Skip to content

Class: ParsedAnalysis

The points a paper reports together for one tested effect, or for one non-result purpose such as the regions of interest an analysis used.

In an original parse, the grouping is the ingestion workflow's proposal, made from a table and its caption. pondie confirms or overrides it with the whole paper in view. The role is required and is set in the ingestion workflow by the roles classifier stage, after the analyses are identified. It is also the ingestion workflow's proposal: pondie's accept confirms it, and a relabel overrides it with a reviewed correction that wins.

URI: parse:ParsedAnalysis

 classDiagram
    class ParsedAnalysis
    click ParsedAnalysis href "../ParsedAnalysis/"
      ParsedAnalysis : anchor_kind

      ParsedAnalysis : cells





        ParsedAnalysis --> "*" CellRef : cells
        click CellRef href "../CellRef/"



      ParsedAnalysis : coordinate_space





        ParsedAnalysis --> "0..1" ReportedSpace : coordinate_space
        click ReportedSpace href "../ReportedSpace/"



      ParsedAnalysis : derived_from

      ParsedAnalysis : description

      ParsedAnalysis : from_prior_study

      ParsedAnalysis : key

      ParsedAnalysis : name

      ParsedAnalysis : name_is_printed

      ParsedAnalysis : origin





        ParsedAnalysis --> "1" CoordinateOrigin : origin
        click CoordinateOrigin href "../CoordinateOrigin/"



      ParsedAnalysis : points





        ParsedAnalysis --> "*" ParsedPoint : points
        click ParsedPoint href "../ParsedPoint/"



      ParsedAnalysis : prior_study_evidence





        ParsedAnalysis --> "*" TextSpan : prior_study_evidence
        click TextSpan href "../TextSpan/"



      ParsedAnalysis : role

      ParsedAnalysis : role_confidence

      ParsedAnalysis : role_source

      ParsedAnalysis : space_as_printed

      ParsedAnalysis : space_basis





        ParsedAnalysis --> "0..1" SpaceBasis : space_basis
        click SpaceBasis href "../SpaceBasis/"



      ParsedAnalysis : split





        ParsedAnalysis --> "0..1" SignSplit : split
        click SignSplit href "../SignSplit/"



      ParsedAnalysis : statistic

      ParsedAnalysis : statistic_df

      ParsedAnalysis : table_id

      ParsedAnalysis : text_spans





        ParsedAnalysis --> "*" TextSpan : text_spans
        click TextSpan href "../TextSpan/"



      ParsedAnalysis : thresholds





        ParsedAnalysis --> "*" Threshold : thresholds
        click Threshold href "../Threshold/"



Slots

Name Cardinality and Range Description Inheritance
key 1
String
The analysis's identity, derived from what was extracted -- its name, `coor... direct
cells *
CellRef
For a table analysis, every cell it was read from: each point's row and colum... direct
text_spans *
TextSpan
For a text or figure analysis, the sentences or legend it was read from: thos... direct
origin 1
CoordinateOrigin
direct
table_id 0..1
String
The ParsedTable the points were read from direct
name 1
String
The analysis's label, as printed where the paper prints one direct
name_is_printed 0..1
Boolean
False when the producer composed the name because the paper printed none, so ... direct
description 0..1
String
direct
coordinate_space 0..1
ReportedSpace
The space of the analysis's points; absent when the paper states none direct
space_as_printed 0..1
String
direct
space_basis 0..1
SpaceBasis
direct
role 1
String or 
CoordinateRole
What the points are for direct
anchor_kind 0..1
String or 
AnchorKind
For an anchor, what it defines direct
from_prior_study 0..1
Boolean
The coordinates come from another publication rather than from this study: a ... direct
prior_study_evidence *
TextSpan
The citation or sentence showing the coordinates come from another publicatio... direct
role_confidence 0..1
Float
The classifier's probability for role (and anchor_kind) direct
role_source 0..1
String
The classifier that set the role, by name and version, such as `set-role-enco... direct
split 0..1
SignSplit
Present on both halves of an analysis split by sign, and only there direct
derived_from *
String
On a revision, the keys of the revised parse's analyses this one was made fro... direct
statistic 0..1
String or 
StatisticKind
The test statistic of this analysis's one test -- t, z, F, chi_square -- what... direct
statistic_df 0..1
String
Degrees of freedom as printed, such as "1, 38" or "27" direct
thresholds *
Threshold
The thresholds the analysis's results survived, at every level the paper stat... direct
points *
ParsedPoint
Empty when the paper names the analysis but reports no coordinates for it -- ... direct

Usages

used by used in type used
CoordinateParse analyses range ParsedAnalysis

Identifier and Mapping Information

Schema Source

  • from schema: https://neurostuff.org/schemas/neuroimaging-paper-parse

Mappings

Mapping Type Mapped Value
self parse:ParsedAnalysis
native parse:ParsedAnalysis

LinkML Source

Direct

name: ParsedAnalysis
description: 'The points a paper reports together for one tested effect, or for one
  non-result purpose such as the regions of interest an analysis used.

  In an original parse, the grouping is the ingestion workflow''s proposal, made from
  a table and its caption. pondie confirms or overrides it with the whole paper in
  view. The role is required and is set in the ingestion workflow by the `roles` classifier
  stage, after the analyses are identified. It is also the ingestion workflow''s proposal:
  pondie''s accept confirms it, and a relabel overrides it with a reviewed correction
  that wins.'
from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse
rank: 1000
attributes:
  key:
    name: key
    description: 'The analysis''s identity, derived from what was extracted -- its
      `name`, `coordinate_space` and `points` -- never from where it was read, so
      the same analysis read from two sources (a table and its HTML copy, a table
      and the sentence restating it) reaches the same key, and a re-run that extracts
      the same content reaches it again. A rename or a changed point is a new key.
      The name is part of the key because differently named analyses can print identical
      points.

      `analysis#<h>`, where `h` is the first 16 hex characters of the sha256 of the
      UTF-8 bytes of the compact JSON array `[<name>, <space>, [<point>, ...]]` (no
      whitespace, non-ASCII characters unescaped). `<name>` is `name` normalized:
      Unicode NFKC, casefolded, every dash (Unicode Pd and U+2212 minus) as `-`, zero-width
      characters and all whitespace dropped, so `PO > Sil` and `PO>Sil` are one name.
      `<space>` is `coordinate_space`, or `""` when absent.

      Each `<point>` is `[<x>, <y>, <z>, <space>, [[<kind>, <value>], ...]]`: the
      coordinates rounded to 0.01 mm and written with two decimals (`"-4.00"`); the
      point''s own `space` when it differs from the analysis''s, else `""`; and each
      of its `values` as its `kind` and its value written to 6 significant digits
      (`"5.8"`, `"3.2e-12"`). Negative zero is written as zero. Each point''s values,
      and then the points, are sorted by their own compact JSON, in code point order.

      An analysis with no points -- a contrast a table lists as "n.s." -- hashes its
      name and space alone. The two halves of a sign split hold different points under
      different names, so each has its own key. `study_schema.keys.analysis_key` computes
      it.

      This key is what the extraction record''s `Analysis.source_table_analysis` and
      `CoordinateSet.id` name, and what neurostore stores as `Analysis.source_id`.
      Two analyses of one parse never share a name, space and points, so never share
      a key.'
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    identifier: true
    domain_of:
    - ParsedAnalysis
    - AnalysisVerdict
    range: string
    required: true
  cells:
    name: cells
    description: 'For a table analysis, every cell it was read from: each point''s
      row and column group, and the rows that name the contrast without coordinates.
      Required for `origin: table`; every point''s row and column group is among them.'
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    domain_of:
    - TableRow
    - ParsedAnalysis
    range: CellRef
    multivalued: true
    inlined_as_list: true
  text_spans:
    name: text_spans
    description: 'For a text or figure analysis, the sentences or legend it was read
      from: those holding its coordinates, or the one stating a result that has none.'
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    domain_of:
    - ParsedAnalysis
    - OmittedAnalysis
    range: TextSpan
    multivalued: true
    inlined_as_list: true
  origin:
    name: origin
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    domain_of:
    - ParsedAnalysis
    range: CoordinateOrigin
    required: true
  table_id:
    name: table_id
    description: The ParsedTable the points were read from. Present exactly when `origin`
      is `table`.
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    domain_of:
    - ParsedTable
    - ParsedAnalysis
    - TableReading
    range: string
  name:
    name: name
    description: The analysis's label, as printed where the paper prints one.
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    domain_of:
    - Producer
    - ParsedAnalysis
    - OmittedAnalysis
    range: string
    required: true
  name_is_printed:
    name: name_is_printed
    description: False when the producer composed the name because the paper printed
      none, so a reader never mistakes a placeholder for the paper's wording.
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    domain_of:
    - ParsedAnalysis
    range: boolean
  description:
    name: description
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    domain_of:
    - ParsedAnalysis
    range: string
  coordinate_space:
    name: coordinate_space
    description: The space of the analysis's points; absent when the paper states
      none. A point states its own only when it differs.
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    domain_of:
    - ParsedAnalysis
    range: ReportedSpace
  space_as_printed:
    name: space_as_printed
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    domain_of:
    - ParsedAnalysis
    range: string
  space_basis:
    name: space_basis
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    domain_of:
    - ParsedAnalysis
    range: SpaceBasis
  role:
    name: role
    description: 'What the points are for. Only a `result` is an analysis with a statistic
      in the extraction record; an `anchor` -- a region of interest, a seed, a stimulation
      target -- becomes a CoordinateSet there, and never enters a studyset as an analysis.

      Required. Set in the ingestion workflow by the `roles` classifier stage, after
      the analyses are identified: a classifier that reads each set and its context,
      named by `role_source`. It is the ingestion workflow''s proposal, which pondie
      audits: accept confirms it, and relabel overrides it with a reviewed correction
      that wins. The extractors never propose a role, and nothing supplies a default.

      Numbers that are not brain coordinates, such as channel numbers or lattice points,
      get no role. They are not a coordinate set.'
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    domain_of:
    - ParsedAnalysis
    - AnalysisVerdict
    required: true
    any_of:
    - range: CoordinateRole
    - range: string
  anchor_kind:
    name: anchor_kind
    description: For an `anchor`, what it defines.
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    domain_of:
    - ParsedAnalysis
    any_of:
    - range: AnchorKind
    - range: string
  from_prior_study:
    name: from_prior_study
    description: 'The coordinates come from another publication rather than from this
      study: a peak quoted for comparison (always true for `role: reference`), another
      study''s peak taken as this study''s ROI or seed, or a peak from the authors''
      own earlier publication given with its statistics, which is `role: result` with
      this flag true so that a meta-analysis can avoid counting the earlier study
      twice. Separate from the role, because a borrowed seed is still the seed this
      study''s analysis used. False when the study defined them itself, including
      an ROI built from its own earlier result, meaning an earlier analysis in this
      same paper such as a localizer.'
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    domain_of:
    - ParsedAnalysis
    range: boolean
  prior_study_evidence:
    name: prior_study_evidence
    description: The citation or sentence showing the coordinates come from another
      publication.
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    domain_of:
    - ParsedAnalysis
    range: TextSpan
    multivalued: true
    inlined_as_list: true
  role_confidence:
    name: role_confidence
    description: The classifier's probability for `role` (and `anchor_kind`).
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    domain_of:
    - ParsedAnalysis
    range: float
    minimum_value: 0
    maximum_value: 1
  role_source:
    name: role_source
    description: The classifier that set the role, by name and version, such as `set-role-encoder@2026-10-09`.
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    domain_of:
    - ParsedAnalysis
    range: string
  split:
    name: split
    description: 'Present on both halves of an analysis split by sign, and only there.
      The split is declared here and nowhere else: a reader never infers it from a
      name.'
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    domain_of:
    - ParsedAnalysis
    range: SignSplit
    inlined: true
  derived_from:
    name: derived_from
    description: On a revision, the keys of the revised parse's analyses this one
      was made from by a split or a merge. Absent on an analysis carried over unchanged.
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    domain_of:
    - ParsedAnalysis
    range: string
    multivalued: true
  statistic:
    name: statistic
    description: 'The test statistic of this analysis''s one test -- t, z, F, chi_square
      -- whatever the columns call it. One analysis is one test. Every value at a
      point that describes the same test belongs to that point: the statistic, its
      p value (uncorrected and corrected), an effect size, the cluster''s p. A different
      test is a different analysis, even on the same rows: an omnibus F and the directional
      post-hoc t that follows it are two analyses, told apart by `column_group` when
      one table prints both.'
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    domain_of:
    - ParsedAnalysis
    any_of:
    - range: StatisticKind
    - range: string
  statistic_df:
    name: statistic_df
    description: Degrees of freedom as printed, such as "1, 38" or "27".
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    domain_of:
    - ParsedAnalysis
    range: string
  thresholds:
    name: thresholds
    description: 'The thresholds the analysis''s results survived, at every level
      the paper states: most often a voxel-level height ("p < .001 uncorrected") together
      with a cluster-level correction ("FWE p < .05"), which one method cannot express.
      Read from the caption, footer or Methods; each threshold cites where.'
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    domain_of:
    - ParsedAnalysis
    range: Threshold
    multivalued: true
    inlined_as_list: true
  points:
    name: points
    description: Empty when the paper names the analysis but reports no coordinates
      for it -- most often because nothing survived the threshold. Whether it found
      an effect is the extraction record's `Analysis.outcome`; an empty list here
      only says that no coordinates were printed.
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    domain_of:
    - ParsedAnalysis
    range: ParsedPoint
    multivalued: true
    inlined_as_list: true

Induced

name: ParsedAnalysis
description: 'The points a paper reports together for one tested effect, or for one
  non-result purpose such as the regions of interest an analysis used.

  In an original parse, the grouping is the ingestion workflow''s proposal, made from
  a table and its caption. pondie confirms or overrides it with the whole paper in
  view. The role is required and is set in the ingestion workflow by the `roles` classifier
  stage, after the analyses are identified. It is also the ingestion workflow''s proposal:
  pondie''s accept confirms it, and a relabel overrides it with a reviewed correction
  that wins.'
from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse
rank: 1000
attributes:
  key:
    name: key
    description: 'The analysis''s identity, derived from what was extracted -- its
      `name`, `coordinate_space` and `points` -- never from where it was read, so
      the same analysis read from two sources (a table and its HTML copy, a table
      and the sentence restating it) reaches the same key, and a re-run that extracts
      the same content reaches it again. A rename or a changed point is a new key.
      The name is part of the key because differently named analyses can print identical
      points.

      `analysis#<h>`, where `h` is the first 16 hex characters of the sha256 of the
      UTF-8 bytes of the compact JSON array `[<name>, <space>, [<point>, ...]]` (no
      whitespace, non-ASCII characters unescaped). `<name>` is `name` normalized:
      Unicode NFKC, casefolded, every dash (Unicode Pd and U+2212 minus) as `-`, zero-width
      characters and all whitespace dropped, so `PO > Sil` and `PO>Sil` are one name.
      `<space>` is `coordinate_space`, or `""` when absent.

      Each `<point>` is `[<x>, <y>, <z>, <space>, [[<kind>, <value>], ...]]`: the
      coordinates rounded to 0.01 mm and written with two decimals (`"-4.00"`); the
      point''s own `space` when it differs from the analysis''s, else `""`; and each
      of its `values` as its `kind` and its value written to 6 significant digits
      (`"5.8"`, `"3.2e-12"`). Negative zero is written as zero. Each point''s values,
      and then the points, are sorted by their own compact JSON, in code point order.

      An analysis with no points -- a contrast a table lists as "n.s." -- hashes its
      name and space alone. The two halves of a sign split hold different points under
      different names, so each has its own key. `study_schema.keys.analysis_key` computes
      it.

      This key is what the extraction record''s `Analysis.source_table_analysis` and
      `CoordinateSet.id` name, and what neurostore stores as `Analysis.source_id`.
      Two analyses of one parse never share a name, space and points, so never share
      a key.'
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    identifier: true
    owner: ParsedAnalysis
    domain_of:
    - ParsedAnalysis
    - AnalysisVerdict
    range: string
    required: true
  cells:
    name: cells
    description: 'For a table analysis, every cell it was read from: each point''s
      row and column group, and the rows that name the contrast without coordinates.
      Required for `origin: table`; every point''s row and column group is among them.'
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    owner: ParsedAnalysis
    domain_of:
    - TableRow
    - ParsedAnalysis
    range: CellRef
    multivalued: true
    inlined: true
    inlined_as_list: true
  text_spans:
    name: text_spans
    description: 'For a text or figure analysis, the sentences or legend it was read
      from: those holding its coordinates, or the one stating a result that has none.'
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    owner: ParsedAnalysis
    domain_of:
    - ParsedAnalysis
    - OmittedAnalysis
    range: TextSpan
    multivalued: true
    inlined: true
    inlined_as_list: true
  origin:
    name: origin
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    owner: ParsedAnalysis
    domain_of:
    - ParsedAnalysis
    range: CoordinateOrigin
    required: true
  table_id:
    name: table_id
    description: The ParsedTable the points were read from. Present exactly when `origin`
      is `table`.
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    owner: ParsedAnalysis
    domain_of:
    - ParsedTable
    - ParsedAnalysis
    - TableReading
    range: string
  name:
    name: name
    description: The analysis's label, as printed where the paper prints one.
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    owner: ParsedAnalysis
    domain_of:
    - Producer
    - ParsedAnalysis
    - OmittedAnalysis
    range: string
    required: true
  name_is_printed:
    name: name_is_printed
    description: False when the producer composed the name because the paper printed
      none, so a reader never mistakes a placeholder for the paper's wording.
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    owner: ParsedAnalysis
    domain_of:
    - ParsedAnalysis
    range: boolean
  description:
    name: description
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    owner: ParsedAnalysis
    domain_of:
    - ParsedAnalysis
    range: string
  coordinate_space:
    name: coordinate_space
    description: The space of the analysis's points; absent when the paper states
      none. A point states its own only when it differs.
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    owner: ParsedAnalysis
    domain_of:
    - ParsedAnalysis
    range: ReportedSpace
  space_as_printed:
    name: space_as_printed
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    owner: ParsedAnalysis
    domain_of:
    - ParsedAnalysis
    range: string
  space_basis:
    name: space_basis
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    owner: ParsedAnalysis
    domain_of:
    - ParsedAnalysis
    range: SpaceBasis
  role:
    name: role
    description: 'What the points are for. Only a `result` is an analysis with a statistic
      in the extraction record; an `anchor` -- a region of interest, a seed, a stimulation
      target -- becomes a CoordinateSet there, and never enters a studyset as an analysis.

      Required. Set in the ingestion workflow by the `roles` classifier stage, after
      the analyses are identified: a classifier that reads each set and its context,
      named by `role_source`. It is the ingestion workflow''s proposal, which pondie
      audits: accept confirms it, and relabel overrides it with a reviewed correction
      that wins. The extractors never propose a role, and nothing supplies a default.

      Numbers that are not brain coordinates, such as channel numbers or lattice points,
      get no role. They are not a coordinate set.'
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    owner: ParsedAnalysis
    domain_of:
    - ParsedAnalysis
    - AnalysisVerdict
    range: string
    required: true
    any_of:
    - range: CoordinateRole
    - range: string
  anchor_kind:
    name: anchor_kind
    description: For an `anchor`, what it defines.
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    owner: ParsedAnalysis
    domain_of:
    - ParsedAnalysis
    range: string
    any_of:
    - range: AnchorKind
    - range: string
  from_prior_study:
    name: from_prior_study
    description: 'The coordinates come from another publication rather than from this
      study: a peak quoted for comparison (always true for `role: reference`), another
      study''s peak taken as this study''s ROI or seed, or a peak from the authors''
      own earlier publication given with its statistics, which is `role: result` with
      this flag true so that a meta-analysis can avoid counting the earlier study
      twice. Separate from the role, because a borrowed seed is still the seed this
      study''s analysis used. False when the study defined them itself, including
      an ROI built from its own earlier result, meaning an earlier analysis in this
      same paper such as a localizer.'
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    owner: ParsedAnalysis
    domain_of:
    - ParsedAnalysis
    range: boolean
  prior_study_evidence:
    name: prior_study_evidence
    description: The citation or sentence showing the coordinates come from another
      publication.
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    owner: ParsedAnalysis
    domain_of:
    - ParsedAnalysis
    range: TextSpan
    multivalued: true
    inlined: true
    inlined_as_list: true
  role_confidence:
    name: role_confidence
    description: The classifier's probability for `role` (and `anchor_kind`).
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    owner: ParsedAnalysis
    domain_of:
    - ParsedAnalysis
    range: float
    minimum_value: 0
    maximum_value: 1
  role_source:
    name: role_source
    description: The classifier that set the role, by name and version, such as `set-role-encoder@2026-10-09`.
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    owner: ParsedAnalysis
    domain_of:
    - ParsedAnalysis
    range: string
  split:
    name: split
    description: 'Present on both halves of an analysis split by sign, and only there.
      The split is declared here and nowhere else: a reader never infers it from a
      name.'
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    owner: ParsedAnalysis
    domain_of:
    - ParsedAnalysis
    range: SignSplit
    inlined: true
  derived_from:
    name: derived_from
    description: On a revision, the keys of the revised parse's analyses this one
      was made from by a split or a merge. Absent on an analysis carried over unchanged.
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    owner: ParsedAnalysis
    domain_of:
    - ParsedAnalysis
    range: string
    multivalued: true
  statistic:
    name: statistic
    description: 'The test statistic of this analysis''s one test -- t, z, F, chi_square
      -- whatever the columns call it. One analysis is one test. Every value at a
      point that describes the same test belongs to that point: the statistic, its
      p value (uncorrected and corrected), an effect size, the cluster''s p. A different
      test is a different analysis, even on the same rows: an omnibus F and the directional
      post-hoc t that follows it are two analyses, told apart by `column_group` when
      one table prints both.'
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    owner: ParsedAnalysis
    domain_of:
    - ParsedAnalysis
    range: string
    any_of:
    - range: StatisticKind
    - range: string
  statistic_df:
    name: statistic_df
    description: Degrees of freedom as printed, such as "1, 38" or "27".
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    owner: ParsedAnalysis
    domain_of:
    - ParsedAnalysis
    range: string
  thresholds:
    name: thresholds
    description: 'The thresholds the analysis''s results survived, at every level
      the paper states: most often a voxel-level height ("p < .001 uncorrected") together
      with a cluster-level correction ("FWE p < .05"), which one method cannot express.
      Read from the caption, footer or Methods; each threshold cites where.'
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    owner: ParsedAnalysis
    domain_of:
    - ParsedAnalysis
    range: Threshold
    multivalued: true
    inlined: true
    inlined_as_list: true
  points:
    name: points
    description: Empty when the paper names the analysis but reports no coordinates
      for it -- most often because nothing survived the threshold. Whether it found
      an effect is the extraction record's `Analysis.outcome`; an empty list here
      only says that no coordinates were printed.
    from_schema: https://neurostuff.org/schemas/neuroimaging-paper-parse/coordinates
    rank: 1000
    owner: ParsedAnalysis
    domain_of:
    - ParsedAnalysis
    range: ParsedPoint
    multivalued: true
    inlined: true
    inlined_as_list: true