Trait zero_sum::analysis::Evaluation

pub trait Evaluation: Sized + Clone + Copy + Display + Add<Output=Self> + Sub<Output=Self> + Mul<Output=Self> + Neg<Output=Self> + Div<Output=Self> + PartialEq + PartialOrd {
    fn null() -> Self;
    fn epsilon() -> Self;
    fn win() -> Self;
    fn max() -> Self;
    fn is_win(&self) -> bool;
}

An evaluation type.

This is usually a tuple around a signed numeric type.

Example

There is a helper macro to facilitate the implementation of tuple structs:

#[macro_use]
extern crate zero_sum;
use std::i32;
use std::ops::{Add, Div, Mul, Neg, Sub};

#[derive(Clone, Copy, PartialEq, PartialOrd)]
struct Eval(i32);

prepare_evaluation_tuple!(Eval); // impl Add, Div, Mul, Neg, Sub, and Display

impl Evaluation for Eval {
    fn null() -> Eval { Eval(0) }
    fn epsilon() -> Eval { Eval(1) }
    fn win() -> Eval { Eval(100000) }
    fn max() -> Eval { Eval(i32::MAX) }
    fn is_win(&self) -> bool { self.0.abs() > 99000 }
}

Required Methods

fn null() -> Self

An empty, or zero evaluation.

fn epsilon() -> Self

The smallest step to consider.

fn win() -> Self

The base value of a win. The evaluation system may add or subtract to it in in order to promote it or discourage it in favor of others in the search.

fn max() -> Self

The maximum value representable. This must be safely negatable.

fn is_win(&self) -> bool

Returns true if this evaluation contains a win. This is usually a check to see if the absolute value is above a certain threshold.

Implementors