WEEKLY REPORT · ISO 2026-W44

PHARE, 26 Oct – 1 Nov 2026

Commits on master
28 ▲ +6 vs last week
PRs merged
22 ▲ +6 vs last week
Issues opened
8 ▼ -22 vs last week
Issues closed
9 ▲ +2 vs last week
Reviews
61 ▼ -40 vs last week
CI builds
1,951 ▼ -92 vs last week

The week at a glance

each dot is one event · hover for details, click to open
Mon 26MTue 27TWed 28WThu 29TFri 30FSat 31SSun 1S
Commits on master
28 PRs merged
22 PRs opened
31 Issues opened
8 Issues closed
9 Reviews
61 CI failures
282
Daily numbers as a table
DayCommitsIssues openedIssues closedPRs openedPRs mergedPRs closedReviewsCommentsCI buildsCI failed
Mon 26 Oct702551171216553
Tue 27 Oct2209229413418
Wed 28 Oct93438016419726
Thu 29 Oct4323208810124
Fri 30 Oct100110066332
Sat 31 Oct1006126349176
Sun 1 Nov4014305180053

Issues

8 opened · 9 closed

Open issues

The axis starts at 130 so the weekly changes stay visible.

Labels on new issues

bug 🔥
3triage
2MHD
2feature
1diagnostics outputs
1pharesee
1I/O
1test
1no label
3

How closed issues were closed

fixed by a PR 9
  • #1500 Quantity centering lookups: per-dimension switches are redundant and MHD table order mismatches the enum completed fixed by PR #1495
  • #1498 Refactor for_N stuff for C++20 completed fixed by PR #1503
  • #1465 Simulator: make the constructor noexcept, move the throwing setup into Simulator::setup() completed fixed by PR #1467
  • #1397 Hybrid: reserve particle array capacity when patches are filled, to cut memory footprint and load imbalance completed fixed by PR #1398
  • #1385 Functional tests: CMake registration and stale-comment inconsistencies completed fixed by PR #1392
  • #1303 density profile not checked for periodicity in MaxwellianFluidModel completed fixed by PR #1413 bug 🔥pharein
  • #1290 auto_dump=false disables more than it should completed fixed by PR #1307 refactoring
  • #1286 Clarify LoadBalancer block pharein completed fixed by PR #1308 refactoringpharein
  • #1198 Inconsistent usage of global refinement_ratio constant completed fixed by PR #1525 refactoringpharesee
All 8 new issues
  • #1529 Magnetic-basis particle initializer reads magnetic_x for all three B components @nicolasaunai
  • #1527 Isothermal electron pressure term generates Bz from B=0 in 2D @acrd bug 🔥triage
  • #1526 Same input, 3+ MPI ranks: two runs give different particles (last bit) @nicolasaunai
  • #1523 Hall MHD: energy leaks through reflective walls (Hall Poynting flux in F_Etot vs zero tangential E in CT), wall pressure drops to zero, NaN @igirault bug 🔥triageMHD
  • #1522 Electromag diagnostics: write B, B0 and B1 in the same phareh5 / vtkhdf file @nicolasaunai featurediagnostics outputsphareseeI/O
  • #1521 MHD shock tests listed under Harris names by phlop (shared file basenames) @nicolasaunai bug 🔥testCIMHD
  • #1518 Stored field layouts have dl=1 / origin=0 after restart @PhilipDeegan
  • #1515 One process-wide scratch buffer for field data instead of hidden static vectors @PhilipDeegan corerefactoringperformance

Pull requests

22 merged · 31 opened · 5 closed unmerged

Open PRs

The axis starts at 20 so the weekly changes stay visible.

Every PR that moved this week

mergedopen draft or closed first human review
Time open (log scale)
1 h1 d1 wk1 mo1 yr
Lines changed, by area (log scale)
101001k10k
1539 build(deps): bump the actions group with 4 updates@dependabot
1538 small suggestions for the new box iterations@igirault
1531 1D tagging: refined patches no longer stick to level-0 patch boundaries@nicolasaunai
1525 python: only use global refinement ratio@PhilipDeegan
1524 make level 0 dl accessible without a time@PhilipDeegan
1509 ci: draft PRs run only the gate@nicolasaunai
1508 for_N_fold in NdArrayViewer::idx and detail::shifted, plus ndarray access bench@PhilipDeegan
1507 OpenCode image for the nightly's AI steps@nicolasaunai
1506 Explain red TeamCity builds in a PR comment (advisory, opt-in by label)@nicolasaunai
1504 Faster particle/mesh interpolation: stride field access and a fused __restrict deposit@PhilipDeegan
1503 Shrink for_N with C++20 templated lambdas; add for_N_fold so client code needs no index_sequence@PhilipDeegan
1502 chore: ruff errors@PhilipDeegan
1495 centerings lookup table instead of switch@PhilipDeegan
1486 Read a field at another centring without copying it (RecenteredField)@igirault
1468 Shared staggering-aware curl/grad/div/laplacian: one curl for Ampere, Faraday, #1327 and #1339, 3-4x faster@igirault
1467 Simulator: make the constructor noexcept, move the throwing setup into Simulator::setup()@PhilipDeegan
1413 Check density periodicity and fix the 1D B periodicity check@PhilipDeegan
1398 optimize ParticleArray allocations with reserving for expected sizes@PhilipDeegan
1392 functional tests cmake/consistency@PhilipDeegan
1308 load balancing cleanup@PhilipDeegan
1307 drop auto_dump from python Simulator@PhilipDeegan
1153 Faster box loops: one pointer per span for copies, vectorized evalOnBox@PhilipDeegan
1542 Carry MHD magnetic energy by the CT Poynting flux to stop the Hall energy leak at reflective walls@igirault
1541 Hybrid dispersion test now checks ω(k) of both branches in 80 s, instead of checking nothing for 2+ hours@nicolasaunai
1540 Lighter 1D functional tests: drop td1d and translat1d, td1dtagged checks and plots more in less time@nicolasaunai
1537 pharesee: read vtkhdf diagnostics into hierarchies@PhilipDeegan
1536 tests: test_run parameterised by diag format, plots into per run dirs@PhilipDeegan
1535 pharesee: compute callbacks take a patch and return FieldData@PhilipDeegan
1534 pharesee: Run without h5py, times keyed by file stem, comparison reports@PhilipDeegan
1533 pharesee: FieldData.field_name -> name@PhilipDeegan
1532 Give MHD test scripts unique names so the nightly runs and logs them correctly@nicolasaunai
1530 Seed particles per cell instead of per patch@nicolasaunai
1528 Isothermal electron pressure no longer creates magnetic field from density noise@nicolasaunai
1520 Performance monitoring introspection@PhilipDeegan
1519 Nightly physics report: a free AI model reads the functional tests' results (scripts, advisory)@nicolasaunai
1517 pharesee: drop merged=True from Run getters in favour of hier.finest()@PhilipDeegan
1516 single source for static buffers @PhilipDeegan
1514 res/eg: 1d jupyter notebooks, executed on CI@PhilipDeegan
1513 pharesee: gaussian filter on finest UniformGrids@PhilipDeegan
1512 pharesee: UniformGrid, finest data of field hierarchies@PhilipDeegan
1511 pharesee: drop ghosts from moments, combine fields with differing ghosts@PhilipDeegan
1510 pharesee: move flat_finest_field to run/utils, split make_interpolator@PhilipDeegan
1505 vtkhdf: stop writing each dump twice, one write per rank (17-31x faster dumps at 1500-3000 ranks)@igirault
1338 remove merged parameter from run.py in favor of hier.finest()@PhilipDeegan
1276 fix: emergency dump one full-hierarchy step, not per-level@igirault
1253 `compute_hier_from`: keep patch data types, drop renaming for binary operations@PhilipDeegan
1195 interpolation/gaussian/notebooks@PhilipDeegan
1172 python3 Vtk hierarchies@PhilipDeegan
coreamrpythontestssimulator/iobuild/cidoc

People

human activity inside the week · darker = more

Who did what

CommitsPRs openedPRs mergedReviews givenIssues openedIssues closedCommentsLines on master
@PhilipDeegan 131714292119 +2,611 −1,273
@nicolasaunai 1110428489 +2,167 −98
@igirault 33331·10 +1,618 −662
@UCaromel ···1··· ·
@acrd ····1·· ·

Bots: coderabbitai (41 reviews, 31 comments), github-advanced-security (16 reviews, 0 comments), github-actions (0 reviews, 9 comments), dependabot (0 reviews, 0 comments), copilot-pull-request-reviewer (1 reviews, 0 comments), dependabot[bot] (0 reviews, 0 comments).

Who reviewed whom

reviewer ↓ · author →igiraultPhilipDeegannicolasaunaiUCaromeldependabot
PhilipDeegan22·43·
nicolasaunai421·21
igirault··3··
UCaromel1····
Median first review27 h
Median time to merge41 h
17 open PRs waiting for a first human review · 1 with no activity for 30 days (as of 2 Nov)

Waiting for review

  • #1171Live, zero-copy access to simulation data from Python@PhilipDeegan open 238.1 d
  • #1267validate restarts against hash produced from initial patch datas@PhilipDeegan open 103.1 d
  • #1396do no duplicate levelghosts across overlapping patch ghost boxes@PhilipDeegan open 22.2 d
  • #1406model temporaries as needed via model::requestTemporaries@PhilipDeegan open 21.1 d
  • #1431amr: level accessors, hide per patch loops behind onLevel@PhilipDeegan open 17.1 d
  • #1432Install PHARE with pip install . (scikit-build-core build backend)@igirault open 16.4 d
  • #1445Validate collisions on refined levels and document them@nicolasaunai open 14.4 d
  • #1444Ion-ion collisions in hybrid runs: Simulation(collisions=...)@nicolasaunai open 14.4 d
  • #1443Exact momentum and energy conservation for collisions with unequal weights@nicolasaunai open 14.4 d
  • #1442Binary Monte-Carlo ion-ion collision operator, in core and tested against NRL rates@nicolasaunai open 14.4 d
  • #1441Deposit ion moments after the particle exchange, so each cell's particles are on one patch first@nicolasaunai open 14.4 d
  • #1501Drop box iterator@PhilipDeegan open 8.8 d
  • #1510pharesee: move flat_finest_field to run/utils, split make_interpolator@PhilipDeegan open 6.2 d
  • #1516single source for static buffers @PhilipDeegan open 6.1 d
  • #1520Performance monitoring introspection@PhilipDeegan open 5.1 d
  • #1537pharesee: read vtkhdf diagnostics into hierarchies@PhilipDeegan open 2.2 d
  • #1545Stop py3_init-densities failing by chance: thresholds measured over many runs@nicolasaunai open 2 h

Stale

  • #1300release: bump SAMRAI and HighFive pins for the upcoming 0.1.0 release@nicolasaunai open 94.9 d

CI and tests

411,702 test executions on TeamCity · 386 failed · 7 ignored
passedfailedcanceledone tick per build, placed at its finish time
Mon 26MTue 27TWed 28WThu 29TFri 30FSat 31SSun 1S
GitHub Actions ci-explainer
141 ✓ · 0 ✗ · 431 canceled review-threads
59 ✓ · 18 ✗ · 62 canceled gate
34 ✓ · 20 ✗ · 31 canceled Coverage
25 ✓ · 6 ✗ · 23 canceled labels
50 ✓ · 0 ✗ · 0 canceled CMake Ubuntu
24 ✓ · 3 ✗ · 23 canceled CodeQL
27 ✓ · 1 ✗ · 20 canceled github_actions in /. - Update #1669682617
1 ✓ · 0 ✗ · 0 canceled pip install
1 ✓ · 0 ✗ · 0 canceled TeamCity gh_pr_clang_samraisub
97 ✓ · 28 ✗ · 14 canceled gh_pr_gcc_samraisub
93 ✓ · 33 ✗ · 13 canceled gh_pr_gcc
84 ✓ · 27 ✗ · 16 canceled · 190,205 tests run gh_pr_clang
89 ✓ · 23 ✗ · 15 canceled · 189,839 tests run gh_master_clang_nightly_samraisub_phlop_asan
5 ✓ · 1 ✗ · 1 canceled gh_master_gcc_nightly_samraisub_phlop
5 ✓ · 1 ✗ · 0 canceled gh_master_gcc_samraisub_weekly_gcov
0 ✓ · 0 ✗ · 4 canceled Build Docker-dep images
0 ✓ · 1 ✗ · 0 canceled build_samrai_upstream_dep
1 ✓ · 0 ✗ · 0 canceled gh_master_gcc_weekly_samraisub
1 ✓ · 0 ✗ · 0 canceled gh_master_clang_weekly_asan
1 ✓ · 0 ✗ · 0 canceled gh_master_gcc_samraisub_weekly_caliper
1 ✓ · 0 ✗ · 0 canceled TeamCity, next CI (shadow) Pipeline Head
82 ✓ · 45 ✗ · 1 canceled · 11,078 tests run clang ASan (build + fast tests)
65 ✓ · 32 ✗ · 1 canceled · 10,218 tests run gcc (build + fast tests)
64 ✓ · 28 ✗ · 4 canceled · 10,362 tests run Nightly page
8 ✓ · 0 ✗ · 0 canceled MHD functional
2 ✓ · 5 ✗ · 0 canceled hybrid harris 2D, MPI
6 ✓ · 1 ✗ · 0 canceled dispersion (dispersion.py), MPI
0 ✓ · 7 ✗ · 0 canceled build (gcc Release)
6 ✓ · 0 ✗ · 0 canceled fast tests (all of level 10 and below but 3D init), serial + MPI
6 ✓ · 0 ✗ · 0 canceled 3D particle init, serial
6 ✓ · 0 ✗ · 0 canceled 3D particle init, MPI
6 ✓ · 0 ✗ · 0 canceled hybrid functional, 1D decks
5 ✓ · 1 ✗ · 0 canceled MHD harris 2D + shock (periodic + with boundaries), MPI
6 ✓ · 0 ✗ · 0 canceled conservation (conserv.py), MPI
5 ✓ · 1 ✗ · 0 canceled OpenCode image
3 ✓ · 0 ✗ · 0 canceled TeamCity, personal builds Ivan · build
1 ✓ · 0 ✗ · 0 canceled

Tests that failed

(empty): test-interpolator
13(empty): test-diagnostics_1d_interp1_refined2
10(empty): test-diagnostics_1d_interp1_refined3
10(empty): test-diagnostics_1d_interp2_refined2
10(empty): test-diagnostics_1d_interp2_refined3
10(empty): test-diagnostics_1d_interp2_refined4
10(empty): test-diagnostics_1d_interp3_refined2
10(empty): test-diagnostics_1d_interp3_refined3
10(empty): test-diagnostics_1d_interp3_refined4
10(empty): test-diagnostics_1d_interp3_refined5
10

Red, then green on the same commit

9 reruns turned a red build green with no failing test: a build or infrastructure problem, not a test.

Code on master

28 commits · lines removed ← → lines added

Where the code changed

build/ci
16
3,092
core
1,503
1,095
tests
214
1,839
amr
69
174
python
135
105
simulator/io
102
97

Commits

28 commits on master (also dots on the timeline)