Open In App

Wavelet Trees | Introduction

Last Updated : 28 Apr, 2023
Improve
Improve
Like Article
Like
Save
Share
Report

A wavelet tree is a data structure that recursively partitions a stream into two parts until we’re left with homogeneous data. The name derives from an analogy with the wavelet transform for signals, which recursively decomposes a signal into low-frequency and high-frequency components. Wavelet trees can be used to answer range queries efficiently.
Consider the problem to find number of elements in a range [L, R] of a given array A which are less than x. One way to solve this problem efficiently is using Persistent Segment Tree data structure. But we can also solve this easily using Wavelet Trees. Let us see how!

Constructing Wavelet TreesEvery node in a wavelet tree is represented by an array which is the subsequence of original array and a range [L, R]. Here [L, R] is the range in which elements of array falls. That is, ‘R’ denotes maximum element in the array and ‘L’ denotes the smallest element. So, the root node will contain the original array in which elements are in range [L, R]. Now we will calculate the middle of the range [L, R] and stable partition the array in two halves for the left and right childs. Therefore, the left child will contains elements that lies in range [L, mid] and right child will contain elements that lies in the range [mid+1, R]. 
Suppose we are given an array of integers. Now we compute the mid (Max + Min / 2) and form two children. 
Left Children: Integers less than/equal to Mid 
Right Children: Integers greater than Mid 
We recursively perform this operation until all node of similar elements are formed.
Given array : 0 0 9 1 2 1 7 6 4 8 9 4 3 7 5 9 2 7 0 5 1 0 
 

Forming a wavelet tree

To construct a Wavelet Tree, let us see what will we need to store at each node. So at each node of the tree, we will store two arrays say S[] and freq[]. The array S[] will be a subsequence of the original array A[] and the array freq[] will store the count of the elements that will go to left and right childs of the node. That is, freq[i] will denote the count of elements from the first i elements of S[] that will go to left child. Therefore, count of elements that will go to right child can be easily calculated as (i – freq[i]).
Below example shows how to maintain freq[] array: 
 

Array : 1 5 2 6 4 4
Mid = (1 + 6) / 2 = 3
Left Child : 1 2
Right Child : 5 6 4 4

To maintain frequency array, we will check if the element is less than Mid or not. If yes, then we will add 1 to last element of frequency array, else 0 and push back again.
For, above array : 
Freq array :{1, 1, 2, 2, 2, 2}
It implies 1 element will go to left child of this node from index 1 and 2, and 2 elements will go to left child from indices 3 to 6. This can be easily depicted from the above given array. 
To compute the number of elements moving to right subtree, we subtract freq[i] from i. 
 

From index 1, 0 elements go to right subtree.
From index 2, 1 element go to right subtree.
From index 3, 1 element go to right subtree.
From index 4, 2 elements go to right subtree.
From index 5, 3 elements go to right subtree.
From index 6, 4 elements go to right subtree.

We can use the stable_partition function and lambda expression in C++ STL to easily stable partition the array around a pivot without distorting the order of elements in original sequence. It is highly recommended to go through the stable_partition and lambda expression articles before moving onto implementation. 
Below is the implementation of construction of Wavelet Trees: 
 

CPP




// CPP code to implement wavelet trees
#include <bits/stdc++.h>
using namespace std;
#define N 100000
 
// Given array
int arr[N];
 
// wavelet tree class
class wavelet_tree {
public:
    // Range to elements
    int low, high;
 
    // Left and Right children
    wavelet_tree *l, *r;
 
    vector<int> freq;
 
    // Default constructor
    // Array is in range [x, y]
    // Indices are in range [from, to]
    wavelet_tree(int* from, int* to, int x, int y)
    {
        // Initialising low and high
        low = x, high = y;
 
        // Array is of 0 length
        if (from >= to)
            return;
 
        // Array is homogeneous
        // Example : 1 1 1 1 1
        if (high == low) {
 
            // Assigning storage to freq array
            freq.reserve(to - from + 1);
 
            // Initialising the Freq array
            freq.push_back(0);
 
            // Assigning values
            for (auto it = from; it != to; it++)
 
                // freq will be increasing as there'll
                // be no further sub-tree
                freq.push_back(freq.back() + 1);
 
            return;
        }
 
        // Computing mid
        int mid = (low + high) / 2;
 
        // Lambda function to check if a number is
        // less than or equal to mid
        auto lessThanMid
            = [mid](int x) { return x <= mid; };
 
        // Assigning storage to freq array
        freq.reserve(to - from + 1);
 
        // Initialising the freq array
        freq.push_back(0);
 
        // Assigning value to freq array
        for (auto it = from; it != to; it++)
 
            // If lessThanMid returns 1(true), we add
            // 1 to previous entry. Otherwise, we add
            // 0 (element goes to right sub-tree)
            freq.push_back(freq.back() + lessThanMid(*it));
 
        // std::stable_partition partitions the array w.r.t
        // Mid
        auto pivot
            = stable_partition(from, to, lessThanMid);
 
        // Left sub-tree's object
        l = new wavelet_tree(from, pivot, low, mid);
 
        // Right sub-tree's object
        r = new wavelet_tree(pivot, to, mid + 1, high);
    }
};
 
// Driver code
int main()
{
    int size = 5, high = INT_MIN;
    int arr[] = { 1, 2, 3, 4, 5 };
    for (int i = 0; i < size; i++)
        high = max(high, arr[i]);
 
    // Object of class wavelet tree
    wavelet_tree obj(arr, arr + size, 1, high);
 
    return 0;
}


Java




// JAVA code to implement wavelet trees
import java.util.*;
 
class WaveletTree {
    int low, high;
    WaveletTree l, r;
    List<Integer> freq;
 
    public WaveletTree(int[] arr, int low, int high)
    {
        // Store the range of values in the tree
        this.low = low;
        this.high = high;
 
        // If there is only one element, create a list with
        // a single element
        if (low == high) {
            freq = new ArrayList<>();
            freq.add(0);
            for (int i = 0; i < arr.length; i++) {
                freq.add(freq.get(i) + 1);
            }
            return;
        }
 
        // Split the input array into two based on the
        // midpoint of the range
        int mid = (low + high) / 2;
        freq = new ArrayList<>();
        freq.add(0);
        int[] leftArr = new int[arr.length];
        int[] rightArr = new int[arr.length];
        int leftIndex = 0, rightIndex = 0;
        for (int i = 0; i < arr.length; i++) {
            if (arr[i] <= mid) {
                leftArr[leftIndex++] = arr[i];
                freq.add(freq.get(i) + 1);
            }
            else {
                rightArr[rightIndex++] = arr[i];
                freq.add(freq.get(i));
            }
        }
 
        // Recursively create Wavelet Trees for the left and
        // right arrays
        l = new WaveletTree(leftArr, low, mid);
        r = new WaveletTree(rightArr, mid + 1, high);
    }
}
// Driver code
public class Main {
    public static void main(String[] args)
    {
        int size = 5, high = Integer.MIN_VALUE;
        int[] arr = { 1, 2, 3, 4, 5 };
        for (int i = 0; i < size; i++)
            high = Math.max(high, arr[i]);
        // Object of class wavelet tree
        WaveletTree obj = new WaveletTree(arr, 1, high);
    }
}


Javascript




// javascript code to implement wavelet trees
const N = 100000;
const INT_MIN = -2147483647;
 
 
// wavelet tree class
class wavelet_tree {
 
    constructor(arr, low, high) {
        // Store the range of values in the tree
        this.low = low;
        this.high = high;
 
        // If there is only one element, create a list with a single element
        if (low == high) {
            this.freq = [];
            this.freq.push(0);
            for (let i = 0; i < arr.length; i++) {
                this.freq.push(this.freq[i] + 1);
            }
            return;
        }
         
        // Split the input array into two based on the midpoint of the range
        let mid = Math.floor((low + high) / 2);
        this.freq = [];
        this.freq.push(0);
        let leftArr = new Array(arr.length);
        let rightArr = new Array(arr.length);
        let leftIndex = 0, rightIndex = 0;
        for (let i = 0; i < arr.length; i++) {
            if (arr[i] <= mid) {
                leftArr[leftIndex++] = arr[i];
                this.freq.push(this.freq[i] + 1);
            } else {
                rightArr[rightIndex++] = arr[i];
                this.freq.push(this.freq[i]);
            }
        }
 
        // Recursively create Wavelet Trees for the left and right arrays
        this.l = new wavelet_tree(leftArr, low, mid);
        this.r = new wavelet_tree(rightArr, mid + 1, high);
    }
}
 
 
let size = 5;
let high = INT_MIN;   
let arr = [1 , 2, 3, 4, 5];
for (let i = 0; i < size; i++)
    high = Math.max(high, arr[i]);   
 
// Object of class wavelet tree
let obj = new wavelet_tree(arr, 1, high);
 
// The code is contributed by Nidhi goel.


Python3




class WaveletTree:
  # Constructor function
    def __init__(self, from_idx, to_idx, low, high, arr):
      # If the array is homogeneous, store the frequency and return
        self.low = low
        self.high = high
        if from_idx >= to_idx:
            return
        if self.high == self.low:
            self.freq = [0] * (to_idx - from_idx + 2)
            for i in range(from_idx, to_idx + 1):
                self.freq[i - from_idx + 1] = self.freq[i - from_idx] + 1
            return
        mid = (self.low + self.high) // 2
         # Compute frequency array
        self.freq = [0] * (to_idx - from_idx + 2)
        for i in range(from_idx, to_idx + 1):
            self.freq[i - from_idx + 1] = self.freq[i - from_idx] + (arr[i] <= mid)
        pivot = from_idx
        while pivot <= to_idx and arr[pivot] <= mid:
            pivot += 1
        self.l = WaveletTree(from_idx, pivot - 1, self.low, mid, arr)
        self.r = WaveletTree(pivot, to_idx, mid + 1, self.high, arr)
 
    def kOrLess(self, l, r, k):
        if l > r or k < self.low:
            return 0
        if self.high <= k:
            return r - l + 1
        LtCount = self.freq[l - 1]
        RtCount = self.freq[r]
        return (
            self.l.kOrLess(LtCount + 1, RtCount, k) +
            self.r.kOrLess(l - LtCount, r - RtCount, k)
        )
 
# Example usage
size = 5
high = float('-inf')
 
# Array : 1 2 3 4 5
arr = [1, 2, 3, 4, 5]
 
for i in range(size):
    high = max(high, arr[i])
 
# Object of class wavelet tree
obj = WaveletTree(0, size - 1, 1, high, arr)
 
# count of elements less than 2 in range [1,3]
print(obj.kOrLess(1, 3, 2))


C#




using System;
using System.Collections.Generic;
 
class WaveletTree {
    int low, high;
    WaveletTree l, r;
    List<int> freq;
 
    public WaveletTree(int[] arr, int low, int high)
    {
        // Store the range of values in the tree
        this.low = low;
        this.high = high;
 
        // If there is only one element, create a list with
        // a single element
        if (low == high) {
            freq = new List<int>();
            freq.Add(0);
            for (int i = 0; i < arr.Length; i++) {
                freq.Add(freq[i] + 1);
            }
            return;
        }
 
        // Split the input array into two based on the
        // midpoint of the range
        int mid = (low + high) / 2;
        freq = new List<int>();
        freq.Add(0);
        int[] leftArr = new int[arr.Length];
        int[] rightArr = new int[arr.Length];
        int leftIndex = 0, rightIndex = 0;
        for (int i = 0; i < arr.Length; i++) {
            if (arr[i] <= mid) {
                leftArr[leftIndex++] = arr[i];
                freq.Add(freq[i] + 1);
            }
            else {
                rightArr[rightIndex++] = arr[i];
                freq.Add(freq[i]);
            }
        }
 
        // Recursively create Wavelet Trees for the left and
        // right arrays
        l = new WaveletTree(leftArr, low, mid);
        r = new WaveletTree(rightArr, mid + 1, high);
    }
}
 
// Driver code
class Program {
    static void Main(string[] args)
    {
        int size = 5, high = int.MinValue;
        int[] arr = { 1, 2, 3, 4, 5 };
        for (int i = 0; i < size; i++)
            high = Math.Max(high, arr[i]);
 
        // Object of class wavelet tree
        WaveletTree obj = new WaveletTree(arr, 1, high);
    }
}


Output

 

Height of the tree: O(log(max(A)) , where max(A) is the maximum element in the array A[].
 

Querying in Wavelet Trees

We have already constructed our wavelet tree for the given array. Now we will move on to our problem to calculate number of elements less than or equal to x in range [ L,R ] in the given array. 
So, for each node we have a subsequence of original array, lowest and highest values present in the array and count of elements in left and right child.
Now, 
 

If high <= x, 
   we return R - L + 1. 
i.e. all the elements in the current range is less than x.

Otherwise, We will use variable LtCount = freq[ L-1 ] (i.e. elements going to left sub-tree from L-1) , RtCount = freq[ R ] (i.e. elements going to right sub-tree from R) 
Now, we recursively call and add the return values of : 
 

left sub-tree with range[ LtCount + 1, RtCount ] and, 
right sub-tree with range[ L - Ltcount,R - RtCount ]

Below is the implementation in C++: 
 

CPP




// CPP program for querying in
// wavelet tree Data Structure
#include <bits/stdc++.h>
using namespace std;
#define N 100000
 
// Given Array
int arr[N];
 
// wavelet tree class
class wavelet_tree {
public:
    // Range to elements
    int low, high;
 
    // Left and Right child
    wavelet_tree* l, *r;
 
    vector<int> freq;
 
    // Default constructor
    // Array is in range [x, y]
    // Indices are in range [from, to]
    wavelet_tree(int* from, int* to, int x, int y)
    {
        // Initialising low and high
        low = x, high = y;
 
        // Array is of 0 length
        if (from >= to)
            return;
 
        // Array is homogeneous
        // Example : 1 1 1 1 1
        if (high == low) {
            // Assigning storage to freq array
            freq.reserve(to - from + 1);
 
            // Initialising the Freq array
            freq.push_back(0);
 
            // Assigning values
            for (auto it = from; it != to; it++)
             
                // freq will be increasing as there'll
                // be no further sub-tree
                freq.push_back(freq.back() + 1);
             
            return;
        }
 
        // Computing mid
        int mid = (low + high) / 2;
 
        // Lambda function to check if a number
        // is less than or equal to mid
        auto lessThanMid = [mid](int x) {
            return x <= mid;
        };
 
        // Assigning storage to freq array
        freq.reserve(to - from + 1);
 
        // Initialising the freq array
        freq.push_back(0);
 
        // Assigning value to freq array
        for (auto it = from; it != to; it++)
 
            // If lessThanMid returns 1(true), we add
            // 1 to previous entry. Otherwise, we add 0
            // (element goes to right sub-tree)
            freq.push_back(freq.back() + lessThanMid(*it));       
 
        // std::stable_partition partitions the array w.r.t Mid
        auto pivot = stable_partition(from, to, lessThanMid);
 
        // Left sub-tree's object
        l = new wavelet_tree(from, pivot, low, mid);
 
        // Right sub-tree's object
        r = new wavelet_tree(pivot, to, mid + 1, high);
    }
 
    // Count of numbers in range[L..R] less than
    // or equal to k
    int kOrLess(int l, int r, int k)
    {
        // No elements int range is less than k
        if (l > r or k < low)
            return 0;
 
        // All elements in the range are less than k
        if (high <= k)
            return r - l + 1;
 
        // Computing LtCount and RtCount
        int LtCount = freq[l - 1];
        int RtCount = freq[r];
 
        // Answer is (no. of element <= k) in
        // left + (those <= k) in right
        return (this->l->kOrLess(LtCount + 1, RtCount, k) +
             this->r->kOrLess(l - LtCount, r - RtCount, k));
    }
 
};
 
// Driver code
int main()
{
    int size = 5, high = INT_MIN;       
    int arr[] = {1, 2, 3, 4, 5};   
     
    // Array : 1 2 3 4 5
    for (int i = 0; i < size; i++)    
        high = max(high, arr[i]);
 
    // Object of class wavelet tree
    wavelet_tree obj(arr, arr + size, 1, high);
 
    // count of elements less than 2 in range [1,3]
    cout << obj.kOrLess(0, 3, 2) << '\n';
 
    return 0;
}


Python3




class WaveletTree:
  # Constructor function
    def __init__(self, from_idx, to_idx, low, high, arr):
      # If the array is homogeneous, store the frequency and return
        self.low = low
        self.high = high
        if from_idx >= to_idx:
            return
        if self.high == self.low:
            self.freq = [0] * (to_idx - from_idx + 2)
            for i in range(from_idx, to_idx + 1):
                self.freq[i - from_idx + 1] = self.freq[i - from_idx] + 1
            return
        mid = (self.low + self.high) // 2
         # Compute frequency array
        self.freq = [0] * (to_idx - from_idx + 2)
        for i in range(from_idx, to_idx + 1):
            self.freq[i - from_idx + 1] = self.freq[i - from_idx] + (arr[i] <= mid)
        pivot = from_idx
        while pivot <= to_idx and arr[pivot] <= mid:
            pivot += 1
        self.l = WaveletTree(from_idx, pivot - 1, self.low, mid, arr)
        self.r = WaveletTree(pivot, to_idx, mid + 1, self.high, arr)
 
    def kOrLess(self, l, r, k):
        if l > r or k < self.low:
            return 0
        if self.high <= k:
            return r - l + 1
        LtCount = self.freq[l - 1]
        RtCount = self.freq[r]
        return (
            self.l.kOrLess(LtCount + 1, RtCount, k) +
            self.r.kOrLess(l - LtCount, r - RtCount, k)
        )
 
# Example usage
size = 5
high = float('-inf')
 
# Array : 1 2 3 4 5
arr = [1, 2, 3, 4, 5]
 
for i in range(size):
    high = max(high, arr[i])
 
# Object of class wavelet tree
obj = WaveletTree(0, size - 1, 1, high, arr)
 
# count of elements less than 2 in range [1,3]
print(obj.kOrLess(1, 3, 2))


Javascript




// Define a WaveletTree class with properties low, high, l, r, and freq
// Default constructor
// Array is in range [x, y]
// Indices are in range [from, to]
class WaveletTree {
    constructor(from, to, x, y) {
        this.low = x;
        this.high = y;
        if (from >= to) return;
         
        // Array is homogeneous
        // Example : 1 1 1 1 1
        if (this.high == this.low) {
            this.freq = [0];
            for (let it = from; it <= to; it++) this.freq.push(this.freq[this.freq.length - 1] + 1);
            return;
        }
        let mid = Math.floor((this.low + this.high) / 2);
         
        // Lambda function to check if a number
        // is less than or equal to mid
        let lessThanMid = x => x <= mid;
        this.freq = [0];
        for (let it = from; it <= to; it++) this.freq.push(this.freq[this.freq.length - 1] + lessThanMid(arr[it]));
         
        // std::stable_partition partitions the array w.r.t Mid
        let pivot = from;
        while (pivot <= to && lessThanMid(arr[pivot])) pivot++;
         
        // Left sub-tree's object
        this.l = new WaveletTree(from, pivot - 1, this.low, mid);
         
        // Right sub-tree's object
        this.r = new WaveletTree(pivot, to, mid + 1, this.high);
    }
 
    // Count of numbers in range[L..R] less than
    // or equal to k
    kOrLess(l, r, k) {
        // No elements int range is less than k
        if (l > r || k < this.low) return 0;
         
        // All elements in the range are less than k
        if (this.high <= k) return r - l + 1;
         
        // Computing LtCount and RtCount
        let LtCount = this.freq[l - 1];
        let RtCount = this.freq[r];
         
        // Answer is (no. of element <= k) in
        // left + (those <= k) in right
        return (
            this.l.kOrLess(LtCount + 1, RtCount, k) +
            this.r.kOrLess(l - LtCount, r - RtCount, k)
        );
    }
}
 
// Example usage
let size = 5,
    high = Number.MIN_SAFE_INTEGER;
     
// Array : 1 2 3 4 5
let arr = [1, 2, 3, 4, 5];
 
for (let i = 0; i < size; i++) high = Math.max(high, arr[i]);
 
// Object of class wavelet tree
let obj = new WaveletTree(0, size - 1, 1, high);
 
// count of elements less than 2 in range [1,3]
console.log(obj.kOrLess(0 + 1 , 3 + 1 , 2));


+

Output

2

Time Complexity: O(log(max(A)) , where max(A) is the maximum element in the array A[]. 
In this post we have discussed about a single problem on range queries without update. In further we will be discussing on range updates also.
References
 

 



Like Article
Suggest improvement
Previous
Next
Share your thoughts in the comments

Similar Reads