multi-map

Map or reduce operators that take a set of iteratees for each step.

Table of Contents

• Introduction
• Peformance
• Usage

Introduction

If you must be as effiecient as possible and you're dealing with multiple map operations, effectively:

const afterOneFn = map(data, fn1);
const afterTwoFn = map(afterOneFn, fn2);
...
return map(afterNFn, fnN);

You can use multiMap instead to get some pretty marginal gains:

return multiMap(data, fn1, fn2, ..., fnN);

The idea is simply that instead of calling a map operation over a data set n times, you just call it once and apply your functions to each iteratee.

Home

Performance

There's some testing in here that is hardly scientific (there's some bias to whichever one is second), where we are running some functions over a large data set (cities around the world);

You can test it yourself by running yarn test, but the output is roughly as follows:

| Method | DataSet Size | Number of Runs | p50 (in milliseconds) | p99 (in milliseconds) | average (in milliseconds) | | ---------- | ------------ | -------------- | --------------------- | --------------------- | ------------------------- | | map | 148040 | 100 | 171 | 3003 | 283.37 | | multiMap | 148040 | 100 | 145 | 569 | 148.39 |

As you can see, the gains are mostly marginal. But if you must "trim that fat..."

Home

Usage

Installation is easy.

Via npm:

npm i @ezweave/multi-map

Via yarn:

yarn add @ezweave/multi-map

There's two variants available, multiMap and multiMapFP. They are functionaly the same, only the FP variant changes the -arity and curries out the input.

E.g. these are equivalent:

import { multiMap, multiMapFP } from "multi-map";

multiMap(data, fn1, fn2, fn3);
multiMapFP(fn1, fn2, fn3)(data);

Home