Trait nlopt::MObjFn

pub trait MObjFn<U>: Fn(&mut [f64], &[f64], Option<&mut [f64]>, &mut U) { }

A trait representing a multi-objective function.

A multi-objective function takes the form of a closure f(result: &mut [f64], x: &[f64], gradient: Option<&mut [f64], user_data: &mut U) -> f64

• result - m-dimensional array to store the value f(x)
• x - n-dimensional array
• gradient - n×m-diconstraint array to store the gradient grad f(x). The n dimension of gradient is stored contiguously, so that df_i / dx_j is stored in gradient[i*n + j]. If gradient is Some(x), the user is required to return a valid gradient, otherwise the optimization will most likely fail.
• user_data - user defined data

Implementors

impl<T, U> MObjFn<U> for T where
    T: Fn(&mut [f64], &[f64], Option<&mut [f64]>, &mut U), 

Loading content...