memoizerific

What is memoizerific?

memoizerific is a JavaScript library that provides a simple and efficient way to memoize functions. Memoization is a technique used to speed up function execution by caching the results of expensive function calls and returning the cached result when the same inputs occur again.

What are memoizerific's main functionalities?

Basic Memoization

This feature allows you to memoize a function with a specified cache size. In this example, the function adds two numbers and caches the result for up to 100 different input combinations.

const memoizerific = require('memoizerific');

const memoizedFunction = memoizerific(100)((a, b) => a + b);

console.log(memoizedFunction(1, 2)); // 3
console.log(memoizedFunction(1, 2)); // 3 (cached result)

Cache Size Limitation

This feature demonstrates how the cache size limitation works. When the cache exceeds the specified size, the oldest cached result is evicted.

const memoizerific = require('memoizerific');

const memoizedFunction = memoizerific(2)((a, b) => a + b);

memoizedFunction(1, 2); // Cache: {(1, 2): 3}
memoizedFunction(2, 3); // Cache: {(1, 2): 3, (2, 3): 5}
memoizedFunction(3, 4); // Cache: {(2, 3): 5, (3, 4): 7} (1, 2) is evicted

Function Argument Handling

This feature shows how memoizerific can handle functions with variable numbers of arguments. The function sums all its arguments and caches the result.

const memoizerific = require('memoizerific');

const memoizedFunction = memoizerific(100)((...args) => args.reduce((sum, val) => sum + val, 0));

console.log(memoizedFunction(1, 2, 3)); // 6
console.log(memoizedFunction(1, 2, 3)); // 6 (cached result)

Other packages similar to memoizerific

lodash.memoize

lodash.memoize is a function from the Lodash library that provides similar memoization capabilities. It allows you to memoize functions and customize the cache key resolver. Compared to memoizerific, lodash.memoize is part of a larger utility library and may offer more flexibility in terms of cache key customization.

fast-memoize

fast-memoize is a lightweight and fast memoization library. It focuses on performance and simplicity, making it a good alternative to memoizerific for scenarios where speed is critical. However, it may not offer as many features as memoizerific, such as cache size limitation.

memoizee

memoizee is a comprehensive memoization library that offers a wide range of features, including cache size limitation, cache expiration, and support for various cache storage mechanisms. It is more feature-rich compared to memoizerific but may be more complex to use.

README

Memoizerific.js

Fastest (see benchmarks), smallest (923b min/gzip), most-efficient, dependency-free, JavaScript memoization lib to memoize JS functions. Fully supports multiple complex object arguments. Implements LRU (least recently used) cache to maintain only the most recent results.

For the browser and nodejs.

Memoization is the process of caching function results, so that they can be returned cheaply, without re-running the function when it is called again with the same arguments.

Install

npm install memoizerific --save

Use

var memoizerific = require('memoizerific');

var myExpensiveFunctionMemoized = memoizerific(50)(function(arg1, arg2, arg3) {
    // so many long expensive calls in here
});

myExpensiveFunction(1, 2, 3); // damn, that took looooong to process
myExpensiveFunction(1, 2, 3); // wow, that one was instant!
myExpensiveFunction(2, 3, 4); // expensive again :(
myExpensiveFunction(2, 3, 4); // woah, this one was dirt cheap, I'll take 2!

Options

There is one option available:

limit: the max number of results to cache.

memoizerific(limit)(fn);

memoizerific(1)(function(){}); // memoize 1 result
memoizerific(10000)(function(){}); // memoize 10,000 results
memoizerific(0)(function(){}); // memoize infinity results (not recommended)

The cache works using LRU logic, purging the least recently used results when the limit is reached.

// memoize 1 result
var myMemoized = memoizerific(1)(function(arg1, arg2, arg3, arg4) {});

myMemoized(1, 2, 3, 'a'); // function runs
myMemoized(1, 2, 3, 'a'); // cached result is returned
myMemoized(1, 2, 3, 'X'); // function runs again, new result is cached, old cached result is purged
myMemoized(1, 2, 3, 'X'); // new cached result is returned
myMemoized(1, 2, 3, 'a'); // function runs again...

Comparison

There are many memoization libs available for JavaScript. Some of them have specialized use-cases, such as memoizing file-system access, or server async requests. While others, such as this one, tackle the more general case of memoizing standard synchronous functions. Following are the minimum criteria I look for in a production-worthy memoization solution:

• Support for multiple arguments: One argument memoizers start to fall short quickly when solving real problems.
• Support for complex arguments: Including large arrays, complex objects, arrays-within-objects, objects-within-arrays, etc. (not just primitives like strings or numbers).
• Dynamic Cache: A cache that grows unimpeded will quickly become a memory leak and source of bugs.
• Consistent performance profile: Many libs perform well within certain parameters, but start to fail wildly in others, usually when a large cache is chosen, or many arguments are used. It is important that performance degrades predictably and linearly as the environment becomes less favorable, to avoid nasty surprises.

Using this list, we can narrow down the field of possible candidates by quite a bit. The popular lodash memoize, for example, only supports one argument out of the box and has no cache-size control. Others support multiple complex arguments, but do not manage the cache-size either:

• :heavy_multiplication_x: Memoizejs (@addyosmani)

• :heavy_multiplication_x: Memoize-strict (@jshanson7)

• :heavy_multiplication_x: Deep-memoize (@rjmk)

• :heavy_multiplication_x: Mem (@sindresorhus)

Three libs with reasonable traction, seem to meet the basic criteria:

• Memoizee (@medikoo)
• LRU-Memoize (@erikras)
• LRU-Memoize (@neilk)

After some quick testing, however, we found the library by @neilk to be producing incorrect results, leaving only two viable candidates. Time to test performance.

Benchmarks

This library is intended for real-world use-cases, and is therefore benchmarked against other libraries using large, complex, real-world data. We don't need any more fibonacci solving tools. Example arguments look like this:

myMemoized(
    { a: 1, b: [{ c: 2, d: { e: 3 }}] }, // 1st argument
    [{ x: 'x', q: 'q', }, { b: 8, c: 9 }, { b: 2, c: [{x: 5, y: 3}, {x: 2, y: 7}] }, { b: 8, c: 9 }, { b: 8, c: 9 }], // 2nd argument
    { z: 'z' }, // 3rd argument
    ... // 4th, 5th... argument
);

We generated sets of thousands of random argument combinations of varying variance (to increase and decrease cache hits and misses) and fed them each the same ones. For the full details and source of the benchmarks see memoize-js-libs-benchmarks.

Data

Following is data from 5000 iterations for each test on firefox 44:

Cache Size	Num Args	Approx. Cache Hits (variance)	LRU-Memoize	Memoizee	Memoizerific	% Faster
10	2	99%	19ms	31ms	10ms	90%
10	2	62%	212ms	319ms	172ms	23%
10	2	7%	579ms	617ms	518ms	12%
100	2	99%	137ms	37ms	20ms	85%
100	2	69%	696ms	245ms	161ms	52%
100	2	10%	1,057ms	649ms	527ms	23%
500	4	95%	476ms	67ms	62ms	8%
500	4	36%	2,642ms	703ms	594ms	18%
500	4	11%	3,619ms	880ms	725ms	21%
1000	8	95%	1,009ms	52ms	65ms	25%
1000	8	14%	10,477ms	659ms	635ms	4%
1000	8	1%	6,943ms	1,501ms	1,466ms	2%

Results

The results from these tests are interesting. While LRU-Memoize performed quite well when there were few arguments, and lots of cache hits, it quickly started to fall apart as the environment became more challenging. Once we got to 4 arguments it became 5x-10x-20x slower than the other two, and began to hit severe levels that could potentially cause real-life problems. I would not recommend it for heavy production use. Memoizee came in a solid second place, being outperformed by Memoizerific by an average of 31%. While not insignificant, in many real-world scenarios this will not make a huge difference, unless the memoized function is being called repetitively in a loop, or through heavy recursion. What counts, is that it degraded nicely, and stayed within reasonable sub 1s levels. It is a sturdy library and I would be comfortable recommending it for production use. Memoizerific was clearly the big performance winner. It was built with complex real-world use in mind, and I would biasedly recommend it.