Issues

• Follow-ups found while fixing the 5 Nov bugs (PRs #1155 to #1168)

  bug
  kgdunn/process-improve#1169 · Filter by author kgdunn opened on Nov 5, 2026

• TPLS: a component that never converges stops at max_iter without a warning, and its result depends on the platform

  area:multivariatebug
  kgdunn/process-improve#1163 · Filter by author kgdunn opened on Nov 5, 2026

• Reference catalogue: carry the pid-book source forward from book commit 28e5b88 to the current book

  area:tests
  kgdunn/process-improve#1154 · Filter by author kgdunn opened on Nov 5, 2026

• Multivariate family: sweep every estimator for consistency, then lock it in with contract tests

  area:multivariatearea:testsenhancement
  kgdunn/process-improve#1149 · Filter by author kgdunn opened on Nov 4, 2026
  3

• examples/multivariate/tpls_new_materials: not run in CI, and one README number has no script

  bug
  kgdunn/process-improve#1145 · Filter by author kgdunn opened on Nov 4, 2026

• Multivariate follow-ups from round 2: missing cells in PCA/PLS, contributions for batch and OPLS, counts and edge cases

  area:multivariatebug
  kgdunn/process-improve#1144 · Filter by author kgdunn opened on Nov 4, 2026

• robust_regression's 3.0 switch to bootstrap intervals: a reminder test, the old numbers under another name, conflevel checks

  bug
  kgdunn/process-improve#1132 · Filter by author kgdunn opened on Nov 3, 2026

• Hosted tool calls: a spawned worker's start-up uses most of the 10 s timeout, and robust regression above about 1,500 points still times out

  bug
  kgdunn/process-improve#1130 · Filter by author kgdunn opened on Nov 3, 2026

• Experiments follow-ups: K-optimal quality varies by BLAS kernel, mixture model_selection searches the wrong terms, the DSD warning on OMARS, Block as a categorical

  bug
  kgdunn/process-improve#1129 · Filter by author kgdunn opened on Nov 3, 2026

• Multivariate follow-ups from the 2026-11-02 decision PRs: missing data, error messages, PRM floors, an unused copy

  area:multivariatebug
  kgdunn/process-improve#1114 · Filter by author kgdunn opened on Nov 3, 2026
  1

• Batch: bring back the sklearn transformers ResampleAligner, DTWAligner and BatchFeatureExtractor from #461

  enhancement
  kgdunn/process-improve#1105 · Filter by author kgdunn opened on Nov 2, 2026

• Sensory excellence: redesign the sensory toolkit

  enhancement
  kgdunn/process-improve#1103 · Filter by author kgdunn opened on Nov 2, 2026
  1

• Catalogue and CI follow-ups: Montgomery's Bartlett test as checks, one BHH cause, and an import without extras

  area:testsenhancement
  kgdunn/process-improve#1100 · Filter by author kgdunn opened on Nov 2, 2026

• sensory_panel_check: NaN F-tests for a MAM with a missing cell, without the warning text

  bug
  kgdunn/process-improve#1099 · Filter by author kgdunn opened on Nov 2, 2026

• Factorial analysis: no agent tools for the Yates and three-level component tables, and no user-guide page

  area:experimentsdocumentationenhancement
  kgdunn/process-improve#1097 · Filter by author kgdunn opened on Nov 2, 2026

• Browser app: a late balance reply after a rename, the time in the downloaded workbook, a grouped ANOVA label, and an e2e race

  bug
  kgdunn/process-improve#1096 · Filter by author kgdunn opened on Nov 2, 2026

• README links are broken

  kgdunn/process-improve#1093 · Filter by author github-actions opened on Nov 2, 2026

• Mutation testing found a change no test catches

  kgdunn/process-improve#1092 · Filter by author github-actions opened on Nov 2, 2026

• Display and labels: MBPCA/MBPLS display_results misaligned; TPLS loadings drop the d_matrix index name

  area:multivariatebug
  kgdunn/process-improve#1083 · Filter by author kgdunn opened on Oct 31, 2026

• Warnings point at the wrong line: MBPCA/MBPLS non-convergence and AdaptivePLS's inner PLS

  area:multivariatebug
  kgdunn/process-improve#1079 · Filter by author kgdunn opened on Oct 31, 2026
  1

• Unclear errors on invalid input: prediction_interval, invert, PLSDA.confusion, empty blocks

  area:multivariatebug
  kgdunn/process-improve#1077 · Filter by author kgdunn opened on Oct 31, 2026

• Make mutation-test kills repeatable: tests with unseeded random folds kill some mutants only by chance

  area:testsenhancement
  kgdunn/process-improve#1073 · Filter by author kgdunn opened on Oct 31, 2026
  1

• PRM and PLS NIPALS read data in small units as having no variance (absolute 1.5e-8 floor)

  area:multivariatebug
  kgdunn/process-improve#1070 · Filter by author kgdunn opened on Oct 31, 2026
  2

• Hotelling's T² limits: rows used to fit the model and new rows need different laws; choose an approach

  kgdunn/process-improve#978 · Filter by author kgdunn opened on Oct 25, 2026
  1

• Mixture designs and models beyond the simplex lattice: the full simplex centroid, axial blends at a chosen distance, McLean and Anderson's design, U-pseudocomponents, ratio variables, component-amount designs, orthogonal blocks (Box and Draper chapters 16 to 18)

  area:experimentsenhancement
  kgdunn/process-improve#973 · Filter by author kgdunn opened on Oct 25, 2026