Struct LabelObservation 

pub struct LabelObservation {
    pub types: BTreeSet<GffType>,
    pub files: usize,
    pub files_with_value: usize,
}

What a corpus was seen to carry under one label.

Fields

types: BTreeSet<GffType>

Types the label was seen holding. A label may hold different types in different files, so this is a set rather than one value.

files: usize

How many files carried the label.

This is what turns a label list into a priority order. A label in every file and a label in one are very different propositions, and without the count they read identically.

files_with_value: usize

How many files carried the label holding something other than its zero value.

Prevalence alone misranks: a label present in every file and empty in all of them carries no information, while a label in a handful of files holding real values does. Ranking needs both.

This orders work; it does not decide what to model

Neither this count nor the file count predicts whether a field is worth modelling, and both have been wrong in both directions. Only what the loader does with a field settles it; these counts are for deciding what to look at first. docs/src/architecture.md works through the three ways the counts mislead, with a real field behind each.

Trait Implementations

impl Clone for LabelObservation

fn clone(&self) -> LabelObservation

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for LabelObservation

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

Formats the value using the given formatter.

impl Default for LabelObservation

fn default() -> LabelObservation

Returns the “default value” for a type.

impl PartialEq for LabelObservation

fn eq(&self, other: &LabelObservation) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl Eq for LabelObservation

impl StructuralPartialEq for LabelObservation