All pull requests

Pull requests

• [R14 integration] Full Python port stack + live R cache + source-faithful fixes (DRAFT)

  workflow-change-approved
  #192·OVVO-Financial opened Nov 6, 2026Owner·

  ·
  6

• [Audit] Measure full Python port stack against rebuilt R 14.0 parity cache

  workflow-change-approved
  #191·OVVO-Financial opened Nov 6, 2026Owner·

  ·
  2

• Regenerate R parity cache for NNS 14.0

  #190·OVVO-Financial opened Nov 5, 2026Owner·

  ·
  2

• partial moments: co-moments reject empty input and recycled targets

  workflow-change-approved
  #189·gitRasheed opened Nov 3, 2026Collaborator·

  ·

• meboot: an exactly linear or constant series no longer errors with rho

  workflow-change-approved
  #188·gitRasheed opened Nov 3, 2026Collaborator·

  ·

• ss: confidence intervals work when a group is constant

  workflow-change-approved
  #187·gitRasheed opened Nov 3, 2026Collaborator·

  ·

• diff: dy_dx uses every bandwidth when small samples repeat one

  workflow-change-approved
  #186·gitRasheed opened Nov 3, 2026Collaborator·

  ·

• mc: exp keeps the rho grid inside lower_rho and upper_rho

  workflow-change-approved
  #185·gitRasheed opened Nov 3, 2026Collaborator·

  ·

• sd: cluster keeps hclust's leaf order, and the variables when nothing is peeled

  workflow-change-approved
  #184·gitRasheed opened Nov 3, 2026Collaborator·

  ·

• sd: efficient-set membership no longer depends on column order

  workflow-change-approved
  #183·gitRasheed opened Nov 3, 2026Collaborator·

  ·

• arma: decide negative_values from the training data only

  workflow-change-approved
  #182·gitRasheed opened Nov 3, 2026Collaborator·

  ·

• cdf: hazard and cumulative hazard at a target match the returned curve

  workflow-change-approved
  #181·gitRasheed opened Nov 3, 2026Collaborator·

  ·

• var: integer-degree VaR roots no longer depend on the data's units

  workflow-change-approved
  #180·gitRasheed opened Nov 3, 2026Collaborator·

  ·

• copula: 0, not 1, when every margin lies on one side of its target

  workflow-change-approved
  #179·gitRasheed opened Nov 3, 2026Collaborator·

  ·

• mode: break tied discrete modes by gravity, not their mean (mirrors R #111)

  workflow-change-approved
  #178·gitRasheed opened Nov 3, 2026Collaborator·

  ·

• nns_stack: score only the deployed fit for a single predictor (book ch24)

  workflow-change-approved
  #176·gitRasheed opened Nov 2, 2026Collaborator·

  ·

• nns_anova: pooled grand mean and median, median robust estimate (book ch15)

  workflow-change-approved
  #175·gitRasheed opened Nov 2, 2026Collaborator·

  ·

• nns_arma_optim: errors, bias and intervals from the returned model (book ch25)

  workflow-change-approved
  #174·gitRasheed opened Nov 2, 2026Collaborator·

  ·

• nns_arma: lower prediction bound is LPM.VaR without abs() (book ch17, ch26)

  workflow-change-approved
  #173·gitRasheed opened Nov 2, 2026Collaborator·

  ·

• nns_meboot: keep the residual scale when the drift changes (book ch20)

  workflow-change-approved
  #172·gitRasheed opened Nov 2, 2026Collaborator·

  ·

• nns_copula: independence anchors from the empirical marginals (book ch12)

  workflow-change-approved
  #171·gitRasheed opened Nov 2, 2026Collaborator·

  ·

• nns_seas: score every component series of each lag (book ch25)

  workflow-change-approved
  #170·gitRasheed opened Nov 2, 2026Collaborator·

  ·

• Vendor NNS-core from the R-fixes follow-up and drop the M.reg underflow guard

  workflow-change-approved
  #169·gitRasheed opened Nov 2, 2026Collaborator·

  ·

• nns_diff stop rule and the stack ALL candidate: mirror the later R fixes

  workflow-change-approved
  #168·gitRasheed opened Nov 2, 2026Collaborator·

  ·

• Stochastic dominance: midpoint check and compensated prefix sums

  workflow-change-approved
  #167·gitRasheed opened Nov 2, 2026Collaborator·

  ·