@redis/bloom

What is @redis/bloom?

@redis/bloom is an npm package that provides probabilistic data structures for Redis, such as Bloom filters, Cuckoo filters, Count-Min Sketch, and Top-K. These data structures are useful for efficiently handling large sets of data with minimal memory usage and are often used in scenarios like caching, deduplication, and approximate counting.

What are @redis/bloom's main functionalities?

Bloom Filter

This example demonstrates how to create a Bloom filter, add an item to it, and check for the existence of an item. Bloom filters are used to test whether an element is a member of a set with a possible false positive rate.

const { createClient } = require('@redis/bloom');

async function bloomFilterExample() {
  const client = createClient();
  await client.connect();

  await client.bf.reserve('myBloom', 0.01, 1000);
  await client.bf.add('myBloom', 'item1');
  const exists = await client.bf.exists('myBloom', 'item1');
  console.log(exists); // true

  await client.disconnect();
}

bloomFilterExample();

Cuckoo Filter

This example shows how to create a Cuckoo filter, add an item, and check for its existence. Cuckoo filters are similar to Bloom filters but support deletion of items.

const { createClient } = require('@redis/bloom');

async function cuckooFilterExample() {
  const client = createClient();
  await client.connect();

  await client.cf.reserve('myCuckoo', 1000);
  await client.cf.add('myCuckoo', 'item1');
  const exists = await client.cf.exists('myCuckoo', 'item1');
  console.log(exists); // true

  await client.disconnect();
}

cuckooFilterExample();

Count-Min Sketch

This example demonstrates how to initialize a Count-Min Sketch, increment the count of an item, and query the count. Count-Min Sketches are used for approximate frequency counting.

const { createClient } = require('@redis/bloom');

async function countMinSketchExample() {
  const client = createClient();
  await client.connect();

  await client.cms.initByDim('myCMS', 2000, 5);
  await client.cms.incrBy('myCMS', 'item1', 5);
  const count = await client.cms.query('myCMS', 'item1');
  console.log(count); // 5

  await client.disconnect();
}

countMinSketchExample();

Top-K

This example shows how to create a Top-K data structure, add items to it, and retrieve the top K items. Top-K is used to maintain a list of the most frequent items.

const { createClient } = require('@redis/bloom');

async function topKExample() {
  const client = createClient();
  await client.connect();

  await client.topk.reserve('myTopK', 3);
  await client.topk.add('myTopK', 'item1', 'item2', 'item3', 'item1');
  const list = await client.topk.list('myTopK');
  console.log(list); // ['item1', 'item2', 'item3']

  await client.disconnect();
}

topKExample();

Other packages similar to @redis/bloom

bloom-filter

The 'bloom-filter' package provides a simple implementation of Bloom filters in JavaScript. It is less feature-rich compared to @redis/bloom and does not integrate with Redis, making it suitable for in-memory operations only.

cuckoo-filter

The 'cuckoo-filter' package offers a JavaScript implementation of Cuckoo filters. Similar to 'bloom-filter', it does not integrate with Redis and is intended for in-memory use cases.

count-min-sketch

The 'count-min-sketch' package provides a JavaScript implementation of the Count-Min Sketch data structure. It is useful for approximate frequency counting but lacks the Redis integration provided by @redis/bloom.

README

@redis/bloom

This package provides support for the RedisBloom module, which adds additional probabilistic data structures to Redis. It extends the Node Redis client to include functions for each of the RediBloom commands.

To use these extra commands, your Redis server must have the RedisBloom module installed.

RedisBloom provides the following probabilistic data structures:

• Bloom Filter: for checking set membership with a high degree of certainty.
• Cuckoo Filter: for checking set membership with a high degree of certainty.
• Count-Min Sketch: Determine the frequency of events in a stream.
• Top-K: Maintain a list of k most frequently seen items.

For complete examples, see bloom-filter.js, cuckoo-filter.js, count-min-sketch.js and topk.js in the Node Redis examples folder.