The index packing API now accepts a `sparse` parameter to generate `PersistentSparseBitMatrix` files alongside existing dense matrices. The sibling cache automatically detects this format via an `is_multi.prsb` marker file and routes queries identically to the dense variant. A new `--sparse` CLI flag exposes the option, with tests verifying end-to-end pipeline correctness and storage equivalence.
Documents a proposed redesign for cross-partition batch resolution, shifting trigger logic to per-destination accumulator thresholds and introducing entropy-based pruning criteria. Adds an `n_layers_per_partition` method to the index, exposing partition metadata with consistent error handling and clarified documentation regarding build-time structural properties.
Shifts sibling annex construction from slot-indexed enumeration to iteration-order traversal by correcting cumulative k-mer offset tracking in batch enumeration. Replaces coarse per-partition parallelism with chunked work distribution to prevent thread starvation on skewed partitions. Decouples custom progress messages from ETA updates to eliminate display clobbering during high-frequency callbacks. Adds regression tests validating batch offset correctness, partial batch handling, and iterator-order consistency across layer builds.
Relocate the `siblings` and `cardcomp` modules from `obikindex` to a dedicated `obikphylo` workspace member. Convert inherent methods on `KmerIndex` into extension traits, update import paths across `obikmer`, and add supporting accessor methods to `obikseq` and `obilayeredmap`. This restructuring reduces the public API surface of `obikindex` while organizing phylogenetic iteration, caching, and distance calculation logic under a dedicated crate.
Transitions the index from MPHF slot-based to physical iteration-order indexing, aligning with the unitig layout. Introduces a streaming-only pipeline for k-mer iteration that adheres to memory constraints by avoiding full in-memory collections. Updates layer and sibling iterators to own an Arc clone of the file reader, making them Send + 'static and safe for concurrent use without borrowing the parent. Exposes batch and k-mer iterator types publicly while simplifying signature syntax with modern lifetime elision.
Transition the minorant flag from a derived value to a stored field within the family mask, resolving a performance regression where on-the-fly reconstruction consumed significant query time. This change introduces O(1) k-mer reconstruction APIs, shifts minorant computation to the index build phase, and enables direct annex-based statistics. Supporting updates include adopting shared ownership for partition caches and refactoring batch processing pipelines.
Refactor `scan_layer_families` across the siblings module to accept a closure callback instead of returning an intermediate collection. This eliminates eager materialization and per-layer buffering by streaming results directly into genome-specific buffers or tally matrices. The update introduces bounded batch processing and scratch buffer reuse to cap peak auxiliary memory, while preserving existing computational behavior, control flow, and error semantics.
The sibling calculation modules now process layer directories sequentially instead of in parallel. This eliminates concurrent processing overhead and prevents interleaved cache sweeps, improving disk I/O and page-cache locality for partition-grouped data access. Progress bar updates and result accumulation have been adapted to the sequential control flow, while core filtering logic and output structures remain unchanged.
Introduces a dedicated `family_scan` submodule to consolidate per-layer family traversal logic. Centralizes path validation, minorant filtering, and conditional matrix instantiation into a shared `scan_layer_families` function. Updates sibling-annex consumers to leverage the new abstraction, reducing inline scanning code. Adds a test fixture to verify numerical consistency across consumer methods for co-occurrence and base-pair metrics.
Extracts directory cleanup, partition initialization, and finalization into dedicated helper methods within KmerIndex. This centralizes force-flag handling and reduces boilerplate across merge, rebuild, and select workflows. Additionally updates the CLI progress bar template to display an ETA indicator following elapsed time.
Introduce the obikmer name-tree subcommand to map numeric leaf labels in phylogenetic tree exports back to taxon names using a reference FASTA file. Correct the --free-loss flag behavior by removing cardinality transition costs from pairwise cost calculations, ensuring sibling gains and losses are priced identically to whole-family events. Update documentation, configuration parameters, and add reference phylogenetic data files.
Establishes foundational infrastructure across multiple crates by introducing unified persistent bit matrix storage with columnar, packed, and implicit variants, alongside De Bruijn graph node encoding and unitig iteration logic. Adds a macro-driven parallel pipeline scheduler featuring NUMA-aware runners, bounded channels, and memory budgets to enforce concurrency limits. Implements streaming nucleotide parsers with pooled page buffers for FASTA, FASTQ, and Genbank formats, complemented by system resource monitoring, progress tracking, and stage profiling utilities. Collectively, these changes provide the core data models, execution frameworks, and I/O pipelines required for downstream k-mer indexing and analysis workloads.
Introduces a new predicate module in obikindex that implements genome metadata predicate parsing, evaluation, and group classification using three-valued logic. Extends the IndexMeta API with methods for single-predicate filtering and group quorum filtering. Updates obikmer command modules to delegate filter construction and matching to the centralized index API, removing local definitions and simplifying call sites.
Replaced the legacy Sankoff parsimony pipeline with a new cardinality-composition decomposition that constructs row-normalized transition probability matrices symmetrized via geometric mean. This ensures reversibility, reduces free parameters from 240 to 120, and guarantees a zero diagonal. Tallies are now explicitly restricted to variable families to align with +ASC-corrected alignment populations. Additionally, fixed `--exclude-genome` handling to re-scan surviving sequences and drop newly monomorphic columns, preventing silent data corruption in downstream tree inference tools.
Implements a calibrated 16-state Sankoff substitution cost matrix and CLI pipeline for evolutionary distance computation, including empirical calibration via saturation-filtered SNP counts. Refactors the sibling scanning stage to use batched transforms for improved synchronization efficiency. Introduces an OS-level advisory directory lock across all index-modifying commands to prevent concurrent write corruption. Updates dependencies and exposes new Sankoff utilities in the public API.
Re-enables the `numa` feature in CI workflows to prevent container/cgroup deadlocks while preserving validation correctness. Fixes concurrent test race conditions by replacing thread-local parameter storage with process-wide atomics and mutex locks. Integrates `tracing-subscriber` for structured logging and adds thread-ID tracking to debug worker lifecycles. Additionally bumps the crate version, updates `.gitignore`, documents experimental evolutionary distance pipelines, and refactors hardcoded test constants.
Introduce `obisys::effective_parallelism()` to read Linux cgroup v1/v2 CPU quotas from sysfs, preventing thread pool oversubscription in containerized environments. Replace direct `std::thread::available_parallelism()` calls across `obikindex` and `obikmer` with this centralized function. Bump `obikmer` version to 1.1.41.
Introduces a `SnpAlignment` struct and helper methods to construct per-genome SNP pseudo-alignments from sibling k-mer data, filtering monomorphic families and encoding bases as IUPAC ambiguity codes. Exposes the type at the crate root for simplified imports. Adds a `--snp` CLI flag to compute and export these alignments as an IUPAC-coded FASTA file. Updates theory documentation to propose a multi-genome framing approach for joint phylogenetic inference, resolving pairwise correspondence ambiguities through positional homology and partial coverage thresholds. Bumps crate version to 1.1.40.
Exposes RawSnpDistanceOutput and implements KmerIndex::raw_snp_distance() to compute pairwise single-copy locus counts under a paralogy-aware rule. The implementation leverages ndarray for parallel matrix aggregation, producing raw p-distance matrices for sanity-checking. A --raw-snp-distance CLI flag is added to export results as CSV, mapping zero-eligible pairs to NA.
Replaces the explicit `SiblingInfo` struct and 3-bit minorant flags with a derived 4-bit presence mask (`FamilyMask`) that tracks observed bases per family. This eliminates redundant file I/O overhead by introducing a `PartitionCache` for batch lookups, simplifies serialization, and updates all downstream builders, stats computation, and tests to operate on the new bitmask representation. Adjusts CLI output to report deduplicated family sizes instead of histograms, ignores generated CSV files, and updates documentation to reflect the fixed canonical reference and new theory.
Add the `obipipeline` crate and replace sequential scatter/gather logic with a concurrent pipeline using `Flat` and `Transform` stages. Introduce `SiblingAnnexStats` API to compute distributions, and add CLI flags to `distance.rs` for constructing the annex and exporting statistics as CSV.
Introduce the siblingannex module in obicompactvec to store per-slot minorant flags and sibling counts in a memory-mapped annex file. Add a scatter-gather pipeline in obikindex to compute these values across index layers and write them to .psib files. Implement central_canonical_neighbors in obikseq for generating strand-aware k-mer variants around the middle base. Expose rolling statistics in obiskbuilder and update dependency graphs accordingly.
Implement the Mash distance metric across the CLI, index, and compact vector traits. This includes adding a `Mash` variant to the `DistanceMetric` enum and `MetricArg` CLI argument, implementing the conversion from Jaccard distances using the standard mutation-rate estimator formula, and updating documentation with supported metrics and algorithmic references. Additionally, add an `entropy` method to rolling statistics for computing order-specific entropy.
Shifts the NUMA-aware runner from a flat, round-robin model to a per-node architecture using dedicated `NodeActivation` channels. Replaces absolute deltas with relative scaling based on the previous growth step's worker count, decoupling growth from node count to fix slow ramp-up and enforce per-node fairness. Updates architecture documentation to reflect these changes and focus tuning questions on `INITIAL`/`GROWTH_DIVISOR` parameters for I/O-bound validation.
Replaces the monolithic CPU scaling threshold with separate CPU and I/O spawn thresholds. Introduces an `IoSample` struct with platform-specific byte reading and a relative throughput growth heuristic. Adds a 0.1s wall-clock guard to `CpuSample` to suppress artificial efficiency spikes, and updates `maybe_activate` to trigger worker scaling when either resource indicates headroom. Bumps `obikmer` to v1.1.33 and updates architecture documentation.
Updates the release workflow to run macOS builds inside a Docker container with explicit registry authentication and adjusted artifact paths. Bumps the obikmer crate version to 1.1.29 and adds *.log to .gitignore. Simplifies NUMA worker spawning by lowering the activation threshold from 0.95 to 0.2, replacing complex stateful tracking with a direct efficiency check, and downgrading progress logging to debug level. Includes general code formatting improvements for readability.
Refactor the release workflow to use a structured container object with authenticated pulls for macOS ARM64 builds. Replace single-worker activation with dynamic upfront provisioning based on node and worker counts. Implement an absolute efficiency gain threshold for scaling checks and add early termination to improve adaptive scaling stability. Bump obikmer crate version to 1.1.27.
Refactor the Gitea release pipeline to generate releases via API and upload binaries using a shared ID. Automate changelog generation by fetching recent commits with `jj log` and producing markdown notes via `aichat`. Convert `hwlocality` to an optional dependency gated by a default `numa` feature, providing fallback implementations for graceful degradation when NUMA support is disabled. Bump obikmer to 1.1.18.
Replaces the fallible NUMA topology builder with an infallible fallback that synthesizes a single-node UMA configuration on failure or absence. Refactors PartitionRunner to pre-spawn dormant workers and dynamically activate them via CPU efficiency thresholds, replacing static upfront spawning with adaptive scaling. Bumps obikmer crate version to 1.1.15.
Integrates an obisys::Reporter across indexing and command modules to capture execution metrics. Replaces discarded timer stops with explicit rep.push() calls, adds timing instrumentation for the pack stage, and prints collected reports after each selection branch.
Introduces a Make-based orchestration for simulating, indexing, merging, filtering, and verifying k-mer counts and presence. Exposes internal builder and iterator APIs publicly, enforces mandatory leading slashes for predicate patterns, registers the `obitaxonomy` crate, and updates tooling configurations alongside documentation.
Introduce `MemoryBitVec` and `MemoryIntVec` for efficient in-memory storage with hybrid compression and overflow handling. Implement `BitSlice`, `BitSliceMut`, `IntSlice`, and `IntSliceMut` traits across persistent and memory-backed types to enable generic slice operations and bitwise/arithmetic overloads. Add `col_persist` and `col_as_memory` methods to `BitMatrix` and `IntMatrix` for efficient column extraction. Align with the new single-pass rebuild architecture by supporting fast kmer filtering and matrix rebuilding. Includes comprehensive tests and profiling instrumentation for the packing phase.
Replaces `rayon` parallel iteration across index, rebuild, reindex, and select modules with a custom `PartitionRunner`. This introduces NUMA-aware task distribution with CPU pinning and round-robin scheduling, eliminating `Arc`, `Mutex`, and atomic synchronization primitives in favor of a flat, pre-spawned worker architecture. Error handling is simplified via `.map_err()` and the `?` operator, while progress bar updates are decoupled into dedicated callbacks.
Replaces the previous partition processing logic with an adaptive, NUMA-aware multi-threaded worker pool that dynamically scales active threads based on real-time CPU efficiency. Introduces pre-spawned, CPU-pinned threads managed via crossbeam channels and Rayon to optimize memory bandwidth and core utilization. Adds a `max_workers()` accessor to aggregate maximum worker capacity across NUMA nodes and updates diagnostics to report active versus maximum worker counts.
Replace NUMA-naive Rayon loops and ad-hoc adaptive pools with a unified `PartitionRunner` that manages a NUMA-aware worker pool. The implementation uses pinned Rayon thread pools per node and activates dormant threads based on real-time CPU efficiency metrics. This standardizes partition-level parallelism, optimizes memory locality, and eliminates cross-socket traffic. Includes architecture documentation and updates mkdocs navigation.
Replaces the global Rayon pool with per-NUMA-node thread pools that pin worker threads to their respective nodes, leveraging Linux first-touch allocation to reduce cross-NUMA memory contention and improve cache locality. Integrates the `hwlocality` crate with a vendored build, includes graceful fallbacks for single-socket or non-Linux systems, and updates dependency constraints. Also adds installation and architecture documentation, and corrects parallelism detection in the partitioner.
Centralize inline spawn checks into a `should_spawn_worker` function with adaptive thresholds. The first worker spawns at <95% CPU efficiency, while subsequent workers only trigger if marginal efficiency gain exceeds 25% of the expected `1/n_workers` (minimum 3%). Also increases the spawn poll interval from 10s to 20s.
This change enhances observability and adaptability in the merge pipeline. Performance timing and debug logging are added to the De Bruijn graph and partition merge layers to track phase durations and pipeline metrics. The merge module replaces blocking receives with timed polls to sample CPU efficiency, dynamically spawning workers when utilization drops below a threshold. A new script is also introduced to parse merge debug logs and generate structured Markdown reports detailing throughput, phase breakdowns, and partition performance.
Applies consistent formatting, whitespace normalization, and indentation standardization to `debruijn.rs` and `merge.rs`. Reorganizes imports and downgrades a unitig traversal log from `info!` to `debug!`. No functional logic or runtime behavior is altered.
Introduce CpuSample to measure process-level CPU efficiency and wall-clock time. Use crossbeam-channel to distribute partition merging tasks to a dynamic worker pool that scales based on CPU utilization, capped at half the available cores. Update diagnostics to track pool usage.
Introduce the `obipipeline` dependency and refactor merge and partition logic to leverage parallel execution. Update `merge_partitions` to use rayon with dynamic memory budgeting and concurrency control via a pilot run. Refactor Pass 1 to concurrently read unitigs, filter kmers through a shared `LayeredMap`, and populate the graph safely. Simplify diagnostics to report total kmer counts and replace manual flags with graph length validation.
Introduce a `peak_rss_bytes()` utility for accurate per-phase RAM measurement. Replace the genome-length heuristic with a dynamic seed expansion ratio based on actual RSS delta. Explicitly drop the `GraphDeBruijn` instance before MPHF construction to prevent resource contention and ensure proper memory management.
This commit improves memory management by respecting Linux cgroup v1/v2 limits and introduces a configurable memory budget for the new `rebuild` subcommand to prevent OOM during index reconstruction. The rebuild process now supports filtering, compaction, and parallelization. Diagnostic capabilities are expanded with debug-level tracing for partition merges, k-mer expansion tracking, and utility flags for label renaming, matrix size breakdowns, per-genome counts, and partition distribution reporting. Accessor methods for active and remaining memory have also been added to the stats struct.
Introduces a memory-aware scheduling strategy for parallel partition merging that replaces unbounded concurrency with a First-Fit Decreasing approach gated by a thread-safe `MemoryBudget` semaphore. An adaptive expansion factor, seeded by a sequential pilot run, dynamically caps concurrent workers to prevent hashbrown OOMs. Adds a `--budget-fraction` CLI flag to configure RAM allocation, enhances the CLI to accept multiple indexes, and introduces comprehensive partition diagnostics including memory utilization tracking, concurrency metrics, and statistical summaries with ASCII histograms. Updates documentation and navigation accordingly.
Introduces allocation-free `sum()` and `count_nonzero()` methods for compact integer vectors, extending the `ColumnWeights` trait with `partial_kmer_counts`. Adds parallel partition scanning to the k-mer index for computing per-genome distinct k-mer counts, and exposes a new `--stats` CLI flag to output these statistics as CSV.
Introduces the `select` CLI command to project and aggregate genome-level k-mer data by column. Adds `filter` as an alias for `rebuild`. The implementation uses parallel partition processing, supports metadata-driven grouping with configurable aggregation operators, and performs atomic in-place rewrites or filtered exports. Updates documentation and navigation accordingly.
Refactor partition iteration to support a generic `on_partition` callback executed after each parallel partition completes. Split the logic into bounded and unbounded paths; the bounded path uses an `AtomicUsize` to enforce row limits, while the unbounded path eliminates atomic contention to improve throughput. Additionally, integrate a progress bar into the dump command by passing an increment callback to `idx.dump()`, ensuring proper initialization and cleanup.
Introduce thread-local `Vec<u8>` buffers to eliminate concurrent I/O contention. Replace the mutable row counter with an `AtomicUsize` and `fetch_update` to enable lock-free early termination when the limit is reached. Collected chunks are then written sequentially to preserve partition ordering.
Introduces `--head N` to the `dump` command for early iteration termination and `--presence-threshold N` to the `distance` command for Jaccard filtering on count indexes. Updates filter defaults to adapt based on explicit ingroup/outgroup declarations. Fixes a Rust type mismatch in the unitig closure and updates partition iteration callbacks to return `bool` for proper early termination support. Documentation is updated accordingly.
Introduce a `--presence-threshold` CLI argument (default: 1) and update `KmerIndex::distance` to accept a `presence_threshold` parameter. This replaces hardcoded zero thresholds, enabling configurable filtering of low-abundance kmers during Jaccard distance calculations.
Eric Coissac