API Reference

This section provides detailed documentation for the LociSimiles Python API, auto-generated from source code docstrings.

Core Modules

Document Module

The Document module provides classes for representing and loading text collections:

• TextSegment - Individual text unit with ID and content
• Document - Container for text segments

Pipeline Module

The Pipelines module provides the main processing pipelines:

• Pipeline - Generic composer: combine any generator + judge
• RetrievalPipeline - Semantic similarity retrieval
• ClassificationPipeline - Text pair classification
• ClassificationPipelineWithCandidateGeneration - Two-stage retrieval + classification
• RuleBasedPipeline - Lexical matching + linguistic filters

Generators Module

The Generators module provides candidate-generation components:

• EmbeddingCandidateGenerator - Semantic embedding similarity
• ExhaustiveCandidateGenerator - All-pairs (no filtering)
• RuleBasedCandidateGenerator - Lexical matching + linguistic filters

Judges Module

The Judges module provides scoring/classification components:

• ClassificationJudge - Transformer-based sequence classification
• ThresholdJudge - Binary decisions from candidate scores
• IdentityJudge - Pass-through (judgment_score = 1.0)

Evaluator Module

The Evaluator module provides tools for assessing detection quality:

• IntertextEvaluator - Main evaluation class

Quick Reference

Loading Documents

from locisimiles import Document

doc = Document("texts.csv")

Saving Results

# Save from a pipeline instance
results = pipeline.run(query=query_doc, source=source_doc, top_k=10)
pipeline.to_csv("results.csv")
pipeline.to_json("results.json")

# Or use standalone functions
from locisimiles.pipeline import results_to_csv, results_to_json
results_to_csv(results, "results.csv")
results_to_json(results, "results.json")

Evaluating Results

from locisimiles import IntertextEvaluator

evaluator = IntertextEvaluator(predictions, ground_truth)
metrics = evaluator.evaluate()