# almost-equal

**Repository Path**: ArkTSCentralRepository/almost-equal

## Basic Information

- **Project Name**: almost-equal
- **Description**: Test if two floats are almost equal
- **Primary Language**: Unknown
- **License**: MIT
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2026-04-01
- **Last Updated**: 2026-04-01

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# almost-equal
基于[almost-equal](https://www.npmjs.com/package/almost-equal)原库1.1.0版本进行适配

> Checks when two floats are almost equal.

## Install
```sh
ohpm install almost-equal
```

## Usage
```javascript
import AlmostEqual from "almost-equal";

const almostEqual = AlmostEqual.almostEqual;

let a = 100;
let b = 100 + 1e-12;

//Check if a == b up to float precision
console.log(almostEqual(a, b, AlmostEqual.FLT_EPSILON, AlmostEqual.FLT_EPSILON))

//Check if a == b up to double precision
console.log(almostEqual(a, b, AlmostEqual.DBL_EPSILON, AlmostEqual.DBL_EPSILON))
```

### `almostEqual(a, b[, absoluteTolerance [, relativeTolerance]])`
Checks if two floats are within the given tolerances of one another using the formula:

    |a - b| < max(absoluteTolerance, min(|a|, |b|) * relativeTolerance)

* `a` and `b` are the two numbers to comapre
* `absoluteTolerance` is a fixed minimal tolerance (set to 0 to ignore)
* `relativeTolerance` is a tolerance that scales with a/b (set to 0 to ignore)

**Returns** `true` if `a` and `b` are approximately equal.

If tolerance argument is omitted, `almostEqual.DBL_EPSILON` value is used by default.

### `almostEqual.FLT_EPSILON`
Floating point (32-bit) epsilon

### `almostEqual.DBL_EPSILON`
Double precision (64-bit) epsilon

Credits
=======
(c) 2013 Mikola Lysenko. MIT License