Struct Problem 

pub struct Problem<'a> {
    pub f: Option<&'a dyn Fn(&[f64]) -> f64>,
    pub residuals: Option<&'a dyn Fn(&[f64]) -> Vec<f64>>,
    pub gradient: Option<&'a dyn Fn(&[f64]) -> Vec<f64>>,
    pub jacobian: Option<&'a dyn Fn(&[f64]) -> Vec<Vec<f64>>>,
    pub x0: Vec<f64>,
    pub bounds: Option<Vec<(f64, f64)>>,
    pub seed: u64,
}

An optimization problem: a scalar objective or a residual vector (least squares), whatever derivatives are available, a start, and optional bounds.

Fields

f: Option<&'a dyn Fn(&[f64]) -> f64>

residuals: Option<&'a dyn Fn(&[f64]) -> Vec<f64>>

gradient: Option<&'a dyn Fn(&[f64]) -> Vec<f64>>

jacobian: Option<&'a dyn Fn(&[f64]) -> Vec<Vec<f64>>>

x0: Vec<f64>

bounds: Option<Vec<(f64, f64)>>

Per-parameter (lo, hi) box, required by differential evolution.

seed: u64

RNG seed for stochastic methods (differential evolution).

Implementations

impl<'a> Problem<'a>

pub fn scalar(f: &'a dyn Fn(&[f64]) -> f64, x0: Vec<f64>) -> Self

A scalar-objective problem.

pub fn least_squares( residuals: &'a dyn Fn(&[f64]) -> Vec<f64>, x0: Vec<f64>, ) -> Self

A least-squares problem min sum r_i(x)^2.

pub fn with_gradient(self, gradient: &'a dyn Fn(&[f64]) -> Vec<f64>) -> Self

pub fn with_jacobian( self, jacobian: &'a dyn Fn(&[f64]) -> Vec<Vec<f64>>, ) -> Self

pub fn with_bounds(self, bounds: Vec<(f64, f64)>) -> Self

pub fn with_seed(self, seed: u64) -> Self