pub fn forward_difference_hessian<P, V>(
problem: &P,
x: &V,
function_precision: f64,
fixed_step: Option<f64>,
) -> Result<<V as DenseMatrixFromFn>::Matrix, P::Error>where
P: CostFunction<Param = V, Output = f64> + MaybeSync,
V: Clone + VectorLen + VectorIndex + DenseMatrixFromFn + MaybeSync,
P::Error: MaybeSend,Expand description
Forward-difference Hessian (one-sided second differences).
Hᵢⱼ = (f(x+hᵢeᵢ+hⱼeⱼ) − f(x+hᵢeᵢ) − f(x+hⱼeⱼ) + f(x)) / hᵢhⱼ, with the
diagonal as the i = j case (f(x+2hᵢeᵢ)). Symmetric n × n by
construction; 1 + n + n(n+1)/2 cost evaluations. Returns Err if any
probe’s cost does.