Issues

A comment after a trailing comma is parsed as a model or data filename, because the file tokens are unanchored regexes that run straight through the '#'

#599

· wshlavacek opened

on Aug 19, 2026

The tutorial README's lesson index stops at 48, so lesson 49 is undiscoverable from the only place lessons are listed

#598

· wshlavacek opened

on Aug 19, 2026

Retire register_bngl() + petab-spec CI machinery once a PyPI petab after 0.8.2 ships native BNGL

gated

#591

· wshlavacek opened

on Aug 19, 2026

ms_segments default of 4 was tuned on the median at radius 0.2; the tail at radius 0.4 favours 8

#585

· wshlavacek opened

on Aug 17, 2026

The shooting suites verify derivatives but not behaviour at pathological parameter points (cf. #578, #581)

#584

· wshlavacek opened

on Aug 17, 2026

Investigate: ADR-0059 rejects simulator-path HMC on an unmeasured cost model ("~1x a solve per leapfrog") — measure it against the bngsim JAX bridge, and settle it with numbers either way

exploratory

#576

· wshlavacek opened

on Aug 14, 2026

Structural identifiability as a pre-fit gate: decide symbolically whether a parameter can EVER be identified from these observables, instead of inferring it numerically from a flat profile

exploratory

#575

· wshlavacek opened

on Aug 14, 2026

Optimal experimental design: turn profile_likelihood's "practically unidentifiable" verdict into the experiment that fixes it, reusing the assemble_fisher_hessian already built for gntr

exploratory

#574

· wshlavacek opened

on Aug 14, 2026

Profile an observable's additive offset — and the coupled (scale, offset) pair — out of the search, completing the ADR-0066/ADR-0108 family (6 of 23 benchmark slugs, up to 42% of the search): method proposal + evaluation requested before we commit

exploratory

#572

· wshlavacek opened

on Aug 13, 2026

BNGSIM_HAS_EVENT_SENS is a version floor at exactly 0.12.2, which the released wheel passes while lacking the commits it gates — the sibling flag's own comment explains why that is unsafe

#558

· wshlavacek opened

on Aug 11, 2026

Replace fixed 10-second waits during distributed-cluster startup and shutdown with real readiness checks

#398

· wshlavacek opened

on May 31, 2026

Navigation Menu

Uh oh!

All issues

A comment after a trailing comma is parsed as a model or data filename, because the file tokens are unanchored regexes that run straight through the '#'

The tutorial README's lesson index stops at 48, so lesson 49 is undiscoverable from the only place lessons are listed

Retire register_bngl() + petab-spec CI machinery once a PyPI petab after 0.8.2 ships native BNGL

ms_segments default of 4 was tuned on the median at radius 0.2; the tail at radius 0.4 favours 8

The shooting suites verify derivatives but not behaviour at pathological parameter points (cf. #578, #581)

Investigate: ADR-0059 rejects simulator-path HMC on an unmeasured cost model ("~1x a solve per leapfrog") — measure it against the bngsim JAX bridge, and settle it with numbers either way

Structural identifiability as a pre-fit gate: decide symbolically whether a parameter can EVER be identified from these observables, instead of inferring it numerically from a flat profile

Optimal experimental design: turn profile_likelihood's "practically unidentifiable" verdict into the experiment that fixes it, reusing the assemble_fisher_hessian already built for gntr

Profile an observable's additive offset — and the coupled (scale, offset) pair — out of the search, completing the ADR-0066/ADR-0108 family (6 of 23 benchmark slugs, up to 42% of the search): method proposal + evaluation requested before we commit

BNGSIM_HAS_EVENT_SENS is a version floor at exactly 0.12.2, which the released wheel passes while lacking the commits it gates — the sibling flag's own comment explains why that is unsafe

Replace fixed 10-second waits during distributed-cluster startup and shutdown with real readiness checks

Uh oh!

Search results