`Bearing.bearing(p: &Point)` computes wrong bearing about geo HOT 4 CLOSED

I think I understand what you're saying, but I'm not able to reconcile it with the code base.

Could you provide an example input / output that aligns with your expectations like:

let point_1 = point(x: ???, y: ???);
let point_2 = point(x: ???, y: ???);
let bearing = point_1.bearing(point_2);
assert_eq!(bearing, ???);

from geo.

Sorry for not coming back to this sooner. After some more testing, I concluded that my hypothesis of N=90 and E=0 is wrong. Here is an example test:

#[test]
fn a_different_angle() {
    let p_1 = point!(x: 52.323434, y: 4.822623);
    let p_2 = point!(x: 52.327790, y: 4.851333);
    let b_1 = p_1.haversine_bearing(p_2);
    assert_relative_eq!(b_1, 76.0, epsilon = 1.0e-6);
}

with this output (geo 0.24.1):

left  = 8.5969044325273
right = 76.0

According to this online calculator, the bearing is 76 degrees (no more significant digits unfortunately). When looking at a map, it's at least clear to me that it shouldn't be 8.6 degrees. Not sure what's going on here, since the other tests pass.

from geo.

Assuming that you are operating in the Netherlands, it looks like you've swapped your latitude (y) and longitude (x).

Feel free to follow up if I'm mistaken.

from geo.

Thanks, that clears it up. It's confusing that lat/lon and x/y are not a one to one mapping. Maybe this is a useful addition to the documentation?

from geo.