One page, one statistics pipeline: analyzeWaferMap / analyzeWaferLot produce a summary, and that summary is handed to a renderer as statsSummary or lotStatsSummary. The three scopes below are the only real difference between them — the analysis call and the display options are otherwise identical, which is why they share a page.

Single wafer — analyzeWaferMap scans five families: rings, quadrants, angular sectors (16 compass directions), failure clusters (contiguous failing dies denser than background), and edge arcs. This wafer has a particle/ESD cluster, a handling arc near the perimeter, and the natural radial gradient that drives ring and sector findings. Click any finding to highlight the affected dies.
Lot gallery — analyzeWaferLot surfaces patterns that repeat across wafers and flags yield outliers. W01, W03 and W05 share the same edge-ring failure (detected as a repeated lot pattern); W06 has stacked failures and is flagged an outlier.
Lot stack — buildWaferMap({ lotStack }) aggregates the lot into one map, then analyzeWaferMap runs on the aggregate. Use this when you need a standalone stacked map outside a gallery, or programmatic access to the findings before rendering. The NE-quadrant Vth drift present on every wafer survives averaging and dominates the result.