Repository navigation
feat(v0.2.0): fracture lineage DAG, persistent precursors, axis calibration, and reproducible pipeline - #17
Merged
Conversation
…ration, and reproducible pipeline
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Completes the v0.2.0 milestone for BraggTrack across beamline ingestion, foundation-model compatibility, 4D fracture/coalescence lineage tracking, and single-command reproducible execution:
ESRF-ID03 Ingestion, Physical Axis Calibration & 3-Axis Kinematics
AxisCalibration(voxel_to_physical(z, y, x) -> (mu, chi, d)) and NeXus/HDF5 auto-discovery (/entry_0000/ESRF-ID03/pco_nf/data,/entry*,NXentry, siblingmu/chi/ddatasets, and generic*.h5/*.nxs/*.hdf5fallback).extract_instance_tablewhile preserving integer voxelbbox_*coordinates for lossless 3D ROI cropping.remove_small_objects,merge_nearby_labels(6-connected face adjacency +bincountcentroids), andrelabel_sequential.kinematics.pywithdelta_d,calibrated_axes, and automatickinematics.csv/kinematics_summary.jsonexport inbraggtrack-track-dataset.Public DINOv2 Default & Dynamic DINOv3 Compatibility
DEFAULT_DINO_MODELto public, ungatedfacebook/dinov2-smallfor zero-auth local, CI, and Colab execution while retaining fullfacebook/dinov3-vitb16-pretrain-lvd1689mcompatibility via dynamic prefix/register token skipping (skip = max(0, seq_len - expected_patches)).braggtrack-segment-datasetto compute temporally smoothed Otsu thresholds (--smooth-window 5) in twoO(1)-memory streaming passes.4D Split/Merge Lineage DAG & Persistent Fracture Precursors
braggtrack/tracking/events.py(resolve_splits_and_merges) resolving1 -> 2+grain fracture splits and2+ -> 1coalescence merges with spatial gating, cosine similarity, fragment share bounds, intensity conservation, and Pass 0 cross-detector Hungarian jump-edge pruning.braggtrack/tracking/precursors.py(compute_fracture_precursors) combining shape anisotropy, plateau-safe 3D ROI sub-peak multiplicity (peak_local_max), multi-view MIP patch cosine dispersion, and contiguousmin_persistencefiltering.generate_fracture_scenariobenchmark andbraggtrack-eventsCLI (events.v1schema).End-to-End Orchestrator, Statistical Ablation, Notebooks & Docs
braggtrack-run(braggtrack/cli/run_pipeline.py) chainingvalidate -> segment -> embed -> track -> eventsand emittingrun_manifest.json(manifest.v1) with SHA-256 input checksums, git commit provenance, and per-stage profiling.--bootstrap-samples,--seed) toscripts/ablation_semantic.py.events_acceptancegate (scripts/check_events_acceptance.py) toscripts/ci_report.py.notebooks/braggtrack_demo.ipynbandnotebooks/dino_segmentation_comparison.ipynb, plusREADME.md,docs/architecture.md,CHANGELOG.md, andpaper.md.Verification
ruff check braggtrack/ tests/ scripts/&ruff format --check braggtrack/ tests/ scripts/— PASSpython scripts/pre_pr_check.py— PASSpytest tests/ -v --cov=braggtrack— PASS (215passed,83%+coverage)python scripts/ci_report.py— PASS (overall=PASSacross all 8 acceptance gates)jupyter nbconvert --to notebook --executeon both notebooks — PASS