Struct SolverConfig

pub struct SolverConfig {
    pub z_start: f64,
    pub z_end: f64,
    pub dy_max: f64,
    pub dz_min: f64,
    pub dtau_max: f64,
    pub nc_z_min: f64,
    pub dtau_max_photon_source: f64,
    pub max_newton_iter: usize,
    pub cn_dcbr: bool,
}

Tunable solver parameters.

Defaults are production-quality and match the values used for all paper runs. The fields that most users will want to change are Self::z_start, Self::z_end, and Self::dtau_max.

Fields

z_start: f64

Upper redshift at which integration begins (solver evolves from z_start down to z_end). Must satisfy z_start > z_end.

z_end: f64

Lower redshift at which integration ends.

dy_max: f64

Maximum fractional change in ln(1+z) per step (limits step size for slow-varying sources). Default 0.02. Historical value 0.005 was calibrated for photon-injection bursts; for smooth (continuous) heat injection scenarios 0.02 gives the same μ/y to 4 significant figures with 5–6× fewer steps, since dtau_max takes over as the binding constraint at z ≲ 1.5×10⁶.

dz_min: f64

Minimum allowed step size in z; smaller values trigger a warning. Default 1e-6.

dtau_max: f64

Maximum Compton optical depth per step. With exact exponential DC/BR (unconditionally stable), this primarily limits Kompaneets CN accuracy. Default 10.0 matches the CLI default (the value used for all paper runs). Raise to ~50 for exploratory runs where modest accuracy loss is acceptable.

nc_z_min: f64

Minimum redshift for number-conserving T-shift subtraction. Default 5e4: only subtract at z > nc_z_min where DC/BR is significant. Set to 0.0 to subtract at all redshifts (CosmoTherm-like behavior).

dtau_max_photon_source: f64

Maximum dtau per step during active photon source injection. Limits the Kompaneets Δn² nonlinearity at the injection spike, which causes timestep-dependent wake formation at intermediate x. Default 1.0 (production quality, ~1% from converged). Set to 10.0 for exploratory/fast runs (~3% from converged).

max_newton_iter: usize

Maximum number of Newton iterations per Kompaneets step. Default 10. Increase for extreme injection parameters.

cn_dcbr: bool

Use Crank-Nicolson (instead of backward Euler) for DC/BR. Second-order in time, matching CosmoTherm’s scheme. May be less stable at very low x where DC/BR rates diverge.

Implementations

impl SolverConfig

pub fn validate(&self) -> Result<(), String>

Validate solver configuration parameters.

Returns Err with a descriptive message if any parameter would cause the solver to fail or produce meaningless results.

Trait Implementations

impl Clone for SolverConfig

fn clone(&self) -> SolverConfig

Returns a copy of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for SolverConfig

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for SolverConfig

fn default() -> Self

Returns the “default value” for a type.