@float-toolkit/core
v2.1.0
Published
Float Toolkit is an NPM package for working with floating-point numbers.
Downloads
92
Maintainers
Readme
Float Toolkit
Float Toolkit is a lightweight, dependency-less NPM package that allows you to work with floats (floating-point numbers, or numbers with decimals).
You can round floats down to any number of digits, and perform accurate math operations (addition, subtraction, multiplication and division) with them.
Features
- Accurate math operations
- Round floats to any number of digits
- No dependencies
- TypeScript type declarations
- Compatible with browsers and Node.js
- > 95% code coverage
Get started
Installation
To add Float Toolkit to your project, run this command:
npm install @float-toolkit/core
Usage
The package export is a class called FloatToolkit
. Once you import it, you can create multiple FloatToolkit
instances.
import FloatToolkit from "@float-toolkit/core";
const ft = new FloatToolkit();
console.log(0.1 + 0.2); // 0.30000000000000004
console.log(ft.round(0.1 + 0.2)); // 0.3
console.log(ft.add([0.1, 0.2])); // 0.3
TypeScript
Float Toolkit is written in TypeScript. As such, you will have full type checking in your TypeScript
projects. The type definitions used by the package are exported on their own or through the FloatToolkit
namespace.
import FloatToolkit from "@float-toolkit/core";
const createFT = (options: FloatToolkit.Options) => new FloatToolkit(5, options);
Why use Float Toolkit?
In JavaScript, certain math operations with floats can return values with slight imperfections. This can result in unexpected behavior.
const result = 0.1 + 0.2; // 0.30000000000000004
console.log(result === 0.3); // false
Often, developers address this issue by accepting a certain range of floats that includes the exact one they are looking for, like so:
console.log(result >= 0.299 && result <= 0.301); // true
This is not ideal, since this condition also becomes true if another value falls in the same range, even if it's not the expected
value. For example, a value of 0.3005
would also pass this condition.
Another options is to round the number using Vanilla JS functions.
const roundedResult = Number(result.toFixed(1)); // 0.3
console.log(roundedResult === 0.3); // true
This alternative fixes the main problem the previous solution has, but it also increases the amount of boilerplate code, since it
requires you to add Number(variable.toFixed(precision))
for every single number.
The Float Toolkit syntax is much cleaner and more concise.
const result = ft.add([0.1, 0.2]); // 0.3
console.log(result === 0.3); // true
Support
Need help using Float Toolkit? Don't hesitate to reach out on GitHub Discussions!
Links
Child packages
Contributing
Before creating an issue, please consider the following:
- Refer to the documentation to make sure the error is actually a bug and not a mistake of your own.
- Make sure the issue hasn't already been reported or suggested.
- After following these steps, you can file an issue using one of our templates. Please make sure to follow our Code of Conduct.
- If you wish to submit a pull request alongside your issue, please follow our contribution guidelines.