bevy_spatial

A bevy plugin to track your entities in spatial indices and query them.

NOTE: You will need to enable at least one of the features.

Currently implemented features: |Feature|Description| |-|-| |kdtree|KD-Tree for spatial lookups which is fully recreated. This is ideal for cases where most entities are moving.| |rstar|R*-Tree for spatial lookups which is updated or recreated based on a threshold of changed entities. Ideal when most entities are static. |

Quickstart using the kdtree feature:

```rust use bevy_spatial::{KDTreeAccess2D, KDTreePlugin2D, SpatialAccess};

[derive(Component)]

struct TrackedByKDTree;

fn main() { App::new() .addplugin(KDTreePlugin2D:: { ..default() }) .addsystem(use_neighbour); // ... }

type NNTree = KDTreeAccess2D; // type alias for brevity

fn useneighbour(tree: Res){ if let Some((pos, entity)) = tree.nearestneighbour(Vec2::ZERO) { // pos: Vec3 // do something with the nearest entity here, most likley you will want a query.get(entity) call } } ```

For more details on usage see Examples

compatible bevy versions

| bevy | bevy_spatial | | ---- | ------------ | | 0.9 | 0.4.0 | | 0.8 | 0.3.0 | | 0.8 | 0.2.1 | | 0.7 | 0.1 |

wasm caveats: the rayon acceleration for kdtree is disabled on wasm, making it a bit slower.

TODOs and Ideas

• benchmarks

• documentation

• Versions of the SpatialAccess functions which return Iterators instead of Vecs

• a feature for linfa_nn to use their abstractions over kdtree/balltree linear search

• a feature for https://github.com/InstantDomain/instant-distance