solvr 0.2.0-beta.2

Advanced computing library for real-world problem solving - optimization, differential equations, interpolation, statistics, and more
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28

//! Levenberg-Marquardt algorithm trait for systems of nonlinear equations.
use crate::DType;

use crate::optimize::roots::{RootOptions, RootTensorResult};
use numr::error::Result;
use numr::runtime::Runtime;
use numr::tensor::Tensor;

/// Levenberg-Marquardt algorithm for systems of nonlinear equations.
///
/// A damped Newton method that interpolates between Newton's method and
/// gradient descent. More robust when initial guess is far from solution.
pub trait LevenbergMarquardtAlgorithms<R: Runtime<DType = DType>> {
    /// Solve F(x) = 0 using Levenberg-Marquardt algorithm.
    ///
    /// # Arguments
    /// * `f` - The function F: R^n -> R^n
    /// * `x0` - Initial guess
    /// * `options` - Solver options
    fn levenberg_marquardt<F>(
        &self,
        f: F,
        x0: &Tensor<R>,
        options: &RootOptions,
    ) -> Result<RootTensorResult<R>>
    where
        F: Fn(&Tensor<R>) -> Result<Tensor<R>>;
}