Struct Reconciled 

pub struct Reconciled {
    pub term_rows: FxHashMap<Box<str>, Vec<usize>>,
    pub names: FxHashMap<Box<str>, Box<str>>,
    pub universe: Vec<usize>,
    pub n_genes_total: usize,
    pub n_genes_matched: usize,
    pub n_terms_kept: usize,
    pub n_terms_total: usize,
}

Gene-sets reconciled against an expression dictionary: term → matched row indices, plus coverage statistics. This is the sparse membership the enrichment scorer consumes (term → its rows in the gene profile).

Fields

term_rows: FxHashMap<Box<str>, Vec<usize>>

term id → matched dictionary row indices (deduped, ascending); only terms retaining ≥ the requested minimum members are kept.

names: FxHashMap<Box<str>, Box<str>>

term id → display name (carried through from the source).

universe: Vec<usize>

All matched dictionary rows that carry ≥1 annotation (the enrichment background “universe”), ascending. Drawn from every source term — NOT only the size-windowed term_rows — so the hypergeometric background is the full annotated set.

n_genes_total: usize

distinct gene keys in the source.

n_genes_matched: usize

gene keys that matched at least one dictionary row.

n_terms_kept: usize

terms retained (≥ min members after reconciliation).

n_terms_total: usize

terms in the source before the min-member filter.

Implementations

impl Reconciled

pub fn match_frac(&self) -> f32

Fraction of source genes matched to the dictionary.

pub fn n_memberships(&self) -> usize

Total membership entries across retained terms.

pub fn log_coverage(&self)

Log a one-line coverage summary — info, or warn when overlap is thin (a near-empty map silently yields no enrichment).

pub fn ensure_coverage(&self, min_frac: f32, min_terms: usize) -> Result<()>

Fail when coverage is too thin to produce meaningful enrichment.