Struct TreePoisson 

pub struct TreePoisson {
    pub opening_angle: f64,
    pub domain: Domain,
    pub softening: f64,
}

Barnes-Hut tree code Poisson solver.

For each grid point the potential is computed via an O(N log N) tree walk with configurable opening angle θ. Smaller θ gives higher accuracy (θ → 0 recovers direct summation) at the cost of more work.

Fields

opening_angle: f64

Opening angle criterion θ for the multipole approximation.

domain: Domain

Computational domain.

softening: f64

Plummer softening length (prevents 1/r singularity).

Implementations

impl TreePoisson

pub fn new(domain: Domain, opening_angle: f64) -> Self

Create a new Barnes-Hut Poisson solver.

opening_angle controls accuracy vs speed (typical values 0.3–1.0). Softening is set to the average cell size.

Trait Implementations

impl PoissonSolver for TreePoisson

fn solve(&self, density: &DensityField, g: f64) -> PotentialField

Solve ∇²Φ = 4πGρ and return the potential field.

fn compute_acceleration(&self, potential: &PotentialField) -> AccelerationField

Compute the gravitational acceleration g = −∇Φ via spectral differentiation or finite differences.