Introduces a unified MatrixBuilder abstraction for persistent bit and integer matrices, replacing custom dispatch enums and boolean flags across consumers. Refactors index, merge, and select layers to adopt explicit merge modes and centralized lifecycle management. Adds a complete k-mer filtering implementation with partition-level processing, progress tracking, and a new CLI subcommand supporting configurable predicates and thresholds.
Introduces `PersistentSparseCompactIntMatrix`, a row-major sparse integer matrix combining a bitmatrix support layer with dual value streams for singleton and multi partitions. Adds builder and packing utilities to convert dense count matrices into the new on-disk format, featuring idempotency guards and crash-safe persistence workflows. Exposes core accessors, iterators, and column weight calculations while adjusting internal visibility for companion modules. Includes unit tests and diagnostic examples to validate equivalence with dense representations and measure compaction metrics.
Extracts file creation, opening, and memory mapping logic into a new `mmap_file` module. This replaces repetitive unsafe operations across persistent data structures with standardized helper functions, ensuring consistent header validation, error handling, and data layout while keeping public APIs unchanged.
Update PersistentSparseBitMatrix documentation for clarity. Introduce select0 to the rank/select module, and add next_geq, contains, successor, and predecessor methods to the Elias-Fano implementation. Include naive reference implementations and unit tests to validate correctness and boundary conditions for all new APIs. All changes are strictly additive.
Extended `nonzero_triples` with a boolean flag to conditionally skip sorting when input slots are already ordered. Updated call sites to pass appropriate flags, enabling a single-pass columnar traversal in the sparse matrix builder. Refactored `build_from_dense` into a batched processing pipeline that reduces repeated file access and improves sequential read performance. Added a diagnostic example to validate the new builder against existing dense matrices.
The command now supports reopening existing indexes instead of failing when the output file exists. Control flow branches between opening an existing index and constructing a new one, moving configuration setup exclusively to the creation path. Directory existence is enforced upfront with proper I/O error propagation. The --force flag retains its original semantics by removing the target directory before proceeding with a fresh build.
Update all dependent crates and internal imports to reference the renamed module. Expose configuration parameters and genome metadata through new public accessor methods in the index API, and update documentation comments to align with the new package name.
Update Cargo manifests, dependency paths, and Rust imports across the workspace to reflect the `obikpartitionner` to `obikpartition` rename. Synchronize architecture and implementation documentation with the new module structure. Fix minor syntax issues in test assertions to ensure compilation compatibility. No behavioral or API changes are introduced.
Replaced scattered direct metadata loading with centralized instance methods on `KmerPartition` to guarantee consistent error mapping and legacy recovery. Introduced `StorageKind`, `LayerContent`, and `EvidenceKind` enums alongside lightweight disk-probe methods that inspect file presence without opening heavy data structures. Updated callers across the index, partitioner, and phylo modules to use the new partition API, and added unit tests validating the introspection behavior.
Refactor `pack_sparse_bit_matrix` to conditionally open either a packed or columnar dense representation based on file existence. This eliminates the previous unconditional pre-conversion step that forced a full packed copy. Update cleanup logic and add tests to verify direct transposition, proper file handling, and idempotency for both formats.
Consolidate row extraction and parallel column reduction into reusable generic utilities within the pairwise module. Replace manual iteration in builder and view methods with iterator-based zipped loops and deferred overflow processing to improve memory access patterns. Optimize distance functions by switching to bulk byte passes for branch-free SIMD vectorization, adding a merge-like helper to correct masked overflow values without secondary allocations. Add a comprehensive test validating the optimized distance paths against naive references across multiple thresholds and overflow scenarios.
Consolidate bit matrix operations by delegating pairwise metric calculations to IntSliceView and col_view implementations. Introduce chunked_presence_count in colgroup to handle threshold-based accumulation efficiently. Add a direct point lookup method to the sparse matrix representation to eliminate buffer allocations, and simplify match arms in persistent accessors accordingly. All public APIs and behavioral contracts remain unchanged.
Extract duplicated traversal logic into a shared `NonzeroSlotsView` trait and `nonzero_triples` helper. Update slice view iterators to consistently yield `(position, value)` tuples and use a temporary buffer with explicit permutation mapping for sorted access. Delegate manual slot sorting and column iteration to the new shared helper, eliminating per-format duplication and eager collection while preserving existing public API signatures.
Refactored `sub_matrix` to delegate column population to `fill_sub_matrix`, replacing manual buffer allocation and explicit permutation loops with direct value assignment via `enumerate_slots_values`. This eliminates intermediate allocations, reduces pipeline overhead, and simplifies control flow while preserving slot ordering and permutation semantics.
Renames `get_batch`, `fill_batch`, and `fill_batch_sorted` to `collect_slots_values`, `fill_slots_values`, and `fill_slots_values_sorted` to align with updated `common_traits 0.13` APIs. Introduces optimized batch retrieval that sorts input indices for sequential mmap access before reordering outputs to match the original query order. Updates test suites to reflect the new method signatures without altering validation logic or coverage.
Replaces nested column-major point lookups with a batched iterator that delegates to format-native traversal strategies. The implementation enforces a single pass per matrix type, using row-major iteration for sparse formats and eager collection for packed/columnar layouts while preserving original slot ordering. Memory allocation is optimized by removing `n_cols`-wide buffers in favor of per-row buffering or lazy iteration. Correctness tests verify iterator output against dense baselines across all supported layouts, and architecture documentation is updated to reflect the new format-agnostic query pattern.
Bumps core dependencies including ndarray, rand, hashbrown, niffler, ureq, sysinfo, indicatif, lru, and remove_dir_all. Adapts source code to accommodate breaking changes by migrating RNG initialization, adjusting HTTP response handling, and replacing the fs4 crate with standard library file locking. Adds a planning document for query benchmarking and sparse index regression tests.
Updated core libraries and ecosystem packages, including the full serde suite and regex dependencies. Restructured syn dependency resolution by pinning 2.0.117 while introducing a standalone 3.0.3 entry for derive macros. Downgraded windows-sys to 0.59.0 and replaced the anes dependency in criterion2 with bpaf and walkdir. Removed an unused IntSliceView import from test modules.
Introduce a `resume` method for persistent bit and int matrix builders to restore dimensions from persisted metadata. This enables incremental column appending across separate build sessions without requiring manual dimension tracking. Consolidate builder lifecycle management in the merge layer using a unified `MatrixBuilder` enum, simplify closure logic, and add instrumentation. Include tests verifying state preservation and data integrity across multiple resume cycles.
Introduce runtime dimension tracking, matrix shape validation, and per-layer diagnostics across multiple modules to improve execution observability. Extract intermediate computation results into local variables, standardize collection patterns in partial methods, and replace naive iteration with optimized pairwise counting. All modifications are strictly additive or structural; public APIs, data models, and core logic remain unchanged.
Introduce a `Sparse` variant for `PersistentBitMatrix` with full method dispatch and implementations of `ColumnWeights` and `BitPartials` traits. Refactor distance matrix computations in layered stores to eagerly collect results before reduction, enabling reliable dimension tracking. Add comprehensive debug logging across phylo commands and distance modules to report matrix shapes and flag out-of-bounds indices during CSV iteration.
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.
Implements a compact, row-major sparse bit matrix backed by memory-mapped components, introducing EliasFano, PersistentFixedIntVec, and PersistentRankSelectBitVec primitives for efficient storage and decoding. Adds a BinaryMatrix trait to unify row-level operations across dense and sparse implementations. Corrects edge-case behaviors for zero-width bit storage and cardinality-0 rows. Delivers reduced on-disk size and faster random row access, with column reads remaining dense-only. Test suites and benchmarks are included but currently marked as ignored.
Adds compiler inlining hints to accessors, iterators, bitwise operations, and distance functions across multiple modules. This optimization aims to reduce call overhead for frequently invoked methods without modifying runtime behavior, API contracts, or data models.
Centralize the layer count validation into PartitionCache and track it via a new fast_mode flag. Extend query tuples to include a pre-resolved destination layer index, enabling a fast-path batch lookup that bypasses per-layer probing when enabled. Refactor neighbor iteration and hit resolution to eliminate duplication and conditionally dispatch to the optimized path based on the cache state.
Relocates `SiblingAnnex`, `FamilyMask`, and `SiblingAnnexBuilder` from `obicompactvec` to the local `obikphylo::siblings` module. Replaces the previous implementation with a memory-mapped version using `memmap2`, featuring a 2-byte-per-slot layout, explicit bitfield manipulation, and support for concurrent atomic writes. Updates all sibling module imports to local paths and adds the `memmap2` dependency to `obikphylo`.
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.
Introduces batch retrieval and sub-matrix extraction methods across vector, view, reader, and matrix types. These implementations optimize cache locality by sorting requested indices for sequential memory access before applying an inverse permutation to restore original order. Includes allocation-free variants that populate caller-provided buffers. Updates architecture documentation to define sibling annex persistence in iteration order and clarify pipeline separation.
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.
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.
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.
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.
Refactor the query pipeline into a two-stage MPHF hit-detection pass followed by a column-major matrix fetch to improve cache efficiency. Introduce a QueryHit enum for event-driven callbacks, decoupling hit detection from data population. Add scan/fetch metrics to QueryStats, update Phase 4 architecture docs, and align tests with the new callback signature.
Add header parsing helpers to extract column counts without memory mapping. Update packing functions to verify existing files match current metadata, preventing stale or widened-column artifacts. Extract inline tests in obilayeredmap to an external module and add comprehensive aggregation tests. Bump obikmer to 1.1.35 and clean up repository configuration.
The pairwise matrix functions now explicitly calculate and overwrite diagonal entries using `f(i,i)`, replacing previous implicit symmetric mirroring or default values. Documentation has been updated to clarify that diagonals represent self-comparison weights, ensuring accurate self-similarity calculations. Additionally, the obikmer crate version has been bumped to 1.1.34.
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.
Introduces `new_ones` and `add_col_ones` methods to directly initialize all-ones bit vectors and matrix columns. This replaces redundant initialization sequences that created zero-filled structures and applied bitwise NOT, with a single pass that writes contiguous 0xFF bytes to disk. The change eliminates inversion overhead, streamlines test setup, and improves performance for filter mask intersection logic while preserving identical semantics.
Refactor bitmatrix, colgroup, and layer modules to replace manual iteration with concise `or_where` predicates and bulk inversion calls. This simplifies the codebase and leverages optimized internal implementations for improved performance.
Adds `FilterMask` and conditional bitwise methods (`*_where`) to `obicompactvec` for composable column-based slot filtering. Extends `obikpartitionner` with a `MatrixGroupOps` trait and `column_mask_expr` method to express aggregate constraints as vectorized masks. Refactors matrix builder management into a unified `Builders` enum and introduces `try_compute_combined_mask`, enabling O(1) slot checks and skipping unnecessary row reads during partitioning and rebuilding passes.
Expands MatrixGroupOps with partial_group_min/max helpers for bitwise reductions and introduces add_col_from methods to persist external vectors as matrix columns. Refactors column aggregation in the partitioner to leverage these group operations directly, replacing iterative row processing with simplified builder lifecycle management and explicit metadata serialization.
Refactors column and matrix access to use unified `BitSliceView` and `IntSliceView` abstractions, replacing legacy `PackedCol`/`IntColView` types. Introduces `BitSlice`/`IntSlice` traits for zero-copy, trait-based bitwise and arithmetic operations across persistent and temporary vector types. Removes deprecated in-memory `MemoryBitVec` and `MemoryIntVec` implementations and their tests, while updating dependent crates to use the new view-based API and `BitSliceMut` trait.
Introduces `TempCompactIntVec` and `TempBitVec` as temporary, file-backed intermediates to replace eager in-memory vectors, enabling OS-level paging under memory pressure. Updates the `MatrixGroupOps` trait to return `io::Result` types, allowing proper error propagation and supporting chunked accumulation for large column groups. Includes builder patterns with `.freeze()` finalization, automatic `TempDir` cleanup on drop, and necessary test updates to handle the new fallible signatures. Also fixes `Cargo.toml` section ordering.
Introduce the `ColGroup` struct and `MatrixGroupOps` trait to manage named subsets of column indices and perform additive aggregations (count, sum, any). Implement these operations for `PersistentBitMatrix` and `PersistentCompactIntMatrix`, applying size-optimized branches for presence counts and direct accumulation for small groups. Additionally, add a `mask_with` trait method that efficiently zero-sets elements based on a mask, optimized for sparse masks with O(n_zeros) complexity. Include comprehensive tests covering overflow handling, slot masking, and result additivity across partitioned data.
This commit introduces `BitColView` and `IntColView` to abstract over Columnar and Packed storage formats, implementing `BitSlice` and `IntSlice` for uniform column access. It adds `col_view()` accessors to `PersistentBitMatrix` and `PackedCompactIntMatrix`, explicitly panicking on implicit variants. The new types are publicly re-exported, and unit tests are added to validate per-element retrieval, aggregation methods, and parity with the original columnar representation.
Refactor `cmp_scalar`, `min`, `max`, `add`, and `diff` to operate directly on the primary byte array, deferring overflow slot resolution to a secondary pass. This eliminates HashMap lookups in the hot path and enables SIMD vectorization. Add six unit tests to validate correct promotion and demotion between storage slots when values cross the 255 threshold.
Implemented `sum()`, `count_nonzero()`, and `iter()` to complete the numeric vector interface. The builder now computes aggregate values across memory-mapped regions and overflow entries, while the reader delegates these operations to its inherent methods. The iterator provides zero-copy access to underlying `u32` elements.
Refactor `BitIter` to process `u64` chunks using word-aligned shifts instead of byte-level operations. Introduce a dedicated `MemoryBitIter` for `MemoryBitVec`, updating its `iter()` and `IntoIterator` implementations accordingly. Hide `MemoryBitIter` from the public API to narrow the crate's interface, while leveraging explicit alignment guarantees for safer and more efficient bit extraction.
Centralizes overflow handling and improves modularity by replacing manual mmap indexing and row loops with composable iterator patterns. This change leverages Rust's iterator traits for efficient, idiomatic column traversal while encapsulating data access in a dedicated view struct.