This crate is designed to make it easy to work with positions and vectors that have meaning as geographic entities.

Documentation

Place this in your Cargo.toml

nav-types = "0.5.2"

and use it to calculate vectors and position:

extern crate nav_types;

use nav_types::WGS84;

let pos_a = WGS84::from_degrees_and_meters(36.12, -86.67, 0.0);
let pos_b = WGS84::from_degrees_and_meters(33.94, -118.40, 0.0);

println!("Distance between a and b: {:.2}m", a.distance(&b));

All position formats can work with vectors as long as the vectors are defined in some coordinate system.

use nav_types::{WGS84, ENU};

let pos_a = WGS84::from_degrees_and_meters(36.12, -86.67, 0.0);

let vec = ENU::new(0.0, 0.0, 10.0);
let pos_a_10m_up = pos_a + vec;

// Or with `NED` vector

let ned_vec = NED::new(0.0, 0.0, -10.0);
let pos_a_10m_up_2 = pos_a + ned_vec;

The source is based on nalgebra and some methods are only available if importing traits from nalgebra.

Currently the only way to calculate vectors between latitude and longitude positions is to convert to ECEF format and calculate the difference there. This conversion happens behind the scenes and for this reason could be a source of some surprise. It is therefor advised to try and use ECEF for as long as possible only converting to and from at the beginning and end.

On my laptop (absolute numbers will differ on different machines, but relative differences should be similar)

running 10 tests
test ecef::add_vector    ... bench:       1,321 ns/iter (+/- 21)
test ecef::difference    ... bench:       1,306 ns/iter (+/- 16)
test ecef::from_nvector  ... bench:          21 ns/iter (+/- 1)
test ecef::from_wgs84    ... bench:         492 ns/iter (+/- 23)
test nvector::add_vector ... bench:       1,486 ns/iter (+/- 61)
test nvector::difference ... bench:       1,352 ns/iter (+/- 230)
test nvector::from_ecef  ... bench:         144 ns/iter (+/- 3)
test nvector::from_wgs84 ... bench:         382 ns/iter (+/- 16)
test wgs84::add_vector   ... bench:       2,154 ns/iter (+/- 302)
test wgs84::difference   ... bench:       2,305 ns/iter (+/- 304)