Expand description
IR evaluation metrics, TREC parsing, batch evaluation, statistical testing. IR evaluation metrics, TREC parsing, batch evaluation, statistical testing.
- Binary relevance: Precision, Recall, MRR, NDCG, AP, ERR, RBP, F-measure
- Graded relevance: NDCG and MAP with multi-level relevance
- TREC format: Load and parse run files and qrels
- Batch evaluation: Process multiple queries at once
- Statistical testing: Paired t-test, confidence intervals, Cohen’s d
- Export: CSV and JSON output
Re-exports§
pub use batch::evaluate_batch_binary;pub use batch::evaluate_trec_batch;pub use batch::BatchResults;pub use batch::QueryResults;pub use binary::DegradationMetrics;pub use export::export_to_csv;pub use statistics::cohens_d;pub use statistics::confidence_interval;pub use statistics::paired_t_test;pub use statistics::TTestResult;pub use trec::group_qrels_by_query;pub use trec::group_runs_by_query;pub use trec::load_qrels;pub use trec::load_trec_runs;pub use trec::Qrel;pub use trec::TrecRun;pub use validation::validate_beta;pub use validation::validate_metric_inputs;pub use validation::validate_persistence;pub use validation::ValidationError;pub use binary::Metrics;pub use export::export_to_json;
Modules§
- batch
- Batch evaluation utilities for processing multiple queries.
- binary
- Binary relevance IR evaluation metrics.
- export
- Export utilities for evaluation results (CSV, JSON).
- graded
- Graded relevance IR evaluation metrics.
- statistics
- Statistical testing utilities for evaluation results.
- trec
- TREC format parsing utilities.
- validation
- Input validation utilities for metrics and evaluation.