Repository navigation

feat(lpse2d): plan 2 of LPSE parity -- three-component fields, light options, quasilinear and FD-IAW modules (stacked on #372) - #378

Open

physicistphil wants to merge 182 commits into

ergodicio:mainfrom

physicistphil:lpse2d/lpse-parity-plan2

Open

feat(lpse2d): plan 2 of LPSE parity -- three-component fields, light options, quasilinear and FD-IAW modules (stacked on #372)#378
physicistphil wants to merge 182 commits into
ergodicio:mainfrom
physicistphil:lpse2d/lpse-parity-plan2

Commits on Sep 14, 2026
- lpse2d: LPSE term parity -- TPD coefficient, thermal noise, Landau forms, spectral IAW
  
  physicistphil
  and
  claude
  committed
  58b5038
- lpse2d: per-operation EPW energy ledger (terms.epw.energy_ledger)
  
  physicistphil
  and
  claude
  committed
  87cd442
- lpse2d: spectral light propagator and collisional light absorption
  
  physicistphil
  and
  claude
  committed
  32f8c44
- lpse2d: combined TPD+SRS solver (terms.epw.solver: combined)
  
  physicistphil
  and
  claude
  committed
  77a545b
Commits on Sep 15, 2026
Commits on Sep 16, 2026
Commits on Sep 17, 2026
Commits on Sep 18, 2026
- feat(lpse2d): quasilinear evolution of the electron distribution -- terms.qle (plan 2 K.1)
  
  physicistphil
  and
  claude
  committed
  078ec22
- style(lpse2d): wrap a qle docstring line
  
  physicistphil
  and
  claude
  committed
  658b30d
- feat(lpse2d): finite-difference IAW solver with plasma-flow profiles (plan 2 I.1, I.2)
  
  physicistphil
  and
  claude
  committed
  43ee4bb
- fix(lpse2d): the tpd-srs-iaw example runs LPSE's combined formulation (plan 2 N.4)
  
  physicistphil
  and
  claude
  committed
  a7b34a0
Commits on Sep 21, 2026
- feat(lpse2d): terms.light.one_way -- directional mask on the evolved pump
  
  physicistphil
  and
  claude
  committed
  bcee98d
- fix(lpse2d): translator reads laser.pulseShape.* and bracketed vectors; reference batch
  
  physicistphil
  and
  claude
  committed
  bea5d0b
- feat(lpse2d): implicit Fokker-Planck step for the QLE; light-only decks translate (plan 2 K.1, M11)
  
  physicistphil
  and
  claude
  committed
  0a9a3d0
- fix(lpse2d): 1-D decks keep a periodic y boundary; compare() tolerates light-only references
  
  physicistphil
  and
  claude
  committed
  3cdb700
- fix(lpse2d): QLE gradient falls back to the plain difference at a plateau's clipped edge
  
  physicistphil
  and
  claude
  committed
  621b27b
- fix(lpse2d): refuse y-face beams in the translator and grazing beams in the FD injector
  
  physicistphil
  and
  claude
  committed
  aee5f9b
- style(lpse2d): wrap a translator message
  
  physicistphil
  and
  claude
  committed
  dcc187e
Commits on Sep 22, 2026
Commits on Sep 23, 2026
Commits on Sep 24, 2026
- fix(lpse2d): absorbing layers per field and injector planes as LPSE; IAW coupling switches
  
  physicistphil
  and
  claude
  committed
  6eda074
- fix(lpse2d): translator keys and defaults as the C++ reads them; LPSE's KAP process; HPE tail and limits
  
  physicistphil
  and
  claude
  committed
  46833f9
- fix(lpse2d): translated grids keep LPSE's node counts; fd-combined decks are reported
  
  physicistphil
  and
  claude
  committed
  8aa0fdf
- feat(lpse2d): translate LPSE's own solver time steps; exact fd-light dt bound
  
  physicistphil
  and
  claude
  committed
  6a73f94
- feat(lpse2d): LPSE step order with time-interpolated sources
  
  physicistphil
  and
  claude
  committed
  914a0d7
- fix(lpse2d): combined solver drives the IAW with the whole field at wp0
  
  physicistphil
  and
  claude
  committed
  8a95853
- fix(lpse2d): IAW absorber acts over the IAW step
  
  physicistphil
  and
  claude
  committed
  c8c8628
- fix(lpse2d): IAW band has its own wavenumber cap
  
  physicistphil
  and
  claude
  committed
  d130908
- fix(lpse2d): LPSE IAW solvers window the drive after the Laplacian
  
  physicistphil
  and
  claude
  committed
  36582f0
- fix(lpse2d): translated density on LPSE's nodes; LPSE's cell size
  
  physicistphil
  and
  claude
  committed
  e8bfef4
- fix(lpse2d): rectangular dealias uses LPSE's anti-aliasing indices
  
  physicistphil
  and
  claude
  committed
  421e0ac
- fix(lpse2d): TPD source windowed after its k-space assembly
  
  physicistphil
  and
  claude
  committed
  c1f7c77
- test(lpse2d): neutralise the pump's own absorber in the FD projection test
  
  physicistphil
  and
  claude
  committed
  8ec5102
- fix(lpse2d): light coupling sources suppressed as LPSE's calculateSources
  
  physicistphil
  and
  claude
  committed
  49a91e6
- feat(lpse2d): HPE resonance at LPSE's expanded Bohm-Gross frequency
  
  physicistphil
  and
  claude
  committed
  b71fc8e
- feat(lpse2d): static pump swells by LPSE's constant factor by default
  
  physicistphil
  and
  claude
  committed
  f874833
- fix(lpse2d): absolute-threshold criterion as LPSE's checkIsAboveThreshold
  
  physicistphil
  and
  claude
  committed
  e3d6b85
- fix(lpse2d): thermal-filamentation source as LPSE's
  
  physicistphil
  and
  claude
  committed
  b4828ad
- feat(lpse2d)!: native defaults follow LPSE (C++ conventions)
  
  physicistphil
  and
  claude
  committed
  1a8194f
- feat(lpse2d): LPSE's combined solver on the FD light propagator
  
  physicistphil
  and
  claude
  committed
  df54490
- test(lpse2d): tests follow the C++ defaults; exchange mask in the rotation test
  
  physicistphil
  and
  claude
  committed
  0001142
- fix(lpse2d): FD light absorbs each field at its own rate
  
  physicistphil
  and
  claude
  committed
  ffb03ff
- test(lpse2d): restart test keeps the flat noise it checks
  
  physicistphil
  and
  claude
  committed
  b866c23
- test(lpse2d): s / p launch comparison at the same light step
  
  physicistphil
  and
  claude
  committed
  306fc34
- test(lpse2d): combined + FD light without an evolved pump is refused (A15)
  
  physicistphil
  and
  claude
  committed
  350c32e
Commits on Sep 30, 2026
- fix(lpse2d): combined solver keeps the k = 0 mode of E1, as LPSE
  
  physicistphil
  and
  claude
  committed
  044784f
Commits on Oct 8, 2026
- fix(lpse2d): combined solver builds its noise kick for the light sub-step
  
  physicistphil
  and
  claude
  committed
  59e01d0
- fix(lpse2d): combined solver injects the pump with the spectral injector
  
  physicistphil
  and
  claude
  committed
  1e3cf4a
- fix(lpse2d): combined-mode field saves and Raman flux maps exclude the EPW
  
  physicistphil
  and
  claude
  committed
  f6b1e1a
- fix(lpse2d): LPSE binary density files are not transposed twice
  
  physicistphil
  and
  claude
  committed
  27418ef
- fix(lpse2d): quasi-1D HPE push keeps periodic crossings for the wall accounting
  
  physicistphil
  and
  claude
  committed
  f6c7429
- test(lpse2d): the test_010 beam-direction translation runs from a checked-in deck
  
  physicistphil
  and
  claude
  committed
  278fda7
- Merge lpse2d/lpse-parity (PR #372 review fixes) into lpse2d/lpse-parity-plan2
  
  physicistphil
  and
  claude
  committed
  4a3e3b1
- fix(lpse2d): FD-IAW resampling aligns the zero frequency on odd grids
  
  physicistphil
  and
  claude
  committed
  0c8008c
- fix(lpse2d): parity CLI reads lpse.metrics in the flattened reference layout
  
  physicistphil
  and
  claude
  committed
  c715248
- test(lpse2d): verify baseline test skips when its run is not on the tracking server
  
  physicistphil
  and
  claude
  committed
  b00d903
- test(lpse2d): pin the noise seed of the in-run QLE flattening test
  
  physicistphil
  and
  claude
  committed
  f5dc1a4
Commits on Oct 9, 2026
- test(lpse2d): QLE flattening is checked over the saved times, criterion fixed in advance
  
  physicistphil
  and
  claude
  committed
  3b582f6
- fix(lpse2d): combined solver injects the Raman seed (drivers.E1)
  
  physicistphil
  and
  claude
  committed
  72b6df4
- fix(lpse2d): HPE wall-energy bins count only energies inside their edges
  
  physicistphil
  and
  claude
  committed
  c387f68
- fix(lpse2d): one normal-incidence beam keeps its frequency shift and phase
  
  physicistphil
  and
  claude
  committed
  e5c7327
Commits on Oct 10, 2026
- Merge lpse2d/lpse-parity (PR #372 review round 2) into lpse2d/lpse-parity-plan2
  
  physicistphil
  and
  claude
  committed
  8588bf1
- docs(lpse2d): the combined-solver seed follows drivers.E1.polarization
  
  physicistphil
  and
  claude
  committed
  033d613
- fix(lpse2d): QLE explicit update is reverse-mode differentiable
  
  physicistphil
  and
  claude
  committed
  c214230
- fix(lpse2d): laser-budget metrics follow the pump's entrance face
  
  physicistphil
  and
  claude
  committed
  aa48996
Commits on Oct 16, 2026
- fix(lpse2d): a disabled LPSE laser translates to no pump
  
  physicistphil
  and
  claude
  committed
  c5fee59
- fix(lpse2d): Raman flux diagnostics use the combined field's carrier
  
  physicistphil
  and
  claude
  committed
  bc78533
- Merge lpse2d/lpse-parity (PR #372 review round 3) into lpse2d/lpse-parity-plan2
  
  physicistphil
  and
  claude
  committed
  de42398
- fix(lpse2d): the E1 light-spectrum probe uses the combined carrier
  
  physicistphil
  and
  claude
  committed
  5eae480
- fix(lpse2d): the FD combined solver injects the Raman seed (drivers.E1)
  
  physicistphil
  and
  claude
  committed
  3b810d8
- fix(lpse2d): FD-IAW PPM fluxes are upwinded per shared face
  
  physicistphil
  and
  claude
  committed
  962e2a1
- fix(lpse2d): the combined E1 light-spectrum probe records the Raman light only
  
  physicistphil
  and
  claude
  committed
  f687c21
- fix(lpse2d): the IAW absorber damps over the whole strided IAW step
  
  physicistphil
  and
  claude
  committed
  e71050f
- fix(lpse2d): the IAW and the Raman absorption see the combined light at wp0
  
  physicistphil
  and
  claude
  committed
  88a65bd
- fix(lpse2d): combined runs skip the separate light solvers and their w1 cutoffs
  
  physicistphil
  and
  claude
  committed
  bac837e
Commits on Oct 17, 2026
- Merge lpse2d/lpse-parity (PR #372 review round 4) into lpse2d/lpse-parity-plan2
  
  physicistphil
  and
  claude
  committed
  fe8c40c
- fix(lpse2d): combined-mode absorption follows the evolved ion density
  
  physicistphil
  and
  claude
  committed
  5a56eb0
- fix(lpse2d): parity batch dry runs record nothing
  
  physicistphil
  and
  claude
  committed
  cc48019
Commits on Oct 19, 2026
- fix(lpse2d): the deck translator runs LPSE's square cells
  
  physicistphil
  and
  claude
  committed
  f123c5e
- fix(lpse2d): the FD pump injector refuses the time controls it does not apply
  
  physicistphil
  and
  claude
  committed
  3fccddb
- Merge lpse2d/lpse-parity (PR #372 review round 5) into lpse2d/lpse-parity-plan2
  
  physicistphil
  and
  claude
  committed
  a415138
- fix(lpse2d): resonance absorption -- one enable predicate, absolute cadence, bounded damping
  
  physicistphil
  and
  claude
  committed
  d4fab87
- fix(lpse2d): the combined field is damped at the Raman absorption
  
  physicistphil
  and
  claude
  committed
  c440b7e
- fix(lpse2d): threshold bisection refuses a zero lower bound
  
  physicistphil
  and
  claude
  committed
  6724180
- Merge lpse2d/lpse-parity (PR #372 review round 6) into lpse2d/lpse-parity-plan2
  
  physicistphil
  and
  claude
  committed
  ae6a6da
- fix(lpse2d): the quasi-1D resonance-absorption filter keeps its passband
  
  physicistphil
  and
  claude
  committed
  8b1430b
- fix(lpse2d): KAP phase histories are prefix-stable as the run grows
  
  physicistphil
  and
  claude
  committed
  38d5783
- fix(lpse2d): restart rebases the light-spectrum probes onto the resumed interval
  
  physicistphil
  and
  claude
  committed
  73c8b7d
- fix(lpse2d): parity batch files keep every window; unknown job keys are refused
  
  physicistphil
  and
  claude
  committed
  df8a935
- fix(lpse2d): the LPSE threshold search refuses a zero starting intensity
  
  physicistphil
  and
  claude
  committed
  7821ca1
Commits on Oct 20, 2026
- fix(lpse2d): a light-only PML run uses the pump's layer
  
  physicistphil
  and
  claude
  committed
  8f5668d
- fix(lpse2d): build the separate EPW solver only on the separate path
  
  physicistphil
  and
  claude
  committed
  2577b05
- fix(lpse2d): refuse fd rotation coupling with the transverse projection
  
  physicistphil
  and
  claude
  committed
  55fd66f
- Merge branch lpse2d/lpse-parity (round-7 fixes) into lpse2d/lpse-parity-plan2
  
  physicistphil
  and
  claude
  committed
  3a096b8
- fix(lpse2d): fd rotation refusal follows plan2's transverse_source default
  
  physicistphil
  and
  claude
  committed
  4223609
- fix(lpse2d): FD combined source takes the EPW window and the pump k-filter
  
  physicistphil
  and
  claude
  committed
  adec784
- docs(lpse2d): the combined depletion is unprojected
  
  physicistphil
  and
  claude
  committed
  d24121d
- fix(lpse2d): put the interpolated field save on a coarser grid, in x-space
  
  physicistphil
  and
  claude
  committed
  19cbe2e
- test(lpse2d): gradient objective takes the EPW energy from the grid
  
  physicistphil
  and
  claude
  committed
  7081182
- fix(lpse2d): the spectral light band ignores the EPW wavenumber cap
  
  physicistphil
  and
  claude
  committed
  952252c
- fix(lpse2d): signed EPW dissipation with HPE allow_growth
  
  physicistphil
  and
  claude
  committed
  ca6ced8
- Merge branch lpse2d/lpse-parity (light band, signed dissipation) into lpse2d/lpse-parity-plan2
  
  physicistphil
  and
  claude
  committed
  7a22883
- fix(lpse2d): combined-mode EPW dissipation and lw heating use the combined damping
  
  physicistphil
  and
  claude
  committed
  9119fd2
- Merge branch lpse2d/lpse-parity (combined-mode damping in diagnostics) into lpse2d/lpse-parity-plan2
  
  physicistphil
  and
  claude
  committed
  d9f0d46
- test(lpse2d): combined boundary-loss diagnostic uses the combined layer
  
  physicistphil
  and
  claude
  committed
  8bd768f
- fix(lpse2d): the FD combined light step obeys the combined bound with the sources off
  
  physicistphil
  and
  claude
  committed
  1f35493
- fix(lpse2d): the FD IAW fine-grid band is the coarse mask mode by mode
  
  physicistphil
  and
  claude
  committed
  72f578b
- docs(lpse2d): HPE gamma limits bound the rate before the energy-conservation multiplier
  
  physicistphil
  and
  claude
  committed
  3e8f4f9
- Merge branch lpse2d/lpse-parity (HPE limit docs) into lpse2d/lpse-parity-plan2
  
  physicistphil
  and
  claude
  committed
  957a791
- fix(lpse2d): the dissipation diagnostic follows the evolved QLE Landau rate
  
  physicistphil
  and
  claude
  committed
  0d50e67
- fix(lpse2d): HPE gamma limits bound the applied rate after energy conservation
  
  physicistphil
  and
  claude
  committed
  971f999
- Merge branch lpse2d/lpse-parity (HPE limits after energy conservation) into lpse2d/lpse-parity-plan2
  
  physicistphil
  and
  claude
  committed
  a694ba0
- fix(lpse2d): the FD IAW restriction to the EPW grid keeps only the IAW band
  
  physicistphil
  and
  claude
  committed
  1d2bac5
- fix(lpse2d): saved EPW ex/ey without the stray 4/(nx ny) factor
  
  physicistphil
  and
  claude
  committed
  9eb7ff5
- docs(lpse2d): scope the saved-field Fourier identities to full-grid saves
  
  physicistphil
  and
  claude
  committed
  984f0f6
- test(lpse2d): run the slow HPE trapping and smoke regressions again
  
  physicistphil
  and
  claude
  committed
  86606c8
- Merge branch lpse2d/lpse-parity (slow HPE tests) into lpse2d/lpse-parity-plan2
  
  physicistphil
  and
  claude
  committed
  c25d7eb
- test(lpse2d): HPE linear closure is measured against the configured Landau rate
  
  physicistphil
  and
  claude
  committed
  4d193ad
- fix(lpse2d): FD-combined inverse-Landau growth is exponential; FD IAW noise obeys the IAW band
  
  physicistphil
  and
  claude
  committed
  85f29cc
Commits on Oct 21, 2026
- test(lpse2d): HPE smoke test with enough particles for its kpeak check
  
  physicistphil
  and
  claude
  committed
  df43c9b
- Merge branch lpse2d/lpse-parity (HPE smoke test particles) into lpse2d/lpse-parity-plan2
  
  physicistphil
  and
  claude
  committed
  8dee419
- docs(lpse2d): list the intentional departures from LPSE
  
  physicistphil
  and
  claude
  committed
  5686b91
- docs(lpse2d): single trailing newline in the overview (end-of-file-fixer)
  
  physicistphil
  and
  claude
  committed
  360d65f
- docs(lpse2d): complete the LPSE departures list from an audit; fix wrong statements
  
  physicistphil
  and
  claude
  committed
  bca2538
Commits on Oct 22, 2026
Commits on Oct 23, 2026
- fix(lpse2d): refresh the prescribed pump before the IAW and EPW read it
  
  physicistphil
  and
  claude
  committed
  5119513
- docs(lpse2d): the prescribed-pump timing in Departures from LPSE
  
  physicistphil
  and
  claude
  committed
  30493b4
- fix(lpse2d): HPE flux parity compares the same time window on both sides
  
  physicistphil
  and
  claude
  committed
  75d3399
- fix(lpse2d): parity verify fails on a non-finite stored or recomputed metric
  
  physicistphil
  and
  claude
  committed
  7099c40
- fix(lpse2d): refuse Poynting maps on a single-point x save grid at setup
  
  physicistphil
  and
  claude
  committed
  a62b52c
- Merge branch lpse2d/lpse-parity (Poynting save-grid check) into lpse2d/lpse-parity-plan2
  
  physicistphil
  and
  claude
  committed
  04b432a