#interpolation #numeric #math #extrapolate

lerp

Generic linear interpolation and extrapolation, automatically implemented for a variety of numeric types

6 releases (breaking)

0.5.0 Jun 17, 2023
0.4.0 Jan 19, 2021
0.3.0 Oct 6, 2020
0.2.0 Aug 6, 2016
0.1.1 Aug 5, 2016

#1096 in Algorithms

Download history 6126/week @ 2026-02-19 9961/week @ 2026-02-26 10255/week @ 2026-03-05 10962/week @ 2026-03-12 11822/week @ 2026-03-19 20255/week @ 2026-03-26 11368/week @ 2026-04-02 13603/week @ 2026-04-09 12195/week @ 2026-04-16 12261/week @ 2026-04-23 10567/week @ 2026-04-30 11700/week @ 2026-05-07 10990/week @ 2026-05-14 12695/week @ 2026-05-21 11579/week @ 2026-05-28 15518/week @ 2026-06-04

52,293 downloads per month
Used in 18 crates (14 directly)

MIT license

12KB
127 lines

Lerp

Build Status

Linear interpolation and iteration, automatically implemented over most float-compatible types.

Just need to know what's halfway between three and five?

use lerp::Lerp;

assert_eq!(3.0.lerp(5.0, 0.5), 4.0);

Want to iterate across some points in that range?

// bring the trait into scope
use lerp::LerpIter;

// iterate and produce four items evenly spaced between 3.0 and 5.0
// note that the default, open iterator does not include both endpoints
// this makes chaining lerping iterators simpler
let items: Vec<_> = 3.0_f64.lerp_iter(5.0, 4).collect();
assert_eq!(vec![3.0, 3.5, 4.0, 4.5], items);

// closed iterators return both ends
assert_eq!(vec![3.0, 5.0], 3.0.lerp_iter_closed(5.0, 2).collect::<Vec<_>>());

Of course, the real benefit is that it's derivation is broad enough that it also covers types such as num::Complex<T>. If you have an array-processing library, and the arrays are T: Add<Output = T> + Mul<F: Float, Output = T>, it'll just work for them as well.

Deriving Lerp

As well as working for individual float values, the crate also provides a derive macro, available with the derive feature, which will be able to generate an implementation automatically.

This derive implementation will lerp each field of the struct independently and assumes a generic implementation of Lerp over Float types. If any of the fields is generic only over one of the float values (f32, f64) that can be specified by the #[lerp(f32)] or #[lerp(f64)] attributes respectively.

If you would like for the lerp implementation to ignore a field (or if it does not derive lerp) you can use the #[lerp(skip)] or #[lerp(ignore)] attributes which will produce the value, untouched from the left value.

Not all types are supported in this derive macro. See the github issue for discussion and more information.

[dependencies]
lerp = { version = "0.4", features = ["derive"] }
use lerp::Lerp;

#[derive(Lerp, PartialEq, Debug)]
struct Data {
    a: f64,
    b: f64
}

assert_eq!(
    Data { a: 0.0, b: 1.0 }.lerp(Data { a: 1.0, b: 0.0 }, 0.5),
    Data { a: 0.5, b: 0.5 }
);

More derive examples can be seen in the tests

Usage

[dependencies]
lerp = "0.4"

Documentation

Auto-built from Travis: https://coriolinus.github.io/lerp-rs/

Dependencies

~72–375KB