frames

License · Docs

You want to move your robot from point A to point B. That's all you ever want to do, really. The question is, how do you model the multiple frames that exist in your environment?

Say you have the following:

In short — Frames solves for (x, y, θ) like so:

use frames::prelude::*;
use nalgebra::{Isometry2, Vector2};
use std::f32::consts::PI;

fn main() -> Result<(), FrameError> {
    let mut frames = Frames::new();

    let field = Frame::new("field");
    let robot = Frame::new("robot");

    frames.add_frame(field, Isometry2::new(Vector2::new(0., 0.), 0.))?;
    frames.add_frame(robot, Isometry2::new(Vector2::new(1., 1.), PI))?;

    let x = Point::new("x");
    frames.add_point_in_context(
        x,
        Isometry2::new(Vector2::new(7., 5.), PI),
        field,
    )?;

    assert_eq!(
        frames.get_point_in_context(x, robot)?,
        Isometry2::new(Vector2::new(6., 4.), 0.)
    );

    Ok(())
}

Features

• Efficient — optimized, no-nonsense calculations.
• Scalable — calculate in any dimensions.
• Type-safe — errors can be seen at compile-time.