WEEKLY REPORT · ISO 2026-W43

PHARE, 19 Oct – 25 Oct 2026

Commits on master
22 ▼ -5 vs last week
PRs merged
16 ▼ -3 vs last week
Issues opened
30 ▲ +17 vs last week
Issues closed
7 ▼ -8 vs last week
Reviews
101 ▲ +73 vs last week
CI builds
2,043 ▲ +991 vs last week

The week at a glance

each dot is one event · hover for details, click to open
Mon 19MTue 20TWed 21WThu 22TFri 23FSat 24SSun 25S
Commits on master
22 PRs merged
16 PRs opened
34 Issues opened
30 Issues closed
7 Reviews
101 CI failures
366
Daily numbers as a table
DayCommitsIssues openedIssues closedPRs openedPRs mergedPRs closedReviewsCommentsCI buildsCI failed
Mon 19 Oct520113032711728
Tue 20 Oct2416203481727
Wed 21 Oct61355029318828
Thu 22 Oct20023091343383
Fri 23 Oct61135010447170
Sat 24 Oct1413108430063
Sun 25 Oct0002009536287

Issues

30 opened · 7 closed

Open issues

The axis starts at 80 so the weekly changes stay visible.

Labels on new issues

pharesee
13bug 🔥
11feature
6performance
3test
2triage
2core
1refactoring
1no label
10

How closed issues were closed

fixed by a PR 5fixed by a commit 1not planned or duplicate 1
  • #1483 Title: Allow access to unset fields; use debug-only checks for usability errors completed fixed by PR #1484
  • #1474 Module hierarchy: core/amr shouldn't depend on simulator-level options; simulator sits at the top completed fixed by PR #1325 feature
  • #1470 Convergence tests never run in CI duplicate closed by @igirault, no linked PR or commit
  • #1466 Electron velocity uses the mass-weighted ion velocity instead of the charge-weighted one completed fixed by PR #1471
  • #1459 bug in gate-sefltest completed fixed by commit 460f0b391 bug 🔥triage
  • #1421 Split diagnostics tests per dim/interp to prevent OOM when compiling completed fixed by PR #1422
  • #1096 configurator reports are per rank completed fixed by PR #1491
All 30 new issues
  • #1500 Quantity centering lookups: per-dimension switches are redundant and MHD table order mismatches the enum @PhilipDeegan
  • #1498 Refactor for_N stuff for C++20 @PhilipDeegan
  • #1497 Diag directory is not unique across tests @PhilipDeegan bug 🔥test
  • #1496 Projection stencils (momentsToEx, BxToEx, ...) are hand-written per quantity pair and never checked against the centerings @igirault corerefactoring
  • #1493 Time-dependent builtin external field (e.g. rotating dipole) @igirault
  • #1483 Title: Allow access to unset fields; use debug-only checks for usability errors @PhilipDeegan
  • #1480 Reject ill-formed boxes (lower > upper) instead of iterating them @PhilipDeegan feature
  • #1478 Level ghost interpolation test is skipped, loads the wrong quantities, and its comments are inverted @UCaromel testtriage
  • #1477 GridLayout::evalOnBox: captureless functions with forwarded state (GPU requirement); remove MeshIndex @PhilipDeegan feature
  • #1474 Module hierarchy: core/amr shouldn't depend on simulator-level options; simulator sits at the top @PhilipDeegan feature
  • #1470 Convergence tests never run in CI @igirault
  • #1466 Electron velocity uses the mass-weighted ion velocity instead of the charge-weighted one @nicolasaunai
  • #1465 Simulator: make the constructor noexcept, move the throwing setup into Simulator::setup() @PhilipDeegan
  • #1462 pharesee cannot read `pharevtkhdf` diagnostics @PhilipDeegan
  • #1461 `compute_hier_from` throws away the patch data type, so it can't be used on VTK (or other non-`FieldData`) hierarchies @PhilipDeegan
  • #1459 bug in gate-sefltest @igirault bug 🔥triage
  • #1458 plot2d: draw only the patches that intersect the view (about 20 lines, pixel-identical, 6–34× faster on zooms) @nicolasaunai phareseeperformance
  • #1457 pharesee: sample a hierarchy at arbitrary points with `hier.at(points, ...)` (trajectories, oblique cuts, planes) @nicolasaunai featurepharesee
  • #1456 pharesee: load only a region of a run (cached PatchIndex, LazyPatchArray, `GetX(t, region=...)`) @nicolasaunai featurephareseeperformanceI/O
  • #1455 pharesee: AMR-correct numpy semantics for hierarchies (scalar reductions, one ufunc dispatcher, pairing by box) @nicolasaunai featurepharesee
  • #1454 `hierarchy_from_func` silently returns `None` in 3D @nicolasaunai bug 🔥pharesee
  • #1453 Small plotting bugs: `plot2d` with no matching patch, `plot1d` return value and default levels, `plot_patches` without t=0 @nicolasaunai bug 🔥pharesee
  • #1452 `Run.GetB`/`GetE` with the default `all_primal=True` read each dataset 2–4 times, and fail in 3D @nicolasaunai bug 🔥phareseeperformance
  • #1451 `selection_box`: physical coordinates truncated to int, tuple form 2D-only, dict form crashes, lost on sub-hierarchies, empty levels kept @nicolasaunai bug 🔥pharesee
  • #1450 `plot2d` colour range: taken at the first time, includes ghost cells, and is computed even when `vmin`/`vmax` are given @nicolasaunai bug 🔥pharesee
  • #1449 `Vi / Ni` and `Vi / 2` crash on file-backed hierarchies (h5py `Dataset` divided directly) @nicolasaunai bug 🔥pharesee
  • #1448 `np.sqrt(h, out=h2)` segfaults: `out=` is not unwrapped in `__array_ufunc__` @nicolasaunai bug 🔥pharesee
  • #1447 Binary ops between hierarchies pair patches by list position; `are_compatible_hierarchies` always returns True @nicolasaunai bug 🔥pharesee
  • #1446 `ops.grad` returns d/d(index): missing grid spacing, wrong by a level-dependent factor dx @nicolasaunai bug 🔥pharesee
  • #1440 Ion-ion Coulomb collisions in hybrid runs @nicolasaunai

Pull requests

16 merged · 34 opened · 3 closed unmerged

Open PRs

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
1491 Write system info once per host instead of once per rank@PhilipDeegan
1489 One gtest xml report per test binary, so shared test directories stop losing or corrupting reports@nicolasaunai
1488 pharesee: vectorise Particles.select (py3_init-3d-particles 90 min -> 3 min)@nicolasaunai
1484 Allow access to unset fields; make usability checks debug-only@PhilipDeegan
1482 Ions: use core::isUsable/isSettable helpers@nicolasaunai
1475 NdArrayViewer: index in size_t and take the shape by reference, so per-point loops vectorize (2-4x on #1468 with clang)@igirault
1473 Particles no longer carry a charge (#1466, 3/3)@nicolasaunai
1472 Ion bulk velocity and mass density as diagnostics only (#1466, 2/3)@nicolasaunai
1471 Electron velocity from the ion charge flux (#1466, 1/3)@nicolasaunai
1460 Fix flaky gate-selftest: stop background git maintenance in fixture repos@igirault
1428 Declare Python dependencies and pin CI lint tools in pyproject.toml@igirault
1422 Split diagnostics tests per dim/interp to prevent OOM when compiling@PhilipDeegan
1414 Gate: flag tests a PR really removes, not edits to how they are registered@nicolasaunai
1325 rm phare_simulator_options.hpp and rework dependency hierarchy for mh…@PhilipDeegan
1318 Minimal boundary conditions@igirault
1314 Ruff errors assessed@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
1501 Drop box iterator@PhilipDeegan
1499 Derive projection stencils from centerings instead of 58 hand-written tables@igirault
1495 centerings lookup table instead of switch@PhilipDeegan
1494 Fix/mhd ghost consistency@UCaromel
1492 Perf/mhd merge schedules@UCaromel
1490 Polytropic electron closure@rochSmets
1487 Ohm vectorises: 5-27x faster@PhilipDeegan
1486 Read a field at another centring without copying it (RecenteredField)@igirault
1485 CI: run builds, CodeQL and coverage as one workflow so each PR push is one row on the Actions page@PhilipDeegan
1481 Perf/mhd kernels pr@UCaromel
1479 GridLayout::evalOnBox: captureless functions with forwarded state, remove MeshIndex@PhilipDeegan
1476 diagnostics: per-quantity accessor/computer registries, compute just before write@PhilipDeegan
1469 Choose the tagged quantities and add Löhner's criterion, within the existing tagging strategies@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
1464 MHD: split B into B0 (external field) + B1 (evolved)@igirault
1463 Conventions for C++ sources@igirault
1445 Validate collisions on refined levels and document them@nicolasaunai
1444 Ion-ion collisions in hybrid runs: Simulation(collisions=...)@nicolasaunai
1443 Exact momentum and energy conservation for collisions with unequal weights@nicolasaunai
1442 Binary Monte-Carlo ion-ion collision operator@nicolasaunai
1441 Deposit ion moments after the particle exchange@nicolasaunai
1229 MHD: split B into B0 (external/background) + B1 (evolved)@igirault
1228 redo finer schedules on regrid@UCaromel
1037 improved plot1d function@UCaromel
coreamrpythontestssimulator/iobuild/cidocother

People

human activity inside the week · darker = more

Who did what

CommitsPRs openedPRs mergedReviews givenIssues openedIssues closedCommentsLines on master
@PhilipDeegan 51156310119 +1,691 −1,904
@nicolasaunai 13117111554 +8,349 −1,168
@igirault 484174115 +6,683 −494
@UCaromel ·3·101·6 ·
@rochSmets ·1····· ·

Bots: coderabbitai (50 reviews, 34 comments), github-advanced-security (25 reviews, 0 comments), github-actions (0 reviews, 16 comments).

Who reviewed whom

reviewer ↓ · author →igiraultPhilipDeegannicolasaunaiUCaromel
PhilipDeegan43·137
igirault·17··
nicolasaunai46·1
UCaromel325·
Median first review11 h
Median time to merge2.2 d
15 open PRs waiting for a first human review · 1 with no activity for 30 days (as of 26 Oct)

Waiting for review

Stale

  • #1300release: bump SAMRAI and HighFive pins for the upcoming 0.1.0 release@nicolasaunai open 87.7 d

CI and tests

464,084 test executions on TeamCity · 20 failed · 4 ignored
passedfailedcanceledone tick per build, placed at its finish time
Mon 19MTue 20TWed 21WThu 22TFri 23FSat 24SSun 25S
GitHub Actions review-threads
115 ✓ · 54 ✗ · 251 canceled gate
61 ✓ · 39 ✗ · 38 canceled Coverage
53 ✓ · 16 ✗ · 43 canceled CodeQL
60 ✓ · 11 ✗ · 40 canceled labels
107 ✓ · 1 ✗ · 0 canceled CMake Ubuntu
48 ✓ · 20 ✗ · 36 canceled build
0 ✓ · 3 ✗ · 1 canceled pip install
1 ✓ · 0 ✗ · 1 canceled Graph Update: pip in /. #1648644929
1 ✓ · 0 ✗ · 0 canceled TeamCity gh_pr_gcc_samraisub
137 ✓ · 20 ✗ · 9 canceled gh_pr_clang_samraisub
131 ✓ · 24 ✗ · 10 canceled gh_pr_clang
115 ✓ · 19 ✗ · 7 canceled · 213,583 tests run gh_pr_gcc
119 ✓ · 13 ✗ · 8 canceled · 216,947 tests run gh_master_gcc_nightly_samraisub_phlop
6 ✓ · 1 ✗ · 0 canceled gh_master_clang_nightly_samraisub_phlop_asan
6 ✓ · 0 ✗ · 1 canceled gh_master_gcc_weekly_samraisub
1 ✓ · 0 ✗ · 0 canceled gh_master_clang_weekly_asan
1 ✓ · 0 ✗ · 0 canceled gh_master_gcc_samraisub_weekly_gcov
1 ✓ · 0 ✗ · 0 canceled gh_master_gcc_samraisub_weekly_caliper
1 ✓ · 0 ✗ · 0 canceled TeamCity, next CI (shadow) gcc (build + fast tests)
87 ✓ · 41 ✗ · 10 canceled · 11,315 tests run clang ASan (build + fast tests)
82 ✓ · 47 ✗ · 8 canceled · 10,650 tests run Pipeline Head
75 ✓ · 57 ✗ · 5 canceled · 11,589 tests run TeamCity, personal builds Phil · build
1 ✓ · 0 ✗ · 0 canceled

Tests that failed

(empty): test-collisions-relaxation
8RecenteredFieldProject/2: RecenteredFieldProject/2.MatchesProjectOnHybridPairs
4(empty): py3_time_step
3(empty): py3_patch_overlap
3(empty): test-recentered-field
2

Red, then green on the same commit

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

Code on master

22 commits · lines removed ← → lines added

Where the code changed

tests
1,568
7,217
core
432
4,389
python
688
1,937
amr
528
1,408
simulator/io
247
555
doc
6
674
build/ci
96
431
other
1
112

Commits

22 commits on master (also dots on the timeline)