Every analytical output this library produces is checked, number by number, against Dr. Thomas A. Bishop’s published Minitab results for the same data. Not “inspired by” and not spot-checked — the chart centers, control limits, signal classifications, capability indices and loss-function components are compared to the reference and the run fails if any of them disagree.
280 assertions pass, 0 fail, 35 have no reference value.
Generated by validation/e2e_bishop_report.py from processbehavior 0.2.0, against validation/PBTESTDATABASE_T100.csv.
Coverage¶
| Analytical Design State | Source column | Assertions | Result |
|---|---|---|---|
| ADS 1 — full replication | PM SDS 1 | 102 | ✅ all pass |
| ADS 2 — no replication | PM SDS 2 | 89 | ✅ all pass |
| ADS 3 — partial replication | PM SDS 3 | 89 | ✅ all pass |
What is covered¶
Charts — center line and both natural process limits, for the primary chart and for each valid chart x residual pair at that design state.
Capability — Pp, Ppk (upper and lower), Cp, Cpk, and the percentage of the distribution beyond each specification limit.
Loss function — the Taguchi loss decomposition and its components.
What is not yet covered¶
ODS 4-6 (incomplete grids) are detected and routed correctly, and their collapse to ADS 1-3 is pinned by tests — an incomplete design is analysed as the complete design that survives tidying, so the numbers above are the ones that apply. What is still outstanding is an end-to-end comparison against Bishop’s Minitab output for the incomplete-grid datasets themselves.
Reproducing this¶
python validation/e2e_bishop_report.pyThat writes this page and a detailed HTML report
(validation/e2e_bishop_report.html) with every compared value, expected
beside actual, grouped by design state.