stats library

Author: amejiarosario

src/data-structures/hash-maps/hash-map.js

@@ -1,3 +1,5 @@
+/* global BigInt */
+
 /**
  * Hash Map data structure implementation
  *
@@ -24,30 +26,54 @@ class HashMap {
   }
 
   /**
-   * Decent hash function where each char ascii code is added with an offset depending on the possition
+   * Use the binary represenation (IEEE 754) as a hashCode
    * @param {any} key
    */
-  static hashCode(key) {
-    let hashValue = 0;
-    const stringTypeKey = `${key}${typeof key}`;
-
-    for (let index = 0; index < stringTypeKey.length; index++) {
-      const charCode = stringTypeKey.charCodeAt(index);
-      hashValue += charCode << (index * 8);
-    }
+  static hashCodeForNumber(number) {
+    const buffer = new ArrayBuffer(8); // 8 bytes for float64
+    const dataView = new DataView(buffer);
+    dataView.setFloat64(0, number); // set as float64
+    const longBits = dataView.getBigInt64(0); // read as long int (BigInt)
+    // gurantee only positive (big) integers
+    return longBits > 0 ? longBits : BigInt(2 ** 63) + (longBits * BigInt(-1));
+  }
 
-    return hashValue;
+  /**
+   * Polynomial hsah codes are used to hash String typed keys.
+   * A string is a sequence of characters encoded
+   * using Unicode's numerical value.
+   * @param {any} key converted to string
+   */
+  hashFunctionForString(key) {
+    return Array
+      .from(key.toString())
+      .reduce((hashIndex, char) => this.hashCodeToIndex((41 * hashIndex) + char.codePointAt(0)), 0);
   }
 
   /**
    * A hash function converts keys into array indices
    * @param {any} key
-   * @returns {Number} array index given the bucket size
+   * @returns {BigInt} array index given the bucket size
    */
   hashFunction(key) {
-    const hashValue = HashMap.hashCode(key);
-    const bucketIndex = hashValue % this.buckets.length;
-    return bucketIndex;
+    if (typeof key === 'number') {
+      return this.hashCodeToIndex(HashMap.hashCodeForNumber(key));
+    }
+    return this.hashFunctionForString(key);
+  }
+
+  /**
+   * Multiply-Add-Divide (MAD) compression
+   * @param {number} hash hashCode
+   * @param {number} size bucket array size
+   * @returns {number} array bucket index
+   */
+  hashCodeToIndex(hash, size = this.buckets.length) {
+    const prime = BigInt(6700417); // prime number > size
+    const a = prime - BigInt(2); // integer from range [1..(p-1)]
+    const b = prime - BigInt(1); // integer from range [0..(p-1)]
+    const hashIndex = (((a * BigInt(hash)) + b) % prime) % BigInt(size);
+    return parseInt(hashIndex, 10);
   }
 
   /**

src/data-structures/hash-maps/hash-map.spec.js

@@ -2,131 +2,133 @@ const HashMap = require('./hash-map');
 // import HashMap from './hash-map';
 
 
-describe('without collisions', () => {
-  let hashMap;
-
-  beforeEach(() => {
-    hashMap = new HashMap();
-  });
-
-  it('gets values', () => {
-    hashMap.set('test', 'one');
-    expect(hashMap.get('test')).toBe('one');
-  });
-
-  it('should increase load factor and size', () => {
-    expect(hashMap.getLoadFactor()).toBe(0);
-    expect(hashMap.size).toBe(0);
-    hashMap.set('test', 'one');
-    expect(hashMap.getLoadFactor()).toBe(1 / 16);
-    expect(hashMap.size).toBe(1);
-  });
-
-  it('should overwrite values and keep same size', () => {
-    hashMap.set('test', 'uno');
-    expect(hashMap.get('test')).toBe('uno');
-    hashMap.set('test', 'dos');
-    expect(hashMap.get('test')).toBe('dos');
-    expect(hashMap.size).toBe(1);
-  });
-
-  it('should return with has', () => {
-    expect(hashMap.has('test')).toBe(false);
-    hashMap.set('test', 'uno');
-    expect(hashMap.has('test')).toBe(true);
-  });
-
-  it('should update keys on deletes', () => {
-    hashMap.set('Despacito', 'Luis Fonsi');
-    hashMap.set('Bailando', 'Enrique Iglesias');
-    hashMap.set('Dura', 'Daddy Yankee');
-
-    expect(hashMap.delete('Bailando')).toBe(true);
-    expect(hashMap.delete('Bailando')).toBe(false);
-    expect(hashMap.get('Bailando')).toBe(undefined);
-
-    expect(hashMap.keys()).toEqual(['Despacito', 'Dura']);
-  });
-});
-
-describe('with many values (and collisions)', () => {
-  let hashMap;
-
-  beforeEach(() => {
-    hashMap = new HashMap(2, 0);
-
-    hashMap.set('Pineapple', 'Pen Pineapple Apple Pen');
-    hashMap.set('Despacito', 'Luis Fonsi');
-    hashMap.set('Bailando', 'Enrique Iglesias');
-    hashMap.set('Dura', 'Daddy Yankee');
-    hashMap.set('Lean On', 'Major Lazer');
-    hashMap.set('Hello', 'Adele');
-    hashMap.set('All About That Bass', 'Meghan Trainor');
-    hashMap.set('This Is What You Came For', 'Calvin Harris ');
-  });
-
-  it('gets values', () => {
-    hashMap.set('test', 'one');
-    expect(hashMap.get('test')).toBe('one');
-    expect(hashMap.get('Dura')).toBe('Daddy Yankee');
-    expect(hashMap.get('Bailando')).toBe('Enrique Iglesias');
-  });
-
-  it('should increase load factor and size', () => {
-    expect(hashMap.getLoadFactor()).toBe(4);
-    expect(hashMap.size).toBe(8);
-    hashMap.set('test', 'one');
-    expect(hashMap.getLoadFactor()).toBe(9 / 2);
-    expect(hashMap.size).toBe(9);
+describe('HashMap Tests', () => {
+  describe('without collisions', () => {
+    let hashMap;
+
+    beforeEach(() => {
+      hashMap = new HashMap();
+    });
+
+    it('gets values', () => {
+      hashMap.set('test', 'one');
+      expect(hashMap.get('test')).toBe('one');
+    });
+
+    it('should increase load factor and size', () => {
+      expect(hashMap.getLoadFactor()).toBe(0);
+      expect(hashMap.size).toBe(0);
+      hashMap.set('test', 'one');
+      expect(hashMap.getLoadFactor()).toBe(1 / 16);
+      expect(hashMap.size).toBe(1);
+    });
+
+    it('should overwrite values and keep same size', () => {
+      hashMap.set('test', 'uno');
+      expect(hashMap.get('test')).toBe('uno');
+      hashMap.set('test', 'dos');
+      expect(hashMap.get('test')).toBe('dos');
+      expect(hashMap.size).toBe(1);
+    });
+
+    it('should return with has', () => {
+      expect(hashMap.has('test')).toBe(false);
+      hashMap.set('test', 'uno');
+      expect(hashMap.has('test')).toBe(true);
+    });
+
+    it('should update keys on deletes', () => {
+      hashMap.set('Despacito', 'Luis Fonsi');
+      hashMap.set('Bailando', 'Enrique Iglesias');
+      hashMap.set('Dura', 'Daddy Yankee');
+
+      expect(hashMap.delete('Bailando')).toBe(true);
+      expect(hashMap.delete('Bailando')).toBe(false);
+      expect(hashMap.get('Bailando')).toBe(undefined);
+
+      expect(hashMap.keys()).toEqual(['Despacito', 'Dura']);
+    });
   });
 
-  it('should overwrite values and keep same size', () => {
-    hashMap.set('test', 'uno');
-    expect(hashMap.get('test')).toBe('uno');
-    hashMap.set('test', 'dos');
-    expect(hashMap.get('test')).toBe('dos');
-    expect(hashMap.size).toBe(9);
+  describe('with many values (and collisions)', () => {
+    let hashMap;
+
+    beforeEach(() => {
+      hashMap = new HashMap(2, 0);
+
+      hashMap.set('Pineapple', 'Pen Pineapple Apple Pen');
+      hashMap.set('Despacito', 'Luis Fonsi');
+      hashMap.set('Bailando', 'Enrique Iglesias');
+      hashMap.set('Dura', 'Daddy Yankee');
+      hashMap.set('Lean On', 'Major Lazer');
+      hashMap.set('Hello', 'Adele');
+      hashMap.set('All About That Bass', 'Meghan Trainor');
+      hashMap.set('This Is What You Came For', 'Calvin Harris ');
+    });
+
+    it('gets values', () => {
+      hashMap.set('test', 'one');
+      expect(hashMap.get('test')).toBe('one');
+      expect(hashMap.get('Dura')).toBe('Daddy Yankee');
+      expect(hashMap.get('Bailando')).toBe('Enrique Iglesias');
+    });
+
+    it('should increase load factor and size', () => {
+      expect(hashMap.getLoadFactor()).toBe(4);
+      expect(hashMap.size).toBe(8);
+      hashMap.set('test', 'one');
+      expect(hashMap.getLoadFactor()).toBe(9 / 2);
+      expect(hashMap.size).toBe(9);
+    });
+
+    it('should overwrite values and keep same size', () => {
+      hashMap.set('test', 'uno');
+      expect(hashMap.get('test')).toBe('uno');
+      hashMap.set('test', 'dos');
+      expect(hashMap.get('test')).toBe('dos');
+      expect(hashMap.size).toBe(9);
+    });
+
+    it('should return with has', () => {
+      expect(hashMap.has('test')).toBe(false);
+      hashMap.set('test', 'uno');
+      expect(hashMap.has('test')).toBe(true);
+    });
   });
 
-  it('should return with has', () => {
-    expect(hashMap.has('test')).toBe(false);
-    hashMap.set('test', 'uno');
-    expect(hashMap.has('test')).toBe(true);
-  });
-});
-
-describe('#rehash', () => {
-  let hashMap;
-
-  beforeEach(() => {
-    hashMap = new HashMap();
-
-    hashMap.set('Pineapple', 'Pen Pineapple Apple Pen');
-    hashMap.set('Despacito', 'Luis Fonsi');
-    hashMap.set('Bailando', 'Enrique Iglesias');
-    hashMap.set('Dura', 'Daddy Yankee');
-    hashMap.set('Lean On', 'Major Lazer');
-    hashMap.set('Hello', 'Adele');
-    hashMap.set('All About That Bass', 'Meghan Trainor');
-    hashMap.set('Wake Me Up', 'Avicii');
-    hashMap.set('Brother', 'Avicii');
-    hashMap.set('Faded', 'Alan Walker');
-    hashMap.set('The Spectre', 'Alan Walker');
-  });
-
-  it('should rehash after 12 items by default', () => {
-    expect(hashMap.getLoadFactor()).toBe(11 / 16);
-    expect(hashMap.buckets.length).toBe(16);
-    hashMap.set('Alone', 'Alan Walker');
-    expect(hashMap.getLoadFactor()).toBe(0.75);
-
-    hashMap.set('Levels', 'Avicii');
-
-    expect(hashMap.getLoadFactor()).toBe(13 / 32);
-    expect(hashMap.buckets.length).toBe(32);
-
-    expect(hashMap.get('Dura')).toBe('Daddy Yankee');
-    expect(hashMap.get('Bailando')).toBe('Enrique Iglesias');
-    expect(hashMap.get('Levels')).toBe('Avicii');
+  describe('#rehash', () => {
+    let hashMap;
+
+    beforeEach(() => {
+      hashMap = new HashMap();
+
+      hashMap.set('Pineapple', 'Pen Pineapple Apple Pen');
+      hashMap.set('Despacito', 'Luis Fonsi');
+      hashMap.set('Bailando', 'Enrique Iglesias');
+      hashMap.set('Dura', 'Daddy Yankee');
+      hashMap.set('Lean On', 'Major Lazer');
+      hashMap.set('Hello', 'Adele');
+      hashMap.set('All About That Bass', 'Meghan Trainor');
+      hashMap.set('Wake Me Up', 'Avicii');
+      hashMap.set('Brother', 'Avicii');
+      hashMap.set('Faded', 'Alan Walker');
+      hashMap.set('The Spectre', 'Alan Walker');
+    });
+
+    it('should rehash after 12 items by default', () => {
+      expect(hashMap.getLoadFactor()).toBe(11 / 16);
+      expect(hashMap.buckets.length).toBe(16);
+      hashMap.set('Alone', 'Alan Walker');
+      expect(hashMap.getLoadFactor()).toBe(0.75);
+
+      hashMap.set('Levels', 'Avicii');
+
+      expect(hashMap.getLoadFactor()).toBe(13 / 32);
+      expect(hashMap.buckets.length).toBe(32);
+
+      expect(hashMap.get('Dura')).toBe('Daddy Yankee');
+      expect(hashMap.get('Bailando')).toBe('Enrique Iglesias');
+      expect(hashMap.get('Levels')).toBe('Avicii');
+    });
   });
 });

src/data-structures/hash-maps/hashmap.perf.js

@@ -1,12 +1,14 @@
+// nodemon src/data-structures/hash-maps/hashmap.perf.js
 const HashTable = require('./hash-map');
 // const asciichart = require ('asciichart');
 // const { Series } = require('pandas-js');
+const Series = require('./stats');
 
 const dict = new HashTable();
 
 const keys = [];
 const values = [];
-for (let i = 0; i < 1e1; i++) {
+for (let i = 0; i < 1e5; i++) {
   keys.push(makeid());
   values.push(parseInt(Math.random()*1e2));
 }
@@ -29,18 +31,11 @@ const usage = dict.buckets.reduce((array, el, index) => {
   return array;
 }, Array(dict.buckets.length).fill(0));
 
-console.log(usage, dict.buckets);
+// console.log(usage, dict.buckets);
 
-const ds = new Series(usage);
-console.log({
-  length: usage.length,
-  max: Math.max.apply(null, usage),
-  min: Math.min.apply(null, usage),
-  count: ds.length,
-  mean: ds.mean(),
-  std: ds.std(),
-  variance: ds.variance(),
-});
+const s = new Series(usage);
+console.log('Bucket array distribution', s.describe());
+console.log('Bucket array frequencies', s.frequencies);
 
 // console.log({elementsByBucket});
 // console.log (asciichart.plot(elementsByBucket));
@@ -49,9 +44,9 @@ console.log({
 
 function makeid() {
   const text = [];
-  const possible = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789ñåé😁🖖1234567890!@#$%^&*()_+{}';<>?";
+  const possible = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789ñåé😁🖖1234567890!@#$%^&*()_+{}';<>? ";
 
-  for (let i = 0; i < 5; i++) {
+  for (let i = 0; i < parseInt((Math.random() * 20) + 1, 10); i += 1) {
     text.push(possible.charAt(Math.floor(Math.random() * possible.length)));
   }