dtst
A collection of data structures to use
Last updated 3 years ago by cchrispy .
MIT · Repository · Bugs · Original npm · Tarball · package.json
$ cnpm install dtst 
SYNC missed versions from official npm registry.

dtst

A collection of data structures to use, with comments and examples. Still a work in progress!

What's included

Usage

Tree

Trees are commonly used data structures for storing hierarchical information, such as a file system on a computer (think about navigating through directories and their sub-directories).

import { Tree } from 'dtst';

// create an instance of Tree
var tree = new Tree();

// create the first node in the tree and store the node instance in variable "root"
var root = tree.addNode('Head node');

// add three child nodes, which are attached to the root node by default
var child_1 = tree.addNode('First child');
var child_2 = tree.addNode('Second child');
var child_3 = tree.addNode('Third child');

// attach children to different nodes by specifying a parent node as a second argument
var grandChild_1 = tree.addNode('Child of child_1', child_1);
var grandChild_2 = tree.addNode('Another child of child_1', child_1);
var grandChild_3 = tree.addNode('Child of child_2', child_2);
var greatGrandChild_1 = tree.addNode('Child of grandChild_3', grandChild_3);

/*
** Tree structure:

   root 
     + child_1
       + grandChild_1
       + grandChild_2
     + child_2
       + grandChild_3
         + greatGrandChild_1
     + child_3

**
*/

// check if the tree contains a value
tree.contains('Second child');   // -> returns child_2
tree.contains('I do not exist'); // -> returns false

// delete a node and its descendants
tree.removeNode(child_2);
tree.contains('Second child');          // -> returns false
tree.contains('Child of child_2');      // -> returns false
tree.contains('Child of grandChild_3'); // -> returns false
var size = tree.size;                   // -> 5

Trie

Tries (also known as radix trees or prefix trees) have an ordered data structure that is efficient for information retrieval. There are several ways to construct a Trie, but they all share a tree structure. Each node can represent a single character string, and traversing down a tree can build words or prefixes. A few useful application for Tries: dictionary suggestions, autocomplete dictionaries, searching through a contact list, or searching through phone directories.

import { Trie } from 'dtst';

// create an instance of Trie
var trie = new Trie();

// populate the Trie with a collection of 100 words
const collection = [
  "abrupt", "acid", "adventurous", "ahead", "amused", "appliance", "attach", "bee", "better", 
  "borrow", "bouncy", "branch", "brave", "bushes", "business", "cagey", "carpenter", "cattle", 
  "chin", "cloth", "confuse", "consist", "cool", "count", "craven", "deep", "digestion", "ear", 
  "escape", "evasive", "even", "examine", "fit", "flood", "frequent", "furtive", "gather", 
  "graceful", "grumpy", "guiltless", "half", "hand", "handy", "hate", "hose", "ill-fated", "jail", 
  "kaput", "likeable", "limit", "look", "malicious", "meaty", "mere", "mine", "moaning", "monkey", 
  "nail", "neck", "nippy", "noise", "obedient", "obeisant", "observant", "order", "park", 
  "penitent", "perform", "press", "queue", "quickest", "quince", "racial", "remember", "remind", 
  "replace", "ripe", "roll", "safe", "satisfy", "school", "serious", "shop", "slip", "slope", 
  "small", "spark", "squeamish", "stiff", "store", "supply", "swim", "taste", "tiny", "trees", 
  "use", "useful", "volcano", "vulgar", "whispering"
];
collection.forEach(word => trie.add(word));

// check if certain words or prefixes exist
trie.contains('monkey');   // -> true
trie.contains('monk');     // -> true
trie.contains('abrupt');   // -> true
trie.contains('abruptly'); // -> false

// autocomplete or predict words based off a given prefix
trie.predict('a');   // -> ["abrupt", "acid", "adventurous", "ahead", "amused", "appliance", "attach"]
trie.predict('con'); // -> ["confuse", "consist"]
trie.predict('z');   // -> []

// remove prefixes and their descendant words
trie.remove('rem');
trie.contains('remember'); // -> false
trie.contains('remind');   // -> false
trie.contains('replace');  // -> true
trie.predict('r');         // -> ["racial", "replace", "ripe", "roll"]

Binary Search Tree

Binary Search Trees efficiently store data in a sorted form that allow fast lookup, addition, and removal of items. BST operations uses the principle of binary search when traversing the tree, so lookup/insertion/deletion operations work approximately at logarithmic time complexity.

import { BinarySearchTree } from 'dtst';

// create an instance of BinarySearchTree
var bst = new BinarySearchTree();

// insert key/value pairs, where the key is a unique number (used to determine how data is sorted)
bst.insert(4, 'four'); // the root node
bst.insert(2, 'two');
bst.insert(3, 'three');
bst.insert(1, 'one');
bst.insert(0, 'zero');
bst.insert(10, 'ten');
bst.insert(8, 'eight');
bst.insert(11, 'eleven');
bst.insert(7, 'seven');
bst.insert(6, 'six');

// search for a value given a key
bst.search(10); // -> 'ten'
bst.search(3);  // -> 'three'
bst.size;       // -> 10

// update a value given its key
bst.update(4, 'I am the root node!');
bst.update(8, 'I ate eight');
bst.search(4);  // -> 'I am the root node!'
bst.search(8);  // -> 'I ate eight'

// remove nodes corresponding to a key
// note: this may cause nodes to be rearranged to maintain the sort pattern
bst.remove(6);
bst.remove(4);  // removes the root node

bst.size;       // -> 8
bst.search(4);  // -> null
bst.search(6);  // -> null
bst.search(8);  // -> 'I ate eight'
bst.head;       // -> Node instance { key: 7, value: 'seven' }

Graph

Graphs are vertices (nodes) connected by edges. In fact, a tree data structure can be considered a type of graph. Graphs are useful for anything network related, routing, finding relationships, etc. Connections with friends on social media, Google Maps finding a route based on shortest routes, internet pages linked by hyperlinks, are all real life applications of graphs.

import { Graph } from 'dtst';

// create an instance of Graph
var graph = new Graph();

// add vertices
let sanFrancisco = graph.addVertex('San Francisco');
let sanJose      = graph.addVertex('San Jose');
let sacramento   = graph.addVertex('Sacramento');
let losAngeles   = graph.addVertex('Los Angeles');

// connect vertices with edges
// edges can be given an optional numeric weight, which defaults to 0
// note: edges are directional, meaning an edge from A to B does not mean an edge from B to A
graph.addEdge(sanFrancisco, sacramento, 6);
graph.addEdge(sanFrancisco, sanJose, 4);
graph.addEdge(sanFrancisco, losAngeles, 16);
graph.addEdge(losAngeles, sanFrancisco, 18);
graph.addEdge(sanJose, sacramento);
graph.addEdge(sanJose, sanFrancisco, 5);

// get neighbors (direct connections) of a vertex
sanFrancisco.getNeighbors(); // -> { 'Sacramento', 'San Jose', 'Los Angeles' }
losAngeles.getNeighbors();   // -> { 'San Francisco' }
sanJose.getNeighbors();      // -> { 'Sacramento', 'San Francisco' }
sacramento.getNeighbors();   // -> { }

// remove an edge between two vertices
graph.removeEdge(sanFrancisco, losAngeles);
sanFrancisco.getNeighbors(); // -> { 'Sacramento', 'San Jose' }
losAngeles.getNeighbors();   // -> { 'San Francisco' }

// remove a vertex, which will also remove connections to and from that vertex
graph.removeVertex(sanFrancisco); // graph.removeVertex('San Francisco') will also work
losAngeles.getNeighbors();        // -> { }
sanJose.getNeighbors();           // -> { 'Sacramento' }

Heap

Heaps are tree-based data structures where the parent-child ordering is consistent across the whole heap. In a min heap, every parent node's key is smaller than its children; therefore the root node will always contain the smallest node. In a max heap, every parent node's key is larger. They are useful for keeping track of the smallest (or largest) value in a stream of data. Heaps are often implemented in priority queues.

import { MinHeap } from 'dtst';

// create an instance of MinHeap
var heap = new MinHeap();

// insert some data into the heap
const numbers = [42, 9, 5, 12, 11, 99, 1, 23, 28, 33, 20];
numbers.forEach(number => heap.insert(number));

// peek at the minimum element (the root element)
heap.peek(); // -> 1

// extract or remove the minimum element
// the heap will be re-arranged with the next smallest element as the root
heap.extract(); // -> 1
heep.peek();    // -> 5

// keys correspond to a particular index location of an array
heap.searchForIndex(5); // -> 1 (5 is the smallest element, thus it's at the first index)

// delete an element based on its index location
heap.delete(1); // deleting the first element is essentially the same as heap.extract()
heap.peek();    // -> 9

Linked List

Linked lists are linear collections of nodes where each node has a reference to the next node in line. In a doubly linked list, each node has a reference to the next and previous node. Linked Lists are advantageous over arrays because insertion/deletion operations involve a simple rearrangement of pointers. The same operations on an array require reorganizing the structure because the data is stored contiguously in memory. Linked lists can be used in creating stacks and queues to avoid pre-allocating memory for all elements.

import { LinkedList } from 'dtst';

// create an instance of LinkedList
var list = new LinkedList();

// insert values into the linked list
// use .search(value) to get a node from the list with the given value to be used as a reference
['A', 'B', 'C', 'D', 'E'].forEach(letter => list.addToTail(letter));
list.addToHead('ZERO');
/***  (HEAD) --> ZERO --- A --- B --- C --- D --- E <-- (TAIL)  ***/
list.insert('foo', list.search('C'));
list.addToTail('F');
list.insert('bar', list.search('F'));
/***  (HEAD) --> ZERO --- A --- B --- C --- foo --- D --- E --- F --- bar <-- (TAIL)  ***/
list.getSize(); // -> 9

// delete nodes from the linked list
list.deleteHead();
list.deleteTail();
/***  (HEAD) --> A --- B --- C --- foo --- D --- E --- F <-- (TAIL)  ***/
list.delete(list.search('foo'));
list.delete(list.search('F'));
/***  (HEAD) --> A --- B --- C --- D --- E <-- (TAIL)  ***/

// reverse the linked list
list.reverse();
/***  (HEAD) --> E --- D --- C --- B --- A <-- (TAIL)  ***/

Current Tags

  • 1.0.1                                ...           latest (3 years ago)

2 Versions

  • 1.0.1                                ...           3 years ago
  • 1.0.0                                ...           3 years ago
Maintainers (1)
Downloads
Today 0
This Week 0
This Month 0
Last Day 0
Last Week 1
Last Month 1
Dependencies (2)
Dev Dependencies (4)
Dependents (0)
None

Copyright 2014 - 2016 © taobao.org |