/*
* Reference:
* http://blog.csdn.net/linhuanmars/article/details/19659525
* comment:
* 非递归解法一般采用层序遍历(相当于图的BFS），因为如果使用其他遍历方式也需要同样的
* 复杂度O(n). 层序遍历理解上直观一些，维护到最后的level便是树的深度。
*
* julia's comment:
*
* 1. use LinkedList to implement the queue
* 2. Read Java Queue poll method /remove method difference
* 3. Let us describe the thought process here:
*
* Use two variable to help counting current level nodes left, next level num of nodes in the queue.
* 3rd variable is the level count starting from 0
* tip:
* 1. queue will have at most two levels nodes, one is current level, one is next level;
* 2. visit current level node in the queue, and meanwhile, push next level node in the queue;
* 2. If the current level node's count to zero, then, switch to next level, set as current level.
*
* 4. online judge:
*
* Time Limit Exceeded
*
*/
public static int maxDepth_Iterative(TreeNode root)
{
if (root == null)
return 0;
int level = 0;
LinkedList<TreeNode> queue = new LinkedList<TreeNode>();
queue.AddFirst(root);
int curNum = 1; //num of nodes left in current level
int nextNum = 0; //num of nodes in next level
while (queue.Count > 0 )
{
TreeNode n = queue.First(); // Java Queue Poll - find first, return null if not existed / C# Queue.Firist
curNum--;
if (n.left != null)
{
queue.AddLast(n.left);
nextNum++;
}
if (n.right != null)
{
queue.AddLast(n.right);
nextNum++;
}
if (curNum == 0) // current level is complete, need to go to next level.
{
curNum = nextNum;
nextNum = 0;
level++;
}
}
return level;
}
/*
* Reference:
* http://blog.csdn.net/linhuanmars/article/details/19659525
*
* julia's comment:
* 1. pass online judge:
* 38 / 38 test cases passed.
Status: Accepted
Runtime: 180 ms
*
*/
public static int MaxDepth(TreeNode root)
{
if (root == null)
return 0;
return Math.Max(MaxDepth(root.left), MaxDepth(root.right)) + 1;
}
/*
* reference:
* http://joycelearning.blogspot.ca/2013/10/java-leetcode-maximum-depth-of-binary.html
*
* julia's comment on August 27, 2015:
* 1. quickly try the idea using ArrayList in C#
* 2. online judge:
* 38 / 38 test cases passed.
Status: Accepted
Runtime: 164 ms
* 3. Interested and curious, why C# LinkedList failed but C# ArrayList success in online judge.
* 4. First time use C# ArrayList input argument as ArrayList as well.
* 5. chanage line 133 while use for, and make the loop easy to understand.
* 6. Describe the process:
* step 1: base case handling
* step 2: create new ArrayList container for current level,
* add root into container, set a variable depth iniitalization value 0
* step 3: Get into the loop,
* A. create a new conainter for next level,
* B. go through each node in the current level container,
* C. add left and right child into next level if not null.
* D. increment level variable;
* E. if next level count is 0, break the loop; in other words, no more level.
* F. copy the next level nodes into the current level. in other words, help to
* get back to loop again, reuse the code.
*/
public static int maxDepth_Iterative_B(TreeNode root)
{
if(root == null) return 0;
// Non-recursive, use level order traversal
ArrayList q = new ArrayList ();
q.Add(root);
int depth = 0;
for (; ; )
{
ArrayList next = new ArrayList();
foreach (TreeNode nd in q)
{
if (nd.left != null)
next.Add(nd.left);
if (nd.right != null)
next.Add(nd.right);
}
depth++;
if (next.Count == 0)
break;
q = new ArrayList(next);
}
return depth;
}
