I came into an article which talking about the LCA algorithms, the code is simple http://leetcode.com/2011/07/lowest-common-ancestor-of-a-binary-tree-part-i.html
// Return #nodes that matches P or Q in the subtree.
int countMatchesPQ(Node *root, Node *p, Node *q) {
if (!root) return 0;
int matches = countMatchesPQ(root->left, p, q) + countMatchesPQ(root->right, p, q);
if (root == p || root == q)
return 1 + matches;
else
return matches;
}
Node *LCA(Node *root, Node *p, Node *q) {
if (!root || !p || !q) return NULL;
if (root == p || root == q) return root;
int totalMatches = countMatchesPQ(root->left, p, q);
if (totalMatches == 1)
return root;
else if (totalMatches == 2)
return LCA(root->left, p, q);
else /* totalMatches == 0 */
return LCA(root->right, p, q);
}
but I was wondering how compute the time complexity of the algorithm,can anyone help me?