1 rollouts.presentation.candidate_validity

rollouts.presentation.candidate_validity

Static descriptive views of admitted, immutable candidate-validity evidence.

Primary outcomes partition materialized candidate rows; independently triggered reasons can overlap. All full-chain summaries are retrospective and count candidate occurrences, not independent scenes. Rendering never acquires stores, changes admission, estimates uncertainty, or mutates the supplied evidence.

1.1 Functions

Name	Description
candidate_outcome_color	Keep semantic colours stable across original/refined and subset views.
candidate_triggered_reasons_figure	Render independently triggered rejection counts, allowing overlap.
candidate_validity_figure	Render primary outcomes through time, selected families and chain totals.

1.1.1 candidate_outcome_color

rollouts.presentation.candidate_validity.candidate_outcome_color(outcome)

Keep semantic colours stable across original/refined and subset views.

1.1.2 candidate_triggered_reasons_figure

rollouts.presentation.candidate_validity.candidate_triggered_reasons_figure(
    evidence,
    *,
    title='All triggered rejection reasons',
)

Render independently triggered rejection counts, allowing overlap.

Every cell is an admitted reason count at its stored zero-based step index, displayed as one-based acquisition number. Reasons are not summed into a partition: a candidate may trigger several. VALID is excluded. Empty or unsupported evidence is labelled explicitly. The returned figure owns one heatmap axis and, when populated, a colourbar.

1.1.3 candidate_validity_figure

rollouts.presentation.candidate_validity.candidate_validity_figure(
    evidence,
    *,
    step_index,
    percentages=False,
    title='Candidate validity',
)

Render primary outcomes through time, selected families and chain totals.

1.1.3.1 Parameters

evidence Domain-admitted candidate rows. outcomes starts with VALID; each primary count partitions materialized rows at its stored step index. step_index Stored zero-based step index to highlight; axes display acquisitions as step_index + 1. A missing step is explicitly unavailable rather than represented as a measured zero. percentages Scale temporal bars by the step’s materialized rows and selected-family bars by that family’s materialized rows. Whole-chain totals stay counts. title Overall figure title.

1.1.3.2 Returns

matplotlib.figure.Figure Four axes with exact descriptive marks, ready for an external canvas. No global plotting backend or figure registry is changed.

# rollouts.presentation.candidate_validity { #aria_nbv.rollouts.presentation.candidate_validity } `rollouts.presentation.candidate_validity` Static descriptive views of admitted, immutable candidate-validity evidence. Primary outcomes partition materialized candidate rows; independently triggered reasons can overlap. All full-chain summaries are retrospective and count candidate occurrences, not independent scenes. Rendering never acquires stores, changes admission, estimates uncertainty, or mutates the supplied evidence. ## Functions | Name | Description | | --- | --- | | [candidate_outcome_color](#aria_nbv.rollouts.presentation.candidate_validity.candidate_outcome_color) | Keep semantic colours stable across original/refined and subset views. | | [candidate_triggered_reasons_figure](#aria_nbv.rollouts.presentation.candidate_validity.candidate_triggered_reasons_figure) | Render independently triggered rejection counts, allowing overlap. | | [candidate_validity_figure](#aria_nbv.rollouts.presentation.candidate_validity.candidate_validity_figure) | Render primary outcomes through time, selected families and chain totals. | ### candidate_outcome_color { #aria_nbv.rollouts.presentation.candidate_validity.candidate_outcome_color } ```python rollouts.presentation.candidate_validity.candidate_outcome_color(outcome) ``` Keep semantic colours stable across original/refined and subset views. ### candidate_triggered_reasons_figure { #aria_nbv.rollouts.presentation.candidate_validity.candidate_triggered_reasons_figure } ```python rollouts.presentation.candidate_validity.candidate_triggered_reasons_figure( evidence, *, title='All triggered rejection reasons', ) ``` Render independently triggered rejection counts, allowing overlap. Every cell is an admitted reason count at its stored zero-based step index, displayed as one-based acquisition number. Reasons are not summed into a partition: a candidate may trigger several. VALID is excluded. Empty or unsupported evidence is labelled explicitly. The returned figure owns one heatmap axis and, when populated, a colourbar. ### candidate_validity_figure { #aria_nbv.rollouts.presentation.candidate_validity.candidate_validity_figure } ```python rollouts.presentation.candidate_validity.candidate_validity_figure( evidence, *, step_index, percentages=False, title='Candidate validity', ) ``` Render primary outcomes through time, selected families and chain totals. #### Parameters {.doc-section .doc-section-parameters} evidence Domain-admitted candidate rows. ``outcomes`` starts with VALID; each primary count partitions materialized rows at its stored step index. step_index Stored zero-based step index to highlight; axes display acquisitions as ``step_index + 1``. A missing step is explicitly unavailable rather than represented as a measured zero. percentages Scale temporal bars by the step's materialized rows and selected-family bars by that family's materialized rows. Whole-chain totals stay counts. title Overall figure title. #### Returns {.doc-section .doc-section-returns} matplotlib.figure.Figure Four axes with exact descriptive marks, ready for an external canvas. No global plotting backend or figure registry is changed.