nimare.studyset.View

class View(store, index=None, point_mask=None, context=None)

Bases: object

A selection of analyses over one immutable store.

Methods

analyses_with_points(point_bool)	Analyses with at least one flagged focus, as a view.
coordinate_block()	Foci for the selection, grouped by analysis.
drop_keys(keys)	Narrow to everything except the named analyses.
frame(name)	Return a memoised pandas frame.
image_block(imtype, *[, policy])	Return the images of one type for the selection.
label_block([annotation])	Return the annotation matrix for the selection.
points_in_mask(mask_data, affine)	Point-level boolean for foci inside an image mask.
points_near(xyz, radius)	Point-level boolean for foci within radius mm of any of xyz.
position_of_row()	Store analysis row -> position in this selection, -1 where absent.
resolve(requirements[, drop_invalid])	Intersect each requirement's validity, then resolve against the result.
row_of_key()	{full analysis id: position in this selection}, memoised.
select(mask_or_positions)	Narrow to a subset of the selected analyses.
select_keys(keys, *[, allow_short, strict])	Narrow to the named analyses, by full id or short analysis id.
select_points(mask)	Narrow to a subset of foci, keeping every analysis.
select_studies(study_keys, *[, exclude, strict])	Narrow by parent study.
text_block([field])	Return one text field for the selection.
with_context(**changes)	Return a view with different execution context.

Properties

store
index
point_mask
context
keys	The full study-analysis ids of the selected analyses.
short_keys	The analysis ids the selected analyses declare, without their study.
study_keys	Unique study ids of the selected analyses.

analyses_with_points(point_bool)

Analyses with at least one flagged focus, as a view.

Counted with bincount over the point-to-analysis column rather than add.reduceat over the offsets: reduceat returns the element at the index when two offsets coincide, so analyses with no foci came back as hits.

coordinate_block()

Foci for the selection, grouped by analysis.

drop_keys(keys)

Narrow to everything except the named analyses.

An id matching nothing is not an error: it is absent from the result either way.

frame(name)

Return a memoised pandas frame. Compatibility surface, not a hot path.

image_block(imtype, *, policy='all')

Return the images of one type for the selection.

property keys

The full study-analysis ids of the selected analyses.

label_block(annotation=None)

Return the annotation matrix for the selection.

points_in_mask(mask_data, affine)

Point-level boolean for foci inside an image mask.

Uses the coordinates in the context’s space: a Talairach focus and its MNI projection are not the same voxel.

points_near(xyz, radius)

Point-level boolean for foci within radius mm of any of xyz.

position_of_row()

Store analysis row -> position in this selection, -1 where absent.

Memoised against the view, because every block build wants it.

resolve(requirements, drop_invalid=True)

Intersect each requirement’s validity, then resolve against the result.

Returns (narrowed view, {name: block}). Because the narrowing produces a view, every block is aligned to the same analyses by construction – there is no parallel-list bookkeeping to get wrong.

row_of_key()

{full analysis id: position in this selection}, memoised.

select(mask_or_positions)

Narrow to a subset of the selected analyses. O(k).

select_keys(keys, *, allow_short=True, strict=True)

Narrow to the named analyses, by full id or short analysis id.

Parameters:

• keys (str or array_like of str) – Analysis ids, each a full "<study id>-<analysis id>" id or the analysis id on its own.

• allow_short (bool, default=True) – Whether an analysis id on its own may be matched.

• strict (bool, default=True) – Whether to raise on an id matching no analysis in this selection.

Raises:

ValueError – If strict and any id matches nothing. The error names them.

select_points(mask)

Narrow to a subset of foci, keeping every analysis.

select_studies(study_keys, *, exclude=False, strict=True)

Narrow by parent study.

strict raises on a study id with no analyses in this selection. It does not apply to exclude, for the reason given on drop_keys().

property short_keys

The analysis ids the selected analyses declare, without their study.

property study_keys

Unique study ids of the selected analyses.

text_block(field='abstract')

Return one text field for the selection.

with_context(**changes)

Return a view with different execution context. No data is touched.