Issues

• method='nf'/'rm' silently ignore Model.set_concentration: the run starts from the XML's seed amounts

  lanl/bngsim#1373 · Filter by author Montana opened on Oct 29, 2026

• nightly: re-baseline differential klu_vs_dense after #1295 (four BioModels now pass)

  gatedtesting
  lanl/bngsim#1371 · Filter by author wshlavacek opened on Oct 28, 2026

• add_stop_condition is silently ignored by run_batch() and run_replicates() (ODE and SSA)

  bugcorrectness concern
  lanl/bngsim#1369 · Filter by author wshlavacek opened on Oct 28, 2026

• SSA reports the wrong concentration after an event resizes a compartment: Result.species divides by the load-time size (exactly 2x after c: 2 -> 4)

  bugcorrectness concernfamily: variable volumes
  lanl/bngsim#1368 · Filter by author wshlavacek opened on Oct 28, 2026

• Evaluate from_bngl dropping post-generate_network actions: 14 of 103 benchmark BNGL models are benchmarked from seed state, not their authors' protocol

  questiontesting
  lanl/bngsim#1366 · Filter by author wshlavacek opened on Oct 28, 2026

• run_replicates / run_batch: ensembles from consecutive base seeds share almost every trajectory

  correctness concernenhancement
  lanl/bngsim#1365 · Filter by author wshlavacek opened on Oct 28, 2026

• parse_bngl_protocol silently misreads continue, n_output_steps, sample_times, default n_steps and par_scan_vals

  bugcorrectness concern
  lanl/bngsim#1364 · Filter by author wshlavacek opened on Oct 28, 2026

• Result.sse_gradient() documents observable-shaped data but silently scores it against species

  bugcorrectness concern
  lanl/bngsim#1363 · Filter by author wshlavacek opened on Oct 28, 2026

• run_replicates and run_batch silently discard a pending Model.set_concentration and start from the seed ICs

  bugcorrectness concern
  lanl/bngsim#1362 · Filter by author wshlavacek opened on Oct 28, 2026

• ODE default atol=1e-8 silently returns wrong (even negative) trajectories for models in molar units

  bugcorrectness concern
  lanl/bngsim#1361 · Filter by author wshlavacek opened on Oct 28, 2026

• Simulator(method="nf"/"rm", xml_path=...) silently ignores Model.set_param: NFsim and RuleMonkey run the XML's original parameters

  bugcorrectness concernhigh priority
  lanl/bngsim#1360 · Filter by author wshlavacek opened on Oct 28, 2026

• get_param on an assignment-rule parameter after a compiled run returns its load-time value, not the rule's

  lanl/bngsim#1358 · Filter by author Montana opened on Oct 28, 2026

• steady_state(method='integration') on the compiled RHS and Jacobian refuses a state at residual 2e-39 and does not retry on difference quotients

  lanl/bngsim#1357 · Filter by author Montana opened on Oct 28, 2026

• Nightly parity: regressions detected

  lanl/bngsim#1355 · Filter by author github-actions opened on Oct 28, 2026
  1

• Generated models (#1306): a floor/round/trunc law's answer depends on the output grid at rtol 1e-8, by up to 1e3x the tolerance; one fails at rtol 1e-12

  lanl/bngsim#1342 · Filter by author Montana opened on Oct 27, 2026

• sbml: reversing listOfSpecies makes MODEL1812040005 fail (P below 0, a*P^z NaN) where the order written runs; RoadRunner runs both

  lanl/bngsim#1340 · Filter by author Montana opened on Oct 27, 2026

• A step across a pole of the right-hand side returns a trajectory where no solution exists

  correctness concern
  lanl/bngsim#1339 · Filter by author wshlavacek opened on Oct 27, 2026

• Tests: KINSOL's and the steady-state march's difference quotients on a jump, and SUNDIALS other than 7.2, are not covered

  family: steady statetesting
  lanl/bngsim#1335 · Filter by author wshlavacek opened on Oct 27, 2026

• Performance: the step of the solvers' own difference quotients, and #1302's +4–36% wall on the fd path (Step 4 of #1002)

  family: steady stateperformance
  lanl/bngsim#1333 · Filter by author wshlavacek opened on Oct 27, 2026

• Forward sensitivity at a parameter that sits on a kink of a rate law returns one side's slope or neither (dX*/dq = 0 for ±47.6), with no error

  bugcorrectness concernfamily: sensitivity terms
  lanl/bngsim#1332 · Filter by author wshlavacek opened on Oct 27, 2026

• A run that slides along a moving threshold is wrong with no error on a differenced Jacobian (X(40) = 19.7 for 25), and times out on the closed form

  bugcorrectness concern
  lanl/bngsim#1331 · Filter by author wshlavacek opened on Oct 27, 2026

• A differenced Jacobian can be wrong at every step a check takes: take exact derivatives where there is no closed form (Step 3 of #1002)

  bugcorrectness concernfamily: steady state
  lanl/bngsim#1330 · Filter by author wshlavacek opened on Oct 27, 2026
  1

• A failed NFsim load leaves the host process in the XML's directory: create_system() changes the cwd and an exception from initializeFromXML skips the restore. The stream swap and global RNG reseed also need to be explicit and contained

  bug
  lanl/bngsim#1314 · Filter by author wshlavacek opened on Oct 27, 2026

• Typed exceptions for "no sparse Jacobian for KLU" and "codegen declined this model"

  enhancement
  lanl/bngsim#1313 · Filter by author wshlavacek opened on Oct 27, 2026

• differential: NFsim (nf_reject) against RuleMonkey (nf_exact) in distribution

  enhancementgatedtesting
  lanl/bngsim#1312 · Filter by author wshlavacek opened on Oct 27, 2026