approx-fix v1.0.2

⭐️ What

Align number, reduce the calculation error.

Get 0.1 + 0.2 as 0.3

It seems like toFixed, but it will only shift while delta is under 1e-8.

📦 Getting Started

Installation

npm install approx-fix
# or
yarn add approx-fix

Usage

approxFix(value, [precision = 2])

precision == 2 is similar to toFixed(2),
it means the number of digits to appear after the decimal point.

// * there are two functions
// * support cjs and esm
// * support object recursion way

// * ---------------- cjs
const { approxFix, approxFixNum } = require('approx-fix');
approxFix([0.1 + 0.2]); // => [0.3]

// * ---------------- esm
// * default is approxFix
import approxFix from 'approx-fix';
approxFix(0.1 + 0.2); // => 0.3
approxFix(0.12301, 3); // => 0.12301
approxFix(0.1230000001, 3); // => 0.123

💡 Why

IEEE754 problem.

0.2 + 0.1 === 0.30000000000000004;
Math.sin(Math.PI) === 1.2246467991473532e-16;

And I want to really get the normal result,
Primarily for the the trigonometric functions,

So I write the lib to save some time.

Well, it's not prefect or totally accurate,
it's just a quick simple auto fix which works for some value.
But it's good enough for my usage.

📖 Description

there are two methods

• approxFixNum
• approxFix

approxFixNum only supports for number values.

approxFix is a wrapper of approxFixNum,
which will support any type of data you pass auto-recursively,
and call approxFixNum while meet number.

(real Number. number string is not well tested and supported)

Rules

While the delta between your number and the 'toFixed' number of it, is under 1e-8,
It will return the fixed number.

Else it should return the same value.

⌨️ Contribution

# git clone and cd into it
git clone https://github.com/seognil-lab/approx-fix

# npm command
npm i
npm run test:watch

📜 References

https://0.30000000000000004.com/

🕗 TODO