Struct Mass 

pub struct Mass<B: Postings> { /* private fields */ }

A Basic Belief Assignment: mass distributed over focal sets.

Invariants from §4.1: m(∅) = 0, and the masses sum to 1. Both are checked on construction rather than assumed, because a mass function that does not sum to 1 produces belief values that look plausible and are meaningless.

Implementations

impl<B: Postings> Mass<B>

pub fn new(focals: Vec<(B, f64)>) -> Result<Self, EvidenceError>

Build a mass function, checking the §4.1 invariants.

pub fn certain(set: B) -> Result<Self, EvidenceError>

All mass on one set — a single source asserting one possibility with full confidence.

pub fn vacuous(frame: B) -> Result<Self, EvidenceError>

All mass on the frame itself: the honest representation of knowing nothing. This is the state a point-estimate probability cannot express, and combining it with anything returns that thing unchanged.

pub fn focals(&self) -> &[(B, f64)]

pub fn belief(&self, a: &B) -> f64

Bel(A) = Σ_{B ⊆ A} m(B) — mass that commits entirely to A. The lower bound.

pub fn plausibility(&self, a: &B) -> f64

Pl(A) = Σ_{B ∩ A ≠ ∅} m(B) — mass not ruling A out. The upper bound.

pub fn interval(&self, a: &B) -> (f64, f64)

The evidential bound [Bel(A), Pl(A)].

pub fn ignorance(&self, a: &B) -> f64

How much of the interval is pure ignorance: Pl − Bel.

Trait Implementations

impl<B: Clone + Postings> Clone for Mass<B>

fn clone(&self) -> Self

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl<B: Debug + Postings> Debug for Mass<B>

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.