Enum FunctionKind 

pub enum FunctionKind {
    TextMatch,
    FTSMatch,
    BayesianMatch,
    BayesianMatchWithPrior,
    KNNMatch,
    FuseLogOdds,
    PositiveEvidencePool,
    BayesianEvidenceFusion,
    GraphPagerank,
    GraphHits,
    GraphBetweenness,
    GraphTraverse,
    GraphNeighbors,
    MultiFieldMatch,
    StagedRetrieval,
    DeepPredict,
    UQAHighlight,
    UQAFacets,
    TraverseMatch,
    TemporalTraverse,
    RPQ,
    GraphCreate,
    GraphDrop,
    GraphExists,
    GraphLabelCreate,
    GraphLabelDrop,
    GraphAlter,
    GraphEdges,
    AttentionFusion,
    LearnedFusion,
    CalibratedVectorMatch,
    SparseThreshold,
    ScoreBM25,
    ScoreBayesianBM25,
    DeepLearn,
    Convolve,
    Pool,
    Flatten,
    Dense,
    Softmax,
    Layer,
    Model,
}

Variants

TextMatch

text_match(field, query_string) - BM25 text retrieval.

FTSMatch

field @@ query - full-text query-string parser over text and vector signals.

BayesianMatch

bayesian_match(field, query_string) - Bayesian BM25 retrieval.

BayesianMatchWithPrior

bayesian_match_with_prior(field, query, prior_field, mode) - Bayesian BM25 adjusted by a document-level external prior.

KNNMatch

knn_match(field, query_vector, k) - top-k cosine KNN.

FuseLogOdds

fuse_log_odds(signal_1, signal_2, ...) - exact signed log-likelihood-ratio addition with one relevance prior.

PositiveEvidencePool

pool_positive_evidence(signal_1, signal_2, ...) - gated, confidence-scaled retrieval pooling without a calibration theorem.

BayesianEvidenceFusion

fuse_bayesian_evidence(signal_1, signal_2, ...) - exact signed log-likelihood-ratio addition with one relevance prior.

GraphPagerank

graph_pagerank([graph_name]) - PageRank over a named graph.

GraphHits

graph_hits([graph_name]) - HITS over a named graph.

GraphBetweenness

graph_betweenness([graph_name]) - betweenness centrality over a named graph.

GraphTraverse

graph_traverse(graph_name, start_vertex, label, max_hops) - BFS traversal scoring.

GraphNeighbors

graph_neighbors(graph_name, vertex_id, label, direction) - 1-hop neighbor expansion.

MultiFieldMatch

multi_field_match(field_1, query_1, field_2, query_2, ...) - per-field Bayesian BM25 probabilities fused with log-odds conjunction.

StagedRetrieval

staged_retrieval(field_1, query_1, top_k_1, field_2, query_2, top_k_2, ...) - cascading BM25 text_match: each stage filters the candidate set from the previous stage and keeps top-k.

DeepPredict

deep_predict(model_name) - runs the saved deep-fusion model.

UQAHighlight

uqa_highlight(field, query [, start_tag, end_tag, max_fragments, fragment_size]) - markup search results around matched terms.

UQAFacets

uqa_facets(field [, field2, ...]) - facet counts over the posting list, computed against the current row context.

TraverseMatch

traverse_match(graph, start, label, max_hops) - BFS traversal emitting (doc_id, score) weighted by hop distance.

TemporalTraverse

temporal_traverse(graph, start, label, max_hops, t_min, t_max)

• traverse_match filtered by edge valid_from/valid_to.

RPQ

rpq(expr, start [, graph]) - evaluate a Regular Path Query (Definition 5.1.2) and emit endpoint vertex ids reachable from start along paths matching expr.

GraphCreate

graph_create(graph_name) - register a new in-memory graph.

GraphDrop

graph_drop(graph_name) - drop a registered graph.

GraphExists

graph_exists(graph_name) - AGE agtype boolean graph probe.

GraphLabelCreate

create_vlabel(graph_name, label_name) / create_elabel(graph_name, label_name) - register an AGE label.

GraphLabelDrop

drop_label(graph_name, label_name [, force]) - drop an AGE label together with its entities.

GraphAlter

alter_graph(graph_name, operation, new_value) - AGE graph alteration (RENAME).

GraphEdges

graph_edges(graph_name [, label]) - emit every edge in the graph as (source, target, label, weight) rows.

AttentionFusion

attention(signal_1, signal_2, ...) - multi-signal attention fusion (single-head).

LearnedFusion

learned_fusion(model, signal_1, ...) - learned per-feature weight fusion using a saved LearnedFusion model.

CalibratedVectorMatch

calibrated_vector_match(field, vector, k [, threshold]) - KNN with calibrated cosine probabilities (Paper 5).

SparseThreshold

sparse_threshold(signal, threshold) - drop scores at or below the threshold and subtract it from survivors.

ScoreBM25

score_bm25([field,] query) - projection helper exposing the current match score.

ScoreBayesianBM25

score_bayesian_bm25([field,] query) - projection helper exposing the current Bayesian BM25 match score.

DeepLearn

deep_learn(model, training_set) - kick off analytical training (Paper 4) for the named deep-fusion model.

Convolve

Deep-fusion construction helpers used inside deep_learn / deep_predict argument expressions:

Pool

Flatten

Dense

Softmax

Layer

Model

Trait Implementations

impl Clone for FunctionKind

fn clone(&self) -> FunctionKind

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Copy for FunctionKind

impl Debug for FunctionKind

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

Formats the value using the given formatter.

impl Eq for FunctionKind

impl PartialEq for FunctionKind

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

Equality operator ==.

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

Inequality operator !=.

impl StructuralPartialEq for FunctionKind