Table of contents

• Relation map
  • Install
  • Link
• Problem and Solve
  • Performance due to time complexity
  • Boilerplate to maintain relationship
• Feedback & Contributions
• Features
  • BiMap
  • OneToManyMap
  • Tree

Relation map

A flexible relational mapping library supporting 1:1 and 1:N

Install

npm i relation-map

yarn add relation-map

Link

npm

github

Problem and Solve

Performance due to time complexity

run from codeSandbox

• When we search a value from typical (unidirectional) map, its time complexity is O(N).
• This is because an exhaustive search is required to access and manipulate values ​​rather than keys.
• But BiMap is O(1). Moreover, updating, deleting and adding are also of the same complexity.

import { BiMap } from "relation-map";

const biMap = new BiMap();
const map = new Map();
const N = 12345;
for (let i = 0; i < N; i++) {
  map.set(String(i), i);
  biMap.set(String(i), i);
}
const mapTime = { total: 0, average: 0 };
const biMapTime = { total: 0, average: 0 };

const start1 = performance.now();
for (let i = 0; i < N; i++) {
  // To find value, A complete search is required.
  for (const v of map.values()) {
    // If find, stop
    if (v === i) break;
  }
}
const end1 = performance.now();
mapTime.total = end1 - start1;
mapTime.average = (end1 - start1) / N;

const start2 = performance.now();
for (let i = 0; i < N; i++) {
  // Find with O(1) complexity.
  biMap.getByRight(i);
}
const end2 = performance.now();

// Log in ms
biMapTime.total = end2 - start2;
biMapTime.average = (end2 - start2) / N;
console.log({ mapTime });
console.log({ biMapTime });

• The difference in results is stark!

{
  mapTime: { total: 340.80774399999996, average: 0.027606945646010527 }
}

{
  biMapTime: { total: 1.2093150000000037, average: 0.000097959902794654 }
}

Boilerplate to maintain relationship

• To solve time complexity, you can consider creating a reference data structure.
• However, each time you update, delete, or add data, the amount of code you will need to write will double.

const map = new Map();

const map1 = new Map();
const map2 = new Map(); // for reference to solve time complexity of map1.
// Add data pair
for (let i = 0; i < 10000; i++) {
  map.set(String(i), i);

  map1.set(String(i), i);
  map2.set(i, String(i)); // Also, need to set it on map2.
}
// Update
map.set("10", 7);
map.delete("7"); // To maintain 1:1

map1.set("10", 7);
map2.set(7, "10");
// Also, need to maintain 1:1 on both.
map1.delete("7");
map2.delete(10);

• This is a simple example.
• However, in actual work, there is the following problem:
  • As the amount of code increases, the probability of bugs increases.
  • As the amount of code increases, stamina will be depleted.
  • As the amount of code increases, project structure will be dirty.
• We can work more simply with a relation map!

const biMap = new BiMap();
// ...
biMap.set("10", 7); // End

Feedback & Contributions

We welcome all types of feedback and contributions! If you have any of the following, please feel free to open an issue

Your input is valuable to us and helps improve the project for everyone. Thank you for contributing!

Features

BiMap

OneToManyMap

Tree