v0.33.0

Breaking

• sectorCount is 4, 8 or 16. The compass names cover 16 bearings, so 32 has no names to give its sectors;
  it is corrected to the default with an 'analysis-option-corrected' warning, like any other unusable value.

Changed

• An expanded map's window widens for the Summary panel. Opening the panel in a map's Expand modal, or in a wafer
  opened from a gallery, makes the window wider by the room the panel needs, so the map keeps its size instead of
  giving up more than half its width; closing the panel narrows it again. The window stays within the screen (a
  floating window moves left if it would run off the edge), does nothing when maximised, and leaves a width the user
  has dragged it to. A wafer opened in a separate browser window is resized the same way where the browser allows it.
• The report builders load when a report is opened. The Summary panel's report button and the gallery's load
  renderSummaryReport with a dynamic import() instead of with the map, which takes about 10 KB gzipped out of what
  rendering a map downloads (core is ~143 KB with this release's additions); the first report opens a moment later while it loads.
  renderWaferReportHtml and renderLotReportHtml are unchanged public exports.
• The Summary panel leads into the report, and the report acts on the map. "What stands out" ends with a Full
  report link that opens the same report the Summary report button does, which begins with that same section. In
  a report opened from the panel, a click on a finding's row closes it and shows that finding on the map, as its
  row in the panel does (on a lot report that is split into several groups, a finding the panel does not hold is
  ignored). Opened on its own, the report is a plain page and its rows do nothing.
• The reports and the Summary panel count and order bins, findings and statistics the same way. One count of dies
  per bin (partial and edge-excluded dies left out) and one order (pass bins first, then fail bins by count) feed the
  panel's bars and the reports' tables, under one title that states the population ("Hard Bin Breakdown — % of dies
  (N=…)"); the reports show the soft-bin breakdown as well when the data has soft bins, in the same columns for a
  wafer and a lot. The reports' findings table is arranged as the panel's list is: each spatial pattern with the
  findings it explains beneath it, then the rest by region, most severe first. The wafer report's Test Values table
  has the panel's full columns (N, min, quartiles, median, mean, max, σ) and the share of dies inside each test's
  limits; a lot's has N, min, mean, max, σ (pooled exactly) and limit yield, and says why it has no quartiles. The
  lot overview names the unweighted mean of wafer yields "Mean per-wafer yield" everywhere.
• The Summary panel and the HTML reports read yields the same way. One computation gives each wafer's
  difference from the lot median (with the lot's outlier rule naming the outliers) and each ring or quadrant's
  difference from the wafer or lot it divides. The panel's wafer and region rows print that difference in points
  and tint a shortfall of 1, 2 or 4 or more points, as the report's tables do; the report's wafer table says
  "low outlier" in words, as the panel always has. The panel's findings mark severity as three dots filled by level
  rather than a colour-only dot, as the reports do. The reports draw bin bars in the colour the map gives that bin
  (the live colours when opened from a panel, the default palette otherwise).
• The Summary panel opens with "What stands out", the same synthesis the HTML reports open with (a headline, up to
  three items ranked by the dies each costs, an "also" line, and what was compared), in place of the findings
  narrative above the findings list. It is a section of its own, so a clean wafer or lot says "Nothing stands out"
  where the findings list is absent. The names in it (a region, a bin) select that finding on the map; the Findings
  section's Detail button opens the same sentences above the full list. Where the wafers in a lot were judged by
  different pass bins, the items rest on yield alone and name no pass bins.
• The weakest finding severity is called "Minor" where it is shown (the Summary panel's severity filter and the
  reports); the value in StatsFinding.severity is unchanged.
• A merged region on a wafer is graded with its constituents' multiple-testing correction. The merged finding's
  p-value is multiplied by the weakest constituent's adjusted-to-raw ratio before severity is assigned, the same rule the
  lot merge applies, and stats.adjustedPValue is set. A merge can no longer grade more severe than the correction its
  parts carry allows, so a few merged findings show a lower severity.
• A lot stack by standard deviation or count carries no test limits. Its values are spreads and tallies, not
  measurements of the test, so result.testDefs has the limits removed and nothing is judged against them: no
  limit fail findings, limit yield, capability index or out-of-spec colouring. Stacks by mean, median, minimum
  and maximum keep them.
• A sector run and a quadrant over the same part of the wafer are one finding. "Sectors W–S" and
  "Quadrant SW" make one statement, so the one with the smaller p-value is shown and the other is listed
  under it (absorbedIds). They must be the same comparison (metric, variable and direction) and cover
  angles that overlap by at least three fifths of their union, by geometry. A single sector inside a
  quadrant, and a mean against a limit fail rate, stay separate. Wafer and lot findings follow the same rule.
• Lot findings merge adjacent regions. A run of adjacent sectors, quadrants or rings that carries one
  signal is one lot finding ("Sectors E–N", "Rings 1–2"), as on a single wafer. Each wafer's regions are
  added up, the rest being the wafer's other regions, and those per-wafer figures are combined across the
  wafers as the single regions are (Stouffer's Z, the same gates). The run is adjusted with the
  Benjamini–Hochberg multiplier of its weakest region and is reported only if it passes on its own;
  otherwise its separate rows stay. "N/M wafers" counts the wafers where the merged region itself differs at
  the analysis's significance level.
• A merged run is compared with the rest of its region family. As for each region on its own: sectors
  leave out the centre dies, so a merged sector run's "rest of the map" is the other sectors' dies.
• Adjacent regions merge per metric. A test's mean finding and its limit fail rate are kept apart when
  adjacent sectors, quadrants and rings are merged, so a run such as sectors E, NE and N reads as one
  finding ("Sectors E–N"), and a quadrant with both a mean and a limit finding lists both. A run of limit
  fail rates merges too, recomputed over the union of its regions. A run of quadrants is named in order
  round the wafer.
• A lot's spatial pattern is decided from the lot. analyzeWaferLot stacks the lot's wafers by die
  position, finds the positions where failures recur (an exact binomial test on each position's 3×3
  neighbourhood, Benjamini–Hochberg across positions), classifies that area with the same classifier and
  thresholds as a single wafer, and reports the pattern once, on every wafer whose failures are concentrated
  in it. A pattern visible on every wafer is reported on every wafer under one label, with the ring, edge-arc
  and cluster findings it explains listed under it. Wafers on a different grid, and lots too small for
  failures to recur significantly, keep the per-wafer count. The wafer-level classifier and its thresholds
  are unchanged (WM-811K: 64.1% overall).
• Outlier wafers have one rule, read by both the lot findings and the Summary panel's
  Wafer Yield list, so they always name the same wafers. From 3 to 7 wafers it is Dixon's Q
  test on the lowest and highest wafer (95% → notable, 99% → unusual); from 8 wafers,
  Tukey's fences over per-wafer yield (1.5 × IQR → notable, 3 × IQR → unusual). Either
  way the wafer must also be at least 3 yield points from the median. The
  finding's stats.method is 'dixon-q' or 'tukey-fence', and effect.effectSize is
  that test's statistic (Q, or the distance from the median in IQR units). The Wafer Yield
  list labels high outliers as well as low ones.

Added

• Compact layout for multi-project wafers. A Compact layout row in the Overlays menu, on a single map and on the
  gallery, draws the dies on a grid with the empty rows and columns removed and each group of dies outlined, so a
  sparse layout fills the map instead of leaving each die a few pixels wide. It is offered when the occupied columns
  and rows repeat at a regular pitch, or at the reticleConfig width and height, or a multiple of it (a product on
  every second reticle), when every wafer supplies the same ones; viewOptions: { compact: true } applies it regardless. Only the layout changes:
  every die is still drawn and counted, so legends, yield and statistics are the same, and hover text and axis
  labels give original die coordinates. A die with no bin keeps its place. The notch marker follows rotation and
  flips, and so does the XY indicator, which sits in a margin beside the grid. A gallery builds one layout from every
  wafer shown, so its cards stay comparable. The wafer outline, ring, quadrant and reticle overlays describe the
  physical wafer and are not drawn in this layout.
• Rotating or flipping a map switches the XY indicator on. The first Rotate or Flip from the Orientation menu, on a
  single map or a gallery, turns on the +X/+Y arrows in whatever layout is showing, so the reader can see which way
  the die coordinates run. They stay on until switched off in the Overlays menu, and Reset orientation does not
  change them.
• **An Axis labels row in the Ov...

v0.32.0

Breaking

• Input values of the wrong type are treated as missing. A bin, site number or test value
  that is not a number, or a pass/fail verdict that is not true/false, is left out of the
  die — the same rule as a value outside the STDF V4 ranges — and still reported as
  input-values-not-numbers. A die whose only bin was text therefore has no verdict rather
  than a fail, which can change yield for such input. The input objects are not modified.
• Test values on dies from buildWaferMap are read-only snapshots. A map holds test values and
  verdicts as one column per test. die.testValues and die.testPass on a built die build a frozen
  object from those columns on each read, keeping nothing on the die, so die.testValues !== die.testValues. Assigning to either field throws a TypeError; changing a key of the frozen
  object throws in strict-mode code and is ignored otherwise. Pass the values to buildWaferMap,
  or copy the die with your own object ({ ...die, testValues: mine }). A spread, structuredClone or JSON.stringify of a built die
  gives plain objects with the same values. A die no longer shares the input record's
  testValues/testPass objects, so changing the input after the build does not change the map.
  Dies a host builds itself keep ordinary objects.
• Pre-built dies get the same input checks as results. Wrong-type and out-of-range
  values are treated as missing, and a die whose coordinates STDF cannot store is kept as an
  unpositioned die.
• WaferMapResult.view is removed. It was marked @internal: the renderers build their own
  draw list whenever they draw, so a result no longer carries one. Read the result's own fields
  (plotMode, metadata, isLotStack, hbinDefs, sbinDefs, testDefs). This makes each
  result smaller (about 86 bytes less per die) and roughly halves the copy the Web Worker makes of
  a result. The internal dataAxisFlip field takes its place.

Changed

• The legend filters to several bins or metadata values at once. Ctrl/Cmd+click on a legend
  entry — the map's legend or the gallery's strip, where Ctrl/Cmd+Enter/Space works from the
  keyboard — adds or removes a value; a plain click shows only that value, or clears the
  filter when it is the only one shown. highlightBin accepts number | number[] and
  highlightMetadataValue accepts string | string[].
• A finding and the legend filter agree. Clicking a finding about a bin filters the legend
  to that bin, and one about yield or a test clears the filter — in a single map and in the
  gallery alike. Changing the legend filter releases the finding, as changing the selection
  does, including a selection made on a gallery card while a lot finding is active.
• Lot regional findings are tested on all wafers' data together. A yield, bin,
  functional pass-rate, limit-fail or test-value difference in a region is combined across
  every wafer (Stouffer's Z over each wafer's own test, weighted by die count) and reported
  with the wafer analysis's gates, redundancy collapse and opposite-region re-test, instead of
  by counting wafers whose own analysis reported it — a pattern present on every wafer but
  too faint on some to pass alone is the lot's pattern. The sentence gives the lot's figure
  and "higher/lower on N/M wafers, all wafers' data combined", N counting the wafers whose
  region differs in that direction; stats.method is 'stouffer-z'. Clusters, edge arcs and
  spatial-pattern labels are still counted by the wafers that report them. Lot findings keep
  absorbed restatements in the list, marked by absorbedIds, as wafer findings do.
• Selected dies are shown by fading the rest of the map. Every unselected die is faded
  towards the map background and the selection is outlined, so selected dies keep their full
  colour and it is clear which dies are selected whatever the selection's shape — a block, a
  ring, an edge arc. A finding highlighted from the Summary panel, a gallery lot finding and
  setSelection are drawn the same way. The fade and outline sit under the axes and legend.
• A map opens in select mode. A drag draws a selection box; hold Space and drag, use the
  arrow keys, or choose Pan in the toolbar to pan.
• Clicking the only selected die again clears the selection. Clicking a die inside a
  larger selection still selects just that die.
• Changing the selection on the map releases an active finding. A click, box select,
  right-click, Esc or clearSelection clears the Summary panel's active finding and its bin
  highlight, so the panel never shows a finding the map no longer highlights.

Fixed

• Regional findings report patterns confined to one region. A bin, limit-fail or
  functional-fail rate that is zero in the rest of the wafer and raised in a region (a bin
  found only at the edge) counts as the largest relative change for the effect gate and for
  severity, for every region family: rings, quadrants, sectors, reticle positions and test
  sites. On a lot whose wafers carry bin 2 only at the edge, the lot finding reads "seen on
  8/8 wafers".
• Yield and functional pass-rate findings are judged by their failure rate. The relative
  effect of a pass rate is measured on its failures: yield 98% → 94% is failures 2% → 6%, and
  reaches the same verdict as the bin rate of the same dies.
• A region is not reported as deviating only because another region deviates more. A
  finding opposite in direction to a stronger finding for the same variable and region family
  must still hold when compared with the rest of the wafer without that region, so an edge
  rich in a bin, a limit fail or high test values no longer makes the inner rings read as low
  in it (or high in yield).
• The soft bin that restates yield is absorbed into the yield finding, as the hard pass bin
  is: the one soft bin every passing die carries and no failing die does.
• An edge ring made of scattered fails is classified as an edge ring. The spatial-pattern
  classifier recognises an edge ring from the failing dies' positions — most fails at the rim
  and spread round at least 60% of it — when no connected group of fails is large enough to
  judge shape by, and before the scratch rule, so a short run along the rim is not a scratch.
  Against WM-811K: detection 86.4%, edge-ring recall 75% (precision 92%), scratch recall 24%.
• Lot spatial-pattern rows count related labels together and state the lot's figures.
  Edge-ring and edge-local wafers count as one edge pattern, and centre and donut as one,
  naming each label's wafer count ("Spatial pattern: edge (edge-ring on 5, edge-local on 2) —
  seen on 7/13 wafers"); the sentence gives the lot's count, not any one wafer's confidence or figures.
  The gallery highlights each counted wafer's failing dies.
• Findings, their severities and "seen on N wafers" counts can change.

Performance

• Insights no longer blocks while its test statistics are computed. The Overview's Test Values
  table is built in slices, showing "Computing test statistics…" until it is ready, and when the
  gallery's summary panel is computing the same statistics at the time, the two share the one
  calculation instead of each doing it.
• Gallery cards are drawn only when on screen. A card below the fold keeps its map up to date
  and is drawn as it scrolls into view. Printing and the gallery PNG draw every card first, so
  both still show the whole gallery. On a 25-wafer, 266k-die lot, with 4 cards on screen:
  switching to value mode takes 0.4 s in Chrome (was 1.1 s) and 0.5 s in WebKit (was 2.0 s);
  the gallery opens in 1.8 s and 2.0 s.
• The gallery draws about twice as fast in WebKit (the desktop app on Linux and macOS, and
  Safari). Die fills and outlines are drawn as many small canvas paths rather than one per
  colour, which WebKit rasterises far faster. On a 25-wafer, 266k-die lot the gallery opens in
  4.0 s (was 7.0 s) and a switch to value or stacked mode takes 1.8–2.3 s (was 3.8–4.4 s).
  Chrome draws identical pixels; in WebKit only anti-aliased edge pixels differ.
• The lot Summary report opens about five times faster on a large lot. Its Test Values table
  finds each test's minimum, mean and maximum in one pass instead of sorting every die's value.
  On a 25-wafer, 266k-die lot the "Summary report" button takes 0.4 s in Chrome (was 6.0 s) and
  0.9 s from click to report in WebKit (was 4.7 s). The report is unchanged (compared as HTML).
  Boxplot quartiles and the histogram's outlier range are found the same way as the Summary
  panel's, by selection, with the same figures.
• Test-value analysis (enableTestValueAnalysis) is two to three times faster on large lots.
  Spec-limit findings count each region's dies in one pass instead of re-reading every die for
  every region, the test list is read column by column, and per-test statistics find their
  quartiles by selection. Findings are unchanged (compared byte for byte on a 266k-die lot with
  5,289 findings). On that lot: 18.9 s → 7.2 s in Chrome, 25.6 s → 10.7 s in WebKit.
• Analysis is about a third faster in Chrome on large lots. Merging findings in adjacent
  regions, pairing hard and soft bins that cover the same dies, and assigning dies to regions
  now compare die positions as numbers rather than as a text key per die. Findings are
  unchanged (compared byte for byte on a 266k-die lot). On that lot, load-time analysis takes
  2.5 s in Chrome (was 3.7 s).
• Stacked modes and mode switches in the gallery are faster. Stacked cards key die positions by
  number rather than by a string per die, and a map no longer rebuilds a key for every die on
  each redraw unless dies are selected. On a 25-wafer, 266k-die lot in Chrome: switching to
  stacked bins 2.0 s → 1.6 s, to value mode 1.3 s → 1.1 s.
• The lot summary panel's test statistics are computed about three times faster. Th...

v0.31.0

Breaking

• The 74 deprecated exports are removed: the 73 deprecated in 0.30.0 (the low-level drawing
  pipeline, the chart-data builders, the region builders and the helpers exported by accident)
  and renderFindingsReportHtml, deprecated in 0.30.3. Each named its replacement in a console
  notice. Upgrading lists every name with
  what to use instead; the main replacements:

  Removed	Use instead
  resolveBinColors, getBinColorScheme	binColorsForMaps(results), or getBinColors() on a map or gallery controller
  valueToViridis, valueToGreyscale, getValueColorScheme	resolveValueColorFn(name, reversed)
  buildYieldData, buildYieldDataCombined	lotYieldSeries on analyzeWaferLot's result
  buildBinParetoData, buildBinClusterData	stats.hardBinCounts, stats.softBinCounts
  buildCapabilityData, buildTestBoxplotData, buildTestTrendData, trendCentre	stats.capability, stats.perTestStats, perWaferTestStats (with computePerTestStats)
  buildTestPassRateData, hasJudgeableTests, computeFunctionalYield	stats.testSpecYield, stats.testFlagYield, stats.functionalYield
  buildRegionYieldData, buildRingRegions, buildQuadrantRegions, classifyDie, getRingLabel	stats.regionYield
  classifyPattern	stats.spatialPattern
  renderSummaryReportHtml, renderLotSummaryReportHtml, renderFindingsReportHtml	renderWaferReportHtml(result, summary?), renderLotReportHtml(results)
  openHtmlReport	openReportModal(html), setReportOpener
  createWafer, generateDies, clipDiesToWafer	buildWaferMap({ layout: true, waferConfig, dieConfig })
  aggregateValues, aggregateBinCounts, getUniqueBins	buildWaferMap's lotStack
  STANDARD_WAFER_DIAMETERS_MM, resolveGridPitch	buildWaferMap's standardDiameters; each die's width/height
  getDieTestValue, isParametricTest, isPositionedDie, metadataCategoricalValue	die.testValues?.[n], testType !== 'F', hasPosition, metadataDisplayValue

  buildView, toCanvas, buildHoverText, buildMapTitle, the transform and affine*
  helpers, the histogram, scatter and correlation data builders, the sector, reticle and
  test-site region builders, parseRegionKey, contrastTextColor, dieHasTestData,
  resolveMetadataColumns, discoverDieMetadataKeys, buildDieListSection and
  DEFAULT_FACET_CURATION have no public replacement: the renderers and the analysis do that
  work themselves.

• 63 types that belonged only to removed functions are removed with them, among them the
  chart-data types, the draw-list types (View, ViewOptions, ViewRect and the rest),
  PitchResult, ToCanvasResult, HitTarget, DieListOptions and the geometry inputs
  WaferSpec, DieSpec and ReticleSpec. Upgrading lists them all. PlotMode stays.

• buildWaferMap takes one argument. The second (WaferMapOptions) set only the starting
  result.plotMode, which the Web Worker never passed, so a build on and off the main thread
  could start in different modes. Set the starting mode with renderWaferMap's or
  renderWaferGallery's viewOptions.plotMode.

• downloadFilename is a prefix for every saved file, on renderWaferMap and
  renderWaferGallery: <downloadFilename>_W05_hard-bin.png, <downloadFilename>_W05_die-list.csv.
  It applies to CSVs, charts, gallery cards and detached windows as well as the map's or gallery's
  PNG, which it no longer names outright. Parts the prefix already names are not repeated. A host
  that matched the exact name its onSaveImage receives should match on the prefix.

• Values outside the STDF V4 ranges are treated as missing. 0.30.4 reported them with the
  input-values-outside-stdf warning and used them as given; buildWaferMap now leaves them out,
  and the warning says what it did:

  • a bin outside 0–32767 or not a whole number: the die has no bin (so it is neither pass nor fail);
  • a coordinate outside ±32767 or not a whole number: the die has no position, in either axis;
  • a test number outside 0–4294967295: that test is left out of every die and of testDefs;
  • a test value that is not finite: that value is left out;
  • a site number outside 0–255: the die has no site;
  • a waferConfig.orientation other than 0, 90, 180 or 270 (or an equivalent such as −90): the
    map is built at 0. WaferConfig.orientation is typed 0 | 90 | 180 | 270.

  The caller's input objects are not modified. A derived test whose testNumber is outside
  0–4294967295 is dropped with a derived-test-invalid warning.

• WaferViewOptions.showPartialDies and isYieldEligibleDie's includePartial are removed.
  No map buildWaferMap builds has partial dies, so neither had anything to act on. A saved
  preference that still carries showPartialDies is ignored. Partial dies a host supplies
  itself are drawn in muted grey and always left out of yield.

Added

• Insights charts draw spec limits as well as test limits. The boxplot, histogram,
  wafer-to-wafer trend and scatter show a test's spec limits (specLow/specHigh, labelled
  LSL/USL, long dashes) beside its test limits (Lo/Hi limit, short dashes). When a test has
  both, a Limits choice (Test + spec, the default; Test limits; Spec limits; None) applies to
  all four charts. Labels that would overlap move to a second row, Axis includes limits
  covers every limit shown, and the scatter marks limits outside its plotted range at the edge,
  as the other charts do. The wafer map still judges pass/fail by the test limits.
• Chart gridlines are lighter (the border colour at 40% opacity), so the data and the limit
  lines stand out from the grid in every theme. The scatter draws its limits in the same amber
  as the other charts, and its limit labels sit on a panel-coloured backing where lines cross
  them.

Changed

• The data-and-stats layer is ~49 KB gzipped, from ~61 KB, with the removed exports gone.

Fixed

• Charts opened from a gallery card's right-click menu are saved under the card's lot and
  wafer, as they are from a single map.

Documentation

• New page, Upgrading: what to change
  for each breaking release, starting with 0.31.0.
• Every version has a GitHub release, created when its tag is pushed, with that version's
  changelog section as its notes.

---

npm: @wafertools/wafermap@0.31.0 ·
Upgrading ·
What's New ·
Full changelog