Safety Results Over Time: test evidence

Requirement-traced qualification evidence for the safety.viz results-over-time module.

Requirement matrix ↗ — the reviewed source specification these tests trace to.

Scope: 25 coverage rows44 distinct requirement IDs
Tests executed: 130 automated checks24 browser · 106 unit
Result: pass all passing10 evidence screenshots
Generated: 2026-07-12 05:50 UTC
Environment: linux 6.17.0-1018-azure · node v22.23.1 · playwright 1.61.1 · chromium 149.0.7827.55
Test run: Actions run #29181741704

Scope & approach

Traceability for the results-over-time module (a Chart.js reimplementation of RhoInc/safety-results-over-time, under #27), per the convention in CONTRIBUTING.md. Requirement IDs (SROT-*) come from the reviewed 58-row matrix at safety.agent docs/requirements/safety-results-over-time.md, whose Evidence Type column routes rows (unit → Vitest, browser → Playwright).

Each table row traces one requirement to the automated test(s) that evidence it: the requirement ID and its source-matrix rows link back to the reviewed specification, the issue column links the implementing work, and the result column shows the recorded outcome of every matching test from the committed evidence.json with its captured screenshots. Browser evidence is captured at fixed conditions (1280×800, device scale 1) on the canonical Linux CI environment.

Browser evidence (Playwright — `tests/e2e/results-over-time.spec.js`)

15 requirement rows · 15 tests

Requirement	Source matrix rows	Issue	Tests & evidence
SROT-FUNC-001/SROT-FUNC-002	SROT-FUNC-001, SROT-FUNC-002	#27	pass SROT-FUNC-001/SROT-FUNC-002: renders measure, group, filter, y-axis, scale, and display controls (#27)
SROT-FUNC-003/SROT-REG-001	SROT-FUNC-003, SROT-REG-001	#27	pass SROT-FUNC-003: participant note updates when a filter is applied (#27)
SROT-DATA-002	SROT-DATA-002	#27	pass SROT-DATA-002: missing and non-numeric results are dropped with a reported count and visible note (#27)
SROT-FUNC-008/SROT-REG-009	SROT-FUNC-008, SROT-REG-009	#27	pass SROT-FUNC-008: box-and-whisker marks render and toggle off with the Box plots control (#27)
SROT-REG-002/SROT-REG-003	SROT-REG-002, SROT-REG-003	#27	pass SROT-REG-002/SROT-REG-003: grouping draws side-by-side boxes with a group-ordered legend (#27)
SROT-REG-010/SROT-REG-012	SROT-REG-010, SROT-REG-012	#27	pass SROT-REG-010/SROT-REG-012: the outlier overlay shows and hides with the Outliers control (#27)
SROT-REG-011	SROT-REG-011	#27	pass SROT-REG-011: outlier points carry a larger hover radius than their resting radius (#27)
SROT-REG-014/SROT-REG-015	SROT-REG-014, SROT-REG-015	#27	pass SROT-REG-014/SROT-REG-015: hovering a box exposes the summary statistics tooltip (#27)
SROT-FUNC-004/SROT-REG-016/SROT-REG-017	SROT-FUNC-004, SROT-REG-016, SROT-REG-017	#27	pass SROT-FUNC-004/SROT-REG-016/SROT-REG-017: y-limit inputs redraw and invert a crossed pair (#27)
SROT-FUNC-005/SROT-REG-020	SROT-FUNC-005, SROT-REG-020	#27	pass SROT-FUNC-005/SROT-REG-020: Reset Limits restores the data extent (#27)
SROT-REG-018	SROT-REG-018	#27	pass SROT-REG-018: the Scale control switches the y-axis between linear and log (#27)
SROT-FUNC-006/SROT-REG-004/SROT-REG-005	SROT-FUNC-006, SROT-REG-004, SROT-REG-005	#27	pass SROT-FUNC-006/SROT-REG-004/SROT-REG-005: the Visits without data control adds empty timepoints (#27)
SROT-FUNC-007/SROT-REG-006/SROT-REG-007	SROT-FUNC-007, SROT-REG-006, SROT-REG-007	#27	pass SROT-FUNC-007/SROT-REG-006/SROT-REG-007: the Unscheduled visits control shows unscheduled timepoints (#27)
SROT-REG-023/SROT-REG-024	SROT-REG-023, SROT-REG-024	#27	pass SROT-REG-023/SROT-REG-024: start_value selects the initial measure and falls back when absent (#27)
SROT-API	— (see legacy-API note)	#27	pass SROT-API: lifecycle API supports init, setData, setSettings, render, resize, and destroy (#27)

Unit evidence (Vitest — `tests/unit/results-over-time/`)

10 requirement rows · 25 tests

Requirement	Source matrix rows	Issue	Tests & evidence
SROT-CFG-004/005/006/007/009	SROT-CFG-004, SROT-CFG-005, SROT-CFG-007, SROT-CFG-009	#27	`configure.test.js` pass results-over-time configure SROT-CFG-004/005/006/007/009: default column mappings match the original renderer (#27) pass results-over-time structureData SROT-CFG-005: computeVisitOrder falls back to alphanumeric order without an order column (#27) pass results-over-time structureData SROT-CFG-005/006: computeVisitOrder sorts by the numeric order column when present (#27)
SROT-CFG-008/012/013/014/015	SROT-CFG-008, SROT-CFG-012, SROT-CFG-013, SROT-CFG-014	#27	`configure.test.js` pass results-over-time configure SROT-CFG-012: filters normalize to { value_col, label } specs and drop blanks (#27) pass results-over-time configure SROT-CFG-013/014: groups always offer a leading None and group_by falls back to it (#27) pass results-over-time configure SROT-CFG-014: an explicit group_by is honored and its column is offered (#27) pass results-over-time configure SROT-CFG-015/DATA-003: box/outlier/visit toggles default to the original renderer values (#27)
SROT-CFG-017/018/019, SROT-DATA-003	SROT-CFG-017, SROT-CFG-019, SROT-DATA-003	#27	`configure.test.js` / `structureData.test.js` pass results-over-time configure SROT-CFG-017/018/019: unscheduled-visit detection defaults are preserved (#27) pass results-over-time structureData SROT-CFG-017/018: parseUnscheduledPattern reads the /.../flags string form (#27) pass results-over-time structureData SROT-CFG-019: isUnscheduledVisit lets an explicit values list take precedence over the pattern (#27)
SROT-DATA-001/002	SROT-DATA-001, SROT-DATA-002	#27	`checkInputs.test.js` / `structureData.test.js` pass results-over-time checkInputs SROT-DATA-001: passes when the required measure, result, and visit columns are present (#27) pass results-over-time checkInputs SROT-DATA-001: throws naming every missing required column (#27) pass results-over-time checkInputs SROT-DATA-002: an empty dataset reports all required columns missing (#27) pass results-over-time structureData SROT-DATA-002: cleanData drops missing and non-numeric results with a reported count (#27)
SROT-CFG-005/006	SROT-CFG-005, SROT-CFG-006	#27	`structureData.test.js` pass results-over-time structureData SROT-CFG-005: computeVisitOrder falls back to alphanumeric order without an order column (#27) pass results-over-time structureData SROT-CFG-005/006: computeVisitOrder sorts by the numeric order column when present (#27)
SROT-REG-010/012	SROT-REG-010, SROT-REG-012	#27	`structureData.test.js` pass results-over-time structureData SROT-REG-010/012: flagOutliers marks values outside the 5th/95th percentiles only when enabled (#27)
SROT-REG-002/010 (grouping-change regression)	SROT-REG-002, SROT-REG-010	#27	`structureData.test.js` pass results-over-time structureData SROT-REG-002/010: flagOutliers recomputes each row group when the grouping changes between renders (#27) pass results-over-time structureData SROT-REG-010/012: flagOutliers marks values outside the 5th/95th percentiles only when enabled (#27)
SROT-REG-021	SROT-REG-021	#27	`structureData.test.js` pass results-over-time structureData SROT-REG-021: measureLabel appends the unit when present (#27)
SROT-REG-015/016/017/020	SROT-REG-015, SROT-REG-016, SROT-REG-017, SROT-REG-020	#27	`getScales.test.js` pass results-over-time getScales SROT-REG-015: statPrecisions gives min/max the base, quantiles +1, and StDev +2 (#27) pass results-over-time getScales SROT-REG-015: yPrecision follows the original log10-range rule (#27) pass results-over-time getScales SROT-REG-016/020: resolveYDomain uses the data extent unless a limit overrides it (#27) pass results-over-time getScales SROT-REG-017: normalizeDomain swaps a crossed lower/upper pair in place (#27)
SROT-REG-003/011/014/015	SROT-REG-003, SROT-REG-011, SROT-REG-014, SROT-REG-015	#27	`getPlugins.test.js` pass results-over-time getPlugins SROT-REG-003: groupColors assigns a stable color per group in order (#27) pass results-over-time getPlugins SROT-REG-011: outlierTooltip names the participant and its value (#27) pass results-over-time getPlugins SROT-REG-014/015: summaryTooltip lists every statistic at the required precision (#27) pass results-over-time getScales SROT-REG-015: statPrecisions gives min/max the base, quantiles +1, and StDev +2 (#27) pass results-over-time getScales SROT-REG-015: yPrecision follows the original log10-range rule (#27) pass results-over-time structureData SROT-REG-014/015: summarize returns the tooltip statistics for a sorted sample (#27)

Visual evidence

Every screenshot below is a committed baseline: the same PNG is the visual-regression baseline the browser suite asserts against and the evidence artifact shown here. Click any capture for the full-resolution image.

SROT DATA 002 invalid data note
SROT FUNC 001 control panel
SROT FUNC 003 participant count
SROT FUNC 004 y limits
SROT FUNC 006 visits without data
SROT FUNC 007 unscheduled visits
SROT FUNC 008 box plots
SROT REG 002 grouped box plots
SROT REG 010 outliers
SROT REG 018 log scale

Source-matrix routing status (58 rows)

Covered: 44 of 58 rows (the browser and unit tables above). The remaining 14 rows are descoped, honestly, for the reasons below.

Violin plots — SROT-FUNC-009, SROT-REG-008 (2 rows): descoped. Violin (kernel-density) marks are off by default in the original renderer and add little over the box-and-whisker distribution summary; Chart.js has no violin primitive, so they would be a large custom-SVG effort for a non-default feature. Box-and-whisker + outliers deliver the population distribution.
Legacy Webcharts / CAT / API rows — SROT-CFG-001, SROT-CFG-002, SROT-API-001, SROT-REG-022, SROT-REG-025 (5 rows): descoped. These describe the legacy Webcharts configuration object, the safetyResultsOverTime() Webcharts-chart factory, and CAT-specific settings workflows. The module ships the shared nextgen lifecycle API instead (SROT-API, module scheme) with explicit settings, matching the histogram precedent; the reviewed matrix flags these rows needs-jeremy-review.
Normal-range mappings — SROT-CFG-010, SROT-CFG-011 (2 rows): descoped. normal_col_low/normal_col_high exist in the original settings but the results-over-time view does not draw a normal range (unlike the histogram), so these mappings are not part of this renderer's contract.
missingValues — SROT-CFG-016 (1 row): descoped. Missing/non-numeric results are removed by numeric cleaning with a reported count (SROT-DATA-002); the explicit missing-value token list is not reproduced.
Enrolled/randomized denominator — SROT-REG-013 (1 row): adapted. The participant note reports "shown of total (%)" for the current measure; the enrolled/randomized population denominator is not present in the trimmed example data contract, so it is not shown.
Log-axis tick overlap — SROT-REG-019 (1 row): delegated. Log-scale tick selection and label spacing are handled by Chart.js's logarithmic scale; the module does not add a custom de-duplication pass, so this is not separately asserted.
Group-variable validation — SROT-REG-026, SROT-REG-027 (2 rows): descoped. Filters whose column is absent from the data are dropped with a warning, but group options are taken from settings as-is (missing-column dropping and duplicate-label de-duplication for the Group-by control are not implemented). A follow-up could extend the filter validation to groups.

Reproducing this report

The evidence set is regenerated from a full test run and committed with the code it qualifies; CI fails when they drift. To verify or rebuild it:

npm ci
npm run evidence:check   # compare a fresh run against the committed evidence
npm run evidence         # regenerate docs/evidence/results-over-time/evidence.json

Screenshot baselines are canonical to the Linux CI runner; the repository's Update evidence baselines workflow is the authoritative way to refresh them. See CONTRIBUTING.md for the traceability convention.