refactor: archive topo2d.js — tested but not wired to index.html (janitor F2)

topo2d.js is moved to archive/topo2d/ along with its unit test. It
had 71.29% mutation score at time of archive but production
index.html never instantiated EEGTopo2D — F2 from the 2026-05-20
janitor dead-code audit.

Stryker mutate scope updated. Aggregate moves from 40.70% → 35.32%
(topo2d was dragging it up by ~5pp). Threshold: lowered break 38 → 30.

archive/topo2d/README.md documents the restore procedure if a
future PR wires topo2d into the viewer's metadata overlay.

Author: bruAristimunha

archive/topo2d/README.md

@@ -0,0 +1,30 @@
+# topo2d.js — archived 2026-05-21
+
+This module implements an MNE/EEGLAB-style 2D EEG topographic map
+renderer. It has full unit-test coverage (71.29% mutation score at
+time of archive) and works as designed when instantiated.
+
+It is archived rather than deleted because production `index.html`
+never instantiated `EEGTopo2D` — the wiring is incomplete on the
+viewer side (there's no overlay slot, no controller that calls
+`Topo2D.init()`). Janitor finding F2 in the 2026-05-20 dead-code
+audit flagged this.
+
+## Restoring
+
+If a future PR wires topo2d into the viewer UI:
+
+1. `git mv archive/topo2d/topo2d.js ./topo2d.js`
+2. `git mv archive/topo2d/unit-topo2d.test.mjs tests/unit-topo2d.test.mjs`
+3. Add `<script src="topo2d.js?v=1"></script>` to `index.html` between
+   `traces.js` and `viewer.js`.
+4. Add the integration code in viewer.js — minimum: a metadata-overlay
+   slot that calls `EEGTopo2D.init(containerEl)` and `setMontage()` /
+   `setSelected()` on channel selection.
+5. Re-add to `stryker.conf.json`'s mutate list and commandRunner.
+
+The original tests still pass against the archived file:
+
+```bash
+node --test archive/topo2d/unit-topo2d.test.mjs
+```

topo2d.js archive/topo2d/topo2d.jstopo2d.js renamed to archive/topo2d/topo2d.js

@@ -1122,3 +1122,48 @@ Neither is on the current Tier 1+2+3 plan; both are documented as
 catching REGRESSIONS in the existing 4 well-mutated files (traces,
 filters, topo2d, bids-recording), not in the placeholder floors for
 viewer + worker.
+
+## Iteration 13 (T11, 2026-05-21) — topo2d archived
+
+Janitor F2 closure: `topo2d.js` was moved to `archive/topo2d/` along
+with its unit test. The module had a high mutation score (71.29%) but
+was never instantiated by production `index.html` — wiring the
+controller into a metadata-overlay slot is UX work that's out of
+scope for the Tier 1+2+3 maturation plan. See
+`archive/topo2d/README.md` for the restore procedure.
+
+### Per-file results (fresh baseline, 3604 mutants, ~18 min runtime)
+
+| File | iter-12 | iter-13 | Δ |
+|---|---:|---:|---:|
+| traces.js          | 66.71% | 66.71% | — |
+| filters.js         | 92.37% | 92.37% | — |
+| topo2d.js          | 71.29% | (archived) | n/a |
+| bids-recording.js  | 76.14% | 74.24% | −1.90 (sampling) |
+| viewer.js          | 0.64% | 0.64% | — |
+| worker.js          | 4.56% | 4.56% | — |
+| **Aggregate**      | 40.70% | **35.32%** | **−5.38** |
+
+### Why the aggregate dropped 5.38pp
+
+`topo2d.js` was carrying 613 mutants at 71.29% — well above the
+aggregate. Removing it drops the high-scoring contribution while
+leaving the low-scoring floors (viewer 0.64%, worker 4.56%) at full
+weight. The 5pp drop is arithmetic, not a regression: per-file
+scores for the four remaining real-tested files (traces, filters,
+bids-recording, worker) are unchanged within sampling noise.
+
+bids-recording's small dip (76.14 → 74.24) is sampling noise from
+Stryker's non-deterministic mutant ordering; no test or source
+changes were made.
+
+### Threshold: lowered 38 → 30
+
+Per the documented decision tree: aggregate (35.32%) is below the
+current break (38), so lower to `aggregate − 5 = 30`. This is NOT a
+regression in the actual code — the gate is recalibrated for the
+narrower scope (5 files: traces, filters, bids-recording, viewer,
+worker). The per-file scores for the well-tested files remain the
+durable measurement.
+
+low threshold also lowered: 42 → 34 (mirrors the previous gap).

tests/unit-topo2d.test.mjs archive/topo2d/unit-topo2d.test.mjstests/unit-topo2d.test.mjs renamed to archive/topo2d/unit-topo2d.test.mjs

@@ -1,17 +1,16 @@
 {
   "$schema": "./node_modules/@stryker-mutator/core/schema/stryker-schema.json",
-  "_comment": "Mutation testing config — see docs/qa-strategy.md § Mutation testing. Scope (v4): traces.js (render path) + filters.js (signal processing) + topo2d.js (electrode-position layout) + bids-recording.js (metadata loader) + viewer.js (JSDOM smoke + partial pure-logic coverage) + worker.js (self-shim smoke + protocol re-implementation harness). Per-file scores are the durable measurement; aggregate is informational.",
+  "_comment": "Mutation testing config — see docs/qa-strategy.md § Mutation testing. Scope (v5): traces.js (render path) + filters.js (signal processing) + bids-recording.js (metadata loader) + viewer.js (JSDOM smoke + partial pure-logic coverage) + worker.js (self-shim smoke + protocol re-implementation harness). topo2d.js was archived 2026-05-21 (janitor F2, not wired into index.html). Per-file scores are the durable measurement; aggregate is informational.",
   "packageManager": "npm",
   "reporters": ["html", "clear-text", "progress", "json"],
   "testRunner": "command",
   "commandRunner": {
-    "command": "node --test tests/unit-traces-draw.test.mjs tests/unit-traces-partial-fill.test.mjs tests/unit-traces-nice-round.test.mjs tests/unit-traces-time-axis.test.mjs tests/unit-traces-scalebar-axis.test.mjs tests/traces.test.mjs tests/unit-filters.test.mjs tests/unit-topo2d.test.mjs tests/unit-bids-recording.test.mjs tests/unit-bids-recording-assemble.test.mjs tests/unit-viewer-jsdom.test.mjs tests/unit-viewer-render-loop.test.mjs tests/unit-viewer-races.test.mjs tests/unit-worker-jsdom.test.mjs tests/unit-worker-protocol.test.mjs tests/unit-worker-faults.test.mjs tests/unit-worker-cache.test.mjs tests/unit-worker-races.test.mjs"
+    "command": "node --test tests/unit-traces-draw.test.mjs tests/unit-traces-partial-fill.test.mjs tests/unit-traces-nice-round.test.mjs tests/unit-traces-time-axis.test.mjs tests/unit-traces-scalebar-axis.test.mjs tests/traces.test.mjs tests/unit-filters.test.mjs tests/unit-bids-recording.test.mjs tests/unit-bids-recording-assemble.test.mjs tests/unit-viewer-jsdom.test.mjs tests/unit-viewer-render-loop.test.mjs tests/unit-viewer-races.test.mjs tests/unit-worker-jsdom.test.mjs tests/unit-worker-protocol.test.mjs tests/unit-worker-faults.test.mjs tests/unit-worker-cache.test.mjs tests/unit-worker-races.test.mjs"
   },
   "coverageAnalysis": "off",
   "mutate": [
     "traces.js",
     "filters.js",
-    "topo2d.js",
     "bids-recording.js",
     "viewer.js",
     "worker.js"
@@ -31,8 +30,8 @@
   "incrementalFile": "reports/stryker-incremental.json",
   "thresholds": {
     "high": 80,
-    "low": 42,
-    "break": 38
+    "low": 34,
+    "break": 30
   },
   "timeoutMS": 60000,
   "concurrency": 4

docs/mutation-survivors-2026-05.md

@@ -75,12 +75,12 @@ const EXPECTED = {
     'lastTotalChannels',
     'meanStd',
   ],
-  // topo2d.js exports via `window.EEGTopo2D = api` only — no
-  // module.exports fallback. Same export-pattern bug as fiff.js had
-  // pre-c57dc88. Documented as janitor F2 (file loaded by tests but
-  // NOT by production index.html). When F2 is decided (wire in or
-  // archive), add topo2d back to this snapshot OR remove from
-  // tests/unit-topo2d.test.mjs.
+  // topo2d.js was archived 2026-05-21 to archive/topo2d/topo2d.js
+  // (janitor F2 closure — file was tested but never instantiated by
+  // production index.html). If a future PR wires topo2d into the
+  // viewer UI, restore per archive/topo2d/README.md and re-add the
+  // entry here. The original export pattern (window.EEGTopo2D = api,
+  // no module.exports fallback) was the same as fiff.js pre-c57dc88.
   '../bids-loader.js': [
     'parseCoordsystem',
     'parseElectrodesTSV',