Not Frequent

Introduction to Tree Algorithms

Authors: Nathan Chen, Siyong Huang, Albert Ye

Introducing a special type of graph: trees.

Edit This Page

Focus Problem – try your best to solve this problem before continuing!

Trees are generally treated very differently from general graph problems.



traversing tree, diameter


Some properties/definitions of trees:

  • A graph is a tree iff it is connected and contains NN nodes and N1N-1 edges
  • A graph is a tree iff every pair of nodes has exactly one simple path between them
  • A graph is a tree iff it is connected and does not contain any cycles

General Tree Terminology:

  • A leaf of a tree is any node in the tree with degree 11
    • If the tree is rooted, the root with a single child is not typically considered a leaf, but depending on the problem, this is not always the case
  • A star graph has two common definitions. Try to understand what they mean - they typically appear in subtasks.
    • Definition 1: Only one node has degree greater than 11
    • Definition 2: Only one node has degree greater than 22
  • A forest is a graph such that each connected component is a tree

Rooted Tree Terminology:

  • A root of a tree is any node of the tree that is considered to be at the 'top'
  • A parent of a node nn is the first node along the path from nn to the root
    • The root does not have a parent. This is typically done in code by setting the parent of the root to be 1-1.
  • The ancestors of a node are its parent and parent's ancestors
    • Typically, a node is considered its own ancestor as well (such as in the subtree definition)
  • The subtree of a node nn are the set of nodes that have nn as an ancestor
    • A node is typically considered to be in its own subtree
    • Note: This is easily confused with subgraph
  • The depth, or level, of a node is its distance from the root

Solution - Subordinates

In this problem we are given the parent of each node of a rooted tree, and we want to compute the subtree size for each node. A subtree is composed of a root node and the subtrees of the root's children. Thus, the size of a subtree is one plus the size of the root's childrens' subtrees.


#include <bits/stdc++.h>
using namespace std;
const int SZ = 2e5;
vector<int> children[SZ];
int subtree_size[SZ], depth[SZ];
void dfs_size(int node) {
subtree_size[node] = 1; // This one represents the root of `node's` subtree



Because Java is so slow, an adjacency list using lists/arraylists results in TLE. Instead, the Java sample code will use the Chinese edge representation.

import java.util.*;
public class Subordinates {
static InputReader in = new InputReader(;
static PrintWriter out = new PrintWriter(System.out);
public static final int MN = 200020;
static int N, M, ans;


In the Python solution, we need to set recursion limit to 2×1052\times10^5.

import sys
sys.setrecursionlimit(200006) # set recursion limit
def dfs(x): # x is the current node
ans = 0 # stores the number of subordinates
for e in edges[x]:
if e != fa[x - 1]:
ans += dfs(e)


StatusSourceProblem NameDifficultyTags
Show TagsConnected Components, Diameter, Tree
Show TagsConnected Components, Tree
Show TagsTree
Show TagsConnected Components, Tree
Show TagsGreedy, Tree
Show TagsTree
Show TagsTree
Show TagsTree
Show TagsTree
Show TagsTree
Show TagsBipartite, Tree
Show TagsDFS, Spanning Tree
Show TagsDFS, Spanning Tree


How is a preorder traversal found for a binary tree?

Question 1 of 3

Module Progress:

Join the USACO Forum!

Stuck on a problem, or don't understand a module? Join the USACO Forum and get help from other competitive programmers!