Somewhat Frequent
0/30

# Binary Jumping

Authors: Benjamin Qi, Neo Wang, Qi Wang

Introduces the problems of finding level ancestors in a tree and computing the lowest common ancestors.

## Binary Jumping

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

### Pro Tip

Binary jumping is more commonly referred to as "binary lifting."

### Solution

To solve this problem, we can use a standard binary lifting implementation where jmp(int x, int d) corresponds to the $d$-th ancestor of $x$.

In our jmp(int x, int d) if our final value of $x$ is $0$, then such a node does not exist and we can simply return $-1$. This is because the lowest number a node can be is $1$ (the root of the tree).

In our implementation, we test if we jump in powers of two by using the $\&$ operator. If the $i$th bit on the right is toggled, then we jump. For example, a jump of $13$ would correspond to the binary number $1101$. We would jump 3 times on bits $0, 2, 3$ (in that order) counting from the right.

Illustration of the jump method

To calculate the rows required for the int up[MS][MX] array, use the formula $\lceil \log_2{N} \rceil$ which in our case simplifies to $\lceil \log_2{(2\cdot 10^5)}\rceil = 18$.

### Pro Tip

It never hurts to add additional rows - or columns, depending on your implementation - (as long as it's reasonable)!

C++

#include <bits/stdc++.h>
using namespace std;
#define FOR(i, a, b) for(int i = (a); i < (b); i++)#define FORE(i, a, b) for(int i = (a); i <= (b); i++)#define F0R(i, a) for(int i = 0; i < (a); i++)#define trav(a, x) for (auto& a : x)
int N, Q;

### Problems

StatusSourceProblem NameDifficultyTags
CSESEasy
Show TagsBinary Jumping
CSESNormal
Show TagsFunctional Graph
CSESNormal
Show Tags2P, Binary Jumping, Binary Search
POINormal
CFNormal
Baltic OINormal
Baltic OINormal
PlatHard
Show TagsBinary Jumping
Baltic OIVery Hard

## Lowest Common Ancestor

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

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

### Solution (Company Queries II)

Resources
CPH

Brief description/solution

SansPapyrus683

Alternative implementation

To find $\textrm{lca}(a, b)$, we can first lift the lower node of $a$ and $b$ to the same depth as the other. Then, we lift both nodes up decrementally. At the end, the parent of either node is the LCA of the two.

C++

 Code Snippet: Benq Template (Click to expand)
int N, Q, T = 1;int depth[200005];int up[200005][20];vi adj[200005];
void dfs(int v) {	FOR(i, 1, 20) {		up[v][i] = up[up[v][i-1]][i-1];

### Alternative Solution (Company Queries II)

As mentioned in the Euler Tour Technique module, let $\texttt{st}, \texttt{en}, \texttt{up}$ be the time-in, time-out, and ancestor table for all nodes in the tree.

These can be filled with a DFS traversal. $\texttt{up}$ can be generated because DFS allows the ancestors to be filled first before traversing the current node.

With this information, we can declare node $a$ an ancestor of $b$ if $\texttt{st}[a] \le \texttt{st}[b]$ and $\texttt{en}[a] \ge \texttt{en}[b]$.

We know that if $a$ is an ancestor of $b$ or $b$ is an ancestor of $a$, the answer will be $a$ or $b$ respectively. Otherwise, we lift one of the nodes up (in this case, $a$) decrementally while it is not the ancestor of $b$. Therefore, if $\texttt{up}[a][i]$ is not an ancestor of $b$, then we can set $a$ to be $\texttt{up}[a][i]$, else, we will decrement $i$ and try to find a lower common ancestor. Afterwards, our answer is the parent of $a$.

C++

 Code Snippet: Benq Template (Click to expand)
int N, Q, T = 1;vi st, en;int up[200005][20];vi adj[200005];
void dfs(int v, int p) {	st[v] = T++;	up[v][0] = p;

### Alternate Solution II (Company Queries II)

We can also find the LCA of two nodes using the Tarjan's Offline LCA algorithm. By taking advantage of the DFS traversal, we can precompute the answers to the queries through forming subtrees and calculating the common parent with a similar structure as Disjoint-Set Union.

C++

 Code Snippet: Benq Template (Click to expand)#include <bits/stdc++.h>using namespace std;
#pragma GCC optimize("O3")
#pragma region	using ll = long long;using db = long double; // or double, if TL is tight
Resources
cp-algo

### Solution (Distance Queries)

Find LCA of node $a$ and $b$ as described above. Then, the distance between the two nodes would be the $\texttt{depth}[a] + \texttt{depth}[b] - 2 \cdot \texttt{depth}[\textrm{lca}(a, b)]$.

C++

 Code Snippet: Benq Template (Click to expand)
int N, Q, T = 1;int depth[200005];int up[200005][20];vi adj[200005];
void dfs(int v) {	FOR(i, 1, 20) {		up[v][i] = up[up[v][i-1]][i-1];

### Problems

#### USACO

StatusSourceProblem NameDifficultyTags
PlatEasy
Show TagsLCA
PlatNormal
Show TagsLCA
Old GoldNormal
Show TagsBinary Jumping, Euler Tour, Small to Large
PlatHard
Show TagsLCA
PlatHard
Show TagsDiameter
PlatHard
Show TagsLCA
PlatVery Hard
Show TagsLCA

#### General

StatusSourceProblem NameDifficultyTags
CFEasy
Show TagsBinary Jumping
CFNormal
Show TagsLCA
Baltic OINormal
CFNormal
Show TagsLCA
CSANormal
Show TagsLCA
CFNormal
Show TagsLCA
DMOJNormal
Show TagsLCA
Show TagsBinary Jumping, DFS, LCA
CFHard
Show TagsBinary Jumping, LCA
TLXHard
Show TagsLCA
TLXHard
Show TagsLCA

### This section is not complete.

Any help would be appreciated! Just submit a Pull Request on Github.

figure out a better way to order these, difficulties aren't really accurate

### 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!