// Copyright 2018 Stefan Kroboth
//
// Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
// http://apache.org/licenses/LICENSE-2.0> or the MIT license <LICENSE-MIT or
// http://opensource.org/licenses/MIT>, at your option. This file may not be
// copied, modified, or distributed except according to those terms.
//! TODO DOCUMENTATION
//!
use std;
use ndarray::{Array1, Array2};
/// `ArgminOperator`
pub struct ArgminOperator<'a> {
/// Operator (for now a simple 2D matrix)
pub operator: &'a Array2<f64>,
/// y of Ax = y
pub y: &'a Array1<f64>,
/// Target cost function value for stopping criterions
pub target_cost: f64,
}
impl<'a> ArgminOperator<'a> {
/// Constructor
pub fn new(operator: &'a Array2<f64>, y: &'a Array1<f64>) -> Self {
ArgminOperator {
operator: operator,
y: y,
target_cost: std::f64::MIN,
}
}
/// Set target cost function value
///
/// If the optimization reaches this value, it will be stopped.
pub fn target_cost(&mut self, target_cost: f64) -> &mut Self {
self.target_cost = target_cost;
self
}
/// Forward application of the operator (A*x)
pub fn apply(&self, x: &Array1<f64>) -> Array1<f64> {
self.operator.dot(x)
}
/// Application of the transpose of the operator (A^T * x)
pub fn apply_transpose(&self, x: &Array1<f64>) -> Array1<f64> {
self.operator.t().dot(x)
}
}