      }
    }
  }
  /**
   * 交换两个节点的值
   * @param nodeId1
   * @param nodeId2
   */
  public void swap(int nodeId1,int nodeId2) {
    int t = numbers[nodeId1 - 1];
    numbers[nodeId1 - 1] = numbers[nodeId2 - 1];
    numbers[nodeId2 - 1] = t;
  }
  /**
   * 创建最大堆
   */
  public void createMaxHeap() {
    //从最后一个非叶节点到第一个节点依次向上调整
    for(int i = this.length / 2; i >= 1; i--) {
      adjust(i);
    }
  }
  public static void main(String[] args) {
    int[] numbers = new int[] { 4,3,6,2,7,1,5 };
    for(int x = 0; x < numbers.length; x++) {
      System.out.print(numbers[x] + " ");
    }
    System.out.println();
    HeapSort heap = new HeapSort(numbers);
    heap.createMaxHeap();
  }
}

public void sort() {
  while(this.length > 1) {
    swap(1,this.length);
    this.length--;
    adjust(1);
  }
  for(int x = 0; x < numbers.length; x++) {
    System.out.print(numbers[x] + " ");
  }
}