{
 "dataset": "assessed-phase-assemblages",
 "schema": "assessed-phase-assemblage-v1",
 "method": {
  "name": "assessed phase-assemblage points (ORCHID)",
  "version": "1.0.0",
  "annotation_run": "run-20261021h-full",
  "annotation_tools_sha256": "c8c3e77469eddd6d9bc28c90060e4d20564dbab33e9ea02b176b26db0a1b4f45",
  "phase_mapper_sha256": "7393ff687870a212358a13810346e1c109cd9a2dcd16d779d7e83ee4ca5a0d3d",
  "phase_mapper": {
   "path": "orchid-project.org/tools/phase_mapper.py",
   "git_blob": "6b40eec5c5f8039a041a5804b09f37c1799eb924",
   "repository": "https://github.com/orchid-calphad/orchid-project.org"
  },
  "builder": "orchid-project.org/tools/build_assessed_assemblages.py"
 },
 "diagrams_with_points": 2178,
 "diagrams_seen": 2272,
 "diagrams_not_annotated": 149,
 "fields_seen": 27118,
 "points": 23129,
 "point_method": {
  "centroid": 21616,
  "pole_of_inaccessibility": 1513
 },
 "points_by_n_phases": {
  "1": 4639,
  "2": 18490
 },
 "points_by_basis": {
  "deterministic_adjacency": 8035,
  "reader_consensus": 6793,
  "deterministic_printed_label": 3205,
  "adjudicator": 2803,
  "deterministic_phase_rules": 2186,
  "expert_reviewed": 87,
  "dft_hull_corroborated": 20
 },
 "points_by_confidence": {
  "high": 14452,
  "medium": 8603,
  "low": 74
 },
 "points_by_review_state": {
  "not reviewed by the expert": 23022,
  "expert decision (tier 5)": 87,
  "expert-reviewed (tier 5)": 20
 },
 "excluded_fields_by_reason": {
  "too_thin_for_margin": 2047,
  "tracing_fault": 740,
  "partly_drawn_truncated_edge": 482,
  "undetermined_in_source": 255,
  "awaiting_expert": 172,
  "diagram_tracing_suspect": 123,
  "axis_component_not_in_system": 89,
  "open": 70,
  "more_than_two_phases": 5,
  "cannot_tell": 4,
  "schematic_field_unbounded_below": 2
 },
 "phase_labels_mapped_to_registry": {
  "mapped": 19823,
  "unmapped": 21796
 },
 "two_phase_tie_lines": {
  "with_tie_line": 11961,
  "without": 6529,
  "omitted_by_reason": {
   "side_assignment_ambiguous": 4395,
   "neighbour_field_unnamed": 908,
   "frame_edge_phase_ambiguous": 732,
   "endpoints_do_not_cover_the_two_phases": 363,
   "boundary_not_a_single_clean_line": 119,
   "isotherm_reaches_cut_off_frame_edge": 12
  },
  "line_compound_endpoints": 8814
 },
 "sgte_crosscheck": {
  "comparisons": 30769,
  "agrees": 17876,
  "disagrees": 3612,
  "not_comparable": 9281,
  "agreement_rate_of_comparable": 0.8319,
  "not_comparable_by_reason": {
   "assemblage phase without registry mapping": 8005,
   "no calculated equilibrium at this cell": 25,
   "order-disorder model: the grid label does not tell the ordered from the disordered state": 278,
   "outside the calculated T-x range": 405,
   "the same registry phase twice (the grid collapses composition sets)": 568
  },
  "points": {
   "no SGTE assessment": 10194,
   "agrees with >=1": 8333,
   "not comparable": 3510,
   "disagrees with all comparable": 1092
  },
  "metric_definitions": {
   "comparison": "one point x one calculated SGTE diagram of its system (nearest grid cell)",
   "comparable": "both phase sets can be expressed as registry phases and the grid cell has an equilibrium: result agrees or disagrees",
   "agreement_rate_of_comparable": "agrees / (agrees + disagrees), over all comparable comparisons (aggregate, not a per-system mean)",
   "points_agreeing_with_at_least_one_rate": "points that agree with >= 1 SGTE assessment / points with >= 1 comparable comparison",
   "median_system_rate": "median over systems with >= 5 comparable comparisons of the system's agreement rate",
   "not_comparable": "counted separately, never in a rate; reasons in not_comparable_by_reason and, per comparison, in sgte_not_comparable.csv"
  },
  "comparable": 21488,
  "points_with_comparable_assessment": 9425,
  "points_agreeing_with_at_least_one": 8333,
  "points_agreeing_with_at_least_one_rate": 0.8841,
  "points_agreeing_with_all_comparable": 7508,
  "points_disagreeing_with_all_comparable": 1092,
  "systems_with_5_comparable": 889,
  "median_system_rate": 0.8889,
  "mean_system_rate": 0.848
 },
 "points_by_boundary_basis": {
  "printed_solid": 20782,
  "mixed": 1292,
  "dashed_estimated": 1028,
  "none": 27
 },
 "diagrams_with_dashed_lines": 1021,
 "ml_folds": {
  "fold_lso": {
   "0": 4669,
   "1": 4644,
   "2": 4516,
   "3": 4274,
   "4": 5026
  },
  "fold_lpo": {
   "0": 4293,
   "1": 4285,
   "2": 5437,
   "3": 4334,
   "4": 4780
  },
  "method": "`chemical_system` (sorted elements, = `system`), `paper_id` (stable source id), `fold_lso` = leave-chemical-system-out fold int(sha256(chemical_system), 16) mod 5, `fold_lpo` = leave-paper-out fold int(sha256(paper_id), 16) mod 5 (tools/ml_folds.py). Subsystems are their own systems (a held-out ternary may have its binaries in training); drop subsystems of held-out systems for strict chemistry extrapolation"
 },
 "espei": {
  "datasets_default": 1068,
  "datasets_with_tie_lines": 1068,
  "points_in_espei": 9877,
  "points_excluded_from_espei": 13252,
  "validation": {
   "espei-zpf": {
    "files": 1068,
    "valid": 1068,
    "loaded": 1068,
    "errors": [],
    "espei": "0.9.1"
   },
   "espei-zpf-with-tie-lines": {
    "files": 1068,
    "valid": 1068,
    "loaded": 1068,
    "errors": [],
    "espei": "0.9.1"
   },
   "espei-zpf-dense": {
    "files": 1,
    "valid": 1,
    "loaded": 1,
    "errors": [],
    "espei": "0.9.1"
   }
  }
 },
 "doi_corrections": {
  "table": "tools/doi_corrections.json",
  "version": "1.0.0",
  "sha256": "1406a2c1b91197ce810bbced42af8ecf80fe2c58b53e5bef60091cdf47655d96",
  "entries": 22,
  "records_with_dropped_original_doi": 89,
  "systems": [
   "AG-PU",
   "CE-SB",
   "CL-EU",
   "CL-RE",
   "I-P",
   "IN-TE",
   "OS-RE",
   "S-SC",
   "SB-SR",
   "SR-TE"
  ]
 },
 "dense_field_grid": {
  "method_version": "1.0.0",
  "diagrams": [
   "CU-ZR-p371"
  ],
  "points": 353,
  "fields": 25,
  "points_by_n_phases": {
   "1": 71,
   "2": 282
  },
  "points_in_espei": 353,
  "points_excluded_from_espei": 0,
  "default_enabled_for": [
   "CU-ZR-p371"
  ],
  "files": "dense-field-grid/ and espei-zpf-dense/",
  "label_overrides": {
   "CU-ZR-p371": {
    "CuZr": "phase:cu-zr:cuzr-b2",
    "Cu51Zr14": "phase:cu-zr:cu51zr14"
   }
  }
 },
 "cite_as": {
  "dataset": "assessed-phase-assemblages",
  "version": "2026.10",
  "release_tag": "v2026.10",
  "license": "CC-BY-4.0",
  "status": "Version 1, machine-extracted: phase fields of expert-drawn (assessed) diagrams named by deterministic rules, model readers, an adjudicator and partial expert review (basis and review state per record); one point deep inside each field; boundary and tie-line positions are drawing-based and of low credence.",
  "title": "ORCHID-Assemblages: assessed phase assemblages of binary systems",
  "doi": "10.5281/zenodo.23735996",
  "url": "https://doi.org/10.5281/zenodo.23735996",
  "text": "R. Arróyave, J. Saal, D. Shin, ORCHID-Assemblages: assessed phase assemblages of binary systems (version 2026.10; Version 1, machine-extracted), Zenodo (2026). https://doi.org/10.5281/zenodo.23735996",
  "bibtex": "@misc{orchid_assessed_assemblages_2026,\n  author    = {Arr{\\'o}yave, Raymundo and Saal, James and Shin, Dongwon},\n  title     = {{ORCHID-Assemblages: assessed phase assemblages of binary systems}},\n  version   = {2026.10},\n  publisher = {Zenodo},\n  year      = {2026},\n  doi       = {10.5281/zenodo.23735996},\n  url       = {https://doi.org/10.5281/zenodo.23735996},\n  note      = {Version 1, machine-extracted}\n}",
  "also_cite": [
   "for every diagram used, its original evaluation (citations[role=original_evaluation], DOI where matched) and H. Okamoto, Desk Handbook: Phase Diagrams for Binary Alloys, 2nd ed., ASM International (2010) as the compilation",
   "B. Hallstedt, The SGTE collection of binary datasets, Calphad 89 (2025) 102833, https://doi.org/10.1016/j.calphad.2025.102833 (CC BY 4.0), and the original assessment named in each file"
  ],
  "part_of": {
   "title": "ORCHID data: umbrella record of the ORCHID datasets",
   "doi": "10.5281/zenodo.23736065",
   "url": "https://doi.org/10.5281/zenodo.23736065",
   "text": "R. Arróyave, J. Saal, D. Shin, ORCHID data: umbrella record of the ORCHID datasets (version 2026.10; Version 1, machine-extracted), Zenodo (2026). https://doi.org/10.5281/zenodo.23736065",
   "bibtex": "@misc{orchid_data_2026,\n  author    = {Arr{\\'o}yave, Raymundo and Saal, James and Shin, Dongwon},\n  title     = {{ORCHID data: umbrella record of the ORCHID datasets}},\n  version   = {2026.10},\n  publisher = {Zenodo},\n  year      = {2026},\n  doi       = {10.5281/zenodo.23736065},\n  url       = {https://doi.org/10.5281/zenodo.23736065},\n  note      = {Version 1, machine-extracted}\n}"
  }
 },
 "files": {
  "bytes_by_file_or_dir": {
   "CITATION.cff": 1868,
   "dense-field-grid": 381061,
   "espei-zpf": 8681278,
   "espei-zpf-dense": 270307,
   "espei-zpf-with-tie-lines": 8659636,
   "espei_exclusions.jsonl": 1813833,
   "exclusions.csv": 322766,
   "exclusions.jsonl": 542755,
   "records.csv": 12639690,
   "records.jsonl.gz": 4771823,
   "records.parquet": 1721635,
   "sgte_not_comparable.csv": 1215857,
   "summary.json": 0
  }
 },
 "parquet": {
  "table": "records.parquet",
  "rows": 23129,
  "csv_sha256": "aa214b354d21442458ebed481f4c06802932d47445bb25c403ea48b7187d7626",
  "content_sha256": "1fe84de89f1ee0d682f7da37975d02e9c8891276e005a7ef78c4cf5cb10e890f",
  "note": "content_sha256 = SHA-256 of the canonical JSON rows (sorted keys); independent of the Parquet encoder"
 }
}