Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
317 commits
Select commit Hold shift + click to select a range
f7a2c7a
MadSpin: run the sequential max-weight scan across cores
oliviermattelaer Jul 17, 2026
5e87a4d
MadSpin: report the per-particle decay efficiency in the progress line
oliviermattelaer Jul 17, 2026
53921e1
MadSpin: only one worker prints the progress lines
oliviermattelaer Jul 17, 2026
8f2a02a
fix ci to use the previous mode to check efficiency
oliviermattelaer Jul 17, 2026
57232cf
CI trigger
oliviermattelaer Jul 17, 2026
1973c3a
Fix CI: install meson+ninja for numpy 2.x f2py build backend
Copilot Jul 17, 2026
787f267
pip dedicated trigger for warm_cache
oliviermattelaer Jul 18, 2026
06a3ae7
debug CI
oliviermattelaer Jul 18, 2026
12a0d61
check
oliviermattelaer Jul 18, 2026
97f6c06
check
oliviermattelaer Jul 18, 2026
e0ef3a5
fix test
oliviermattelaer Jul 18, 2026
3bbb5d7
fix test
oliviermattelaer Jul 18, 2026
24afea3
fix test
oliviermattelaer Jul 18, 2026
7464778
fix test
oliviermattelaer Jul 18, 2026
c75a356
fix test
oliviermattelaer Jul 18, 2026
9d660d1
fix weird issue
oliviermattelaer Jul 18, 2026
0723f50
test
oliviermattelaer Jul 18, 2026
e44f0e1
test
oliviermattelaer Jul 18, 2026
5fda6dc
MadSpin plan: record the madspin (full offshell) sequential design
oliviermattelaer Jul 18, 2026
2fa4752
MadSpin plan: resolve the madspin normalisation via Olivier's telesco…
oliviermattelaer Jul 18, 2026
04a9e0d
MadSpin: implement the madspin (offshell) sequential path -- gated off
oliviermattelaer Jul 18, 2026
4548c7a
MadSpin: mass-set-level accept/reject for offshell sequential (still …
oliviermattelaer Jul 18, 2026
22c980e
MadSpin plan: correct the madspin conclusion -- density mode is the i…
oliviermattelaer Jul 18, 2026
fbaed1a
MadSpin: enable madspin/full sequential -- it works for resonant decays
oliviermattelaer Jul 18, 2026
e71678d
MadSpin sequential: wider max-weight margin, larger scan pool reserva…
oliviermattelaer Jul 19, 2026
7ba9070
MadSpin sequential: fix nb_core-times over-refill in the parallel max…
oliviermattelaer Jul 19, 2026
6ea6568
MadSpin sequential: fix the parallel scan reading only the first pool…
oliviermattelaer Jul 19, 2026
4c6aeac
MadSpin sequential: even parallel scan split (round nevents to a mult…
oliviermattelaer Jul 19, 2026
45e3274
MadSpin: run the joint (non-sequential) max-weight scan multi-process…
oliviermattelaer Jul 19, 2026
3eed34c
MadSpin: identify the worker in the per-shard "done" log line
oliviermattelaer Jul 19, 2026
addf765
fix all all issue
oliviermattelaer Jul 20, 2026
974ae31
fix a jacobian typo
oliviermattelaer Jul 20, 2026
d3b3802
fix the jacobian issue for madspin mode
oliviermattelaer Jul 21, 2026
b60faf1
better help for madspin option
oliviermattelaer Jul 21, 2026
31bde16
Merge pull request #323 from mg5amcnlo/madspin_sequential
oliviermattelaer Jul 21, 2026
aedfb49
Merge pull request #321 from mg5amcnlo/madspin_parralel
oliviermattelaer Jul 21, 2026
584c6b2
CI: run the unit and acceptance tests that no workflow triggered
oliviermattelaer Jul 21, 2026
24be592
change MadSpin defaults: spinmode=madspin, density_keep_jacobian=True…
oliviermattelaer Jul 21, 2026
92e67ca
Fix the failures revealed by running the previously untriggered tests
oliviermattelaer Jul 22, 2026
f3f4e34
Update the test_pptt_fksreal reference for the signed-zero normalisation
oliviermattelaer Jul 22, 2026
7822f30
corrections to LI density matrices
valentindurupt Jul 23, 2026
b2aac96
test_write_model: restore sys.path after importing the written model
oliviermattelaer Jul 23, 2026
c6d23fd
Merge pull request #324 from mg5amcnlo/claude/madspin-sequential-deca…
oliviermattelaer Jul 23, 2026
61c3823
fix lhapdf path registration so patched .info files are actually used
oliviermattelaer Jul 23, 2026
17368fd
log the reason when an lhapdf python import candidate is rejected
oliviermattelaer Jul 23, 2026
1f3405a
Merge pull request #327 from mg5amcnlo/claude/lhapdf-metadata-error-4…
oliviermattelaer Jul 23, 2026
375123c
reactivate buffering for f2py mode
oliviermattelaer Jul 25, 2026
29c8546
Merge pull request #326 from mg5amcnlo/claude/untriggered-test-functi…
oliviermattelaer Jul 25, 2026
c322f0a
Fix density matrix normalisation in density_debug check
Jul 27, 2026
292019c
Fix issue with too many open files
Jul 27, 2026
dbcb92f
MadSpin sequential: reproducible parallel decay-pool refill + deadloc…
oliviermattelaer Jul 27, 2026
b4c52ee
Computation of loop-induced density matrices ok
valentindurupt Jul 28, 2026
e2a90c4
Adding test for madspin and for density mode LI + bug corrections
valentindurupt Jul 29, 2026
040e63c
adding cache to test madspin LI
valentindurupt Jul 29, 2026
de4b786
Add P1TR/P1TL single-helicity propagators for massive vectors
oliviermattelaer Jul 29, 2026
c5b889c
Potential fix for pull request finding
valentindurupt Jul 30, 2026
c8bf6e2
Merge pull request #329 from mg5amcnlo/polarization_PM_propagator
oliviermattelaer Jul 30, 2026
beac03f
Merge remote-tracking branch 'origin/madspin_density' into madspin_va…
valentindurupt Jul 30, 2026
4e889ee
correction sequential_accept_reject
valentindurupt Jul 30, 2026
93f1ef0
correcting for test unittest_madspin_sequential
valentindurupt Jul 30, 2026
9d3dcec
Fix acceptancetest_madspin_LI: parallel-make race + torn MadLoopParam…
oliviermattelaer Jul 30, 2026
ba128f3
Address the Copilot review findings on PR #330
oliviermattelaer Jul 30, 2026
37bcee9
MadSpin: put back the defensive misc.sprint debug prints
oliviermattelaer Jul 30, 2026
e474d58
MadSpin: seed the RNG of the forked decay-generation children
oliviermattelaer Jul 31, 2026
275462e
Merge pull request #330 from mg5amcnlo/madspin_valentin
valentindurupt Aug 13, 2026
9dbcbbc
Fix wrong polarisation state in LO me_frame with a single selected pa…
oliviermattelaer Aug 13, 2026
64cd43f
UpdateNotes: fix English typos throughout the file
oliviermattelaer Aug 13, 2026
7810e10
Update release notes for version 3.7.3
oliviermattelaer Aug 13, 2026
554e41d
Merge pull request #333 from mg5amcnlo/claude/me-frame-single-leg-boost
oliviermattelaer Aug 13, 2026
5e0d0a2
MadSpin sequential: restore the offshell rate factor in the mass-set …
oliviermattelaer Aug 14, 2026
2aac8f4
MadSpin sequential: version the offshell cache, tighten the Z fit's d…
oliviermattelaer Aug 14, 2026
a24008a
MadSpin sequential: normalise the offshell mass weight by |M_prod|^2 …
oliviermattelaer Aug 14, 2026
7779950
MadSpin sequential: one bound over all the angles, reusing the mass set
oliviermattelaer Aug 14, 2026
3a6d520
MadSpin sequential: Z_hat is converged at the default probe depth
oliviermattelaer Aug 14, 2026
eda57c8
MadSpin sequential: verify the weight identity deterministically
oliviermattelaer Aug 14, 2026
9168426
MadSpin sequential: drop variant B
oliviermattelaer Aug 14, 2026
a8e2202
MadSpin: one 'unweighting' option for the four accept/reject schemes
oliviermattelaer Aug 14, 2026
2cff6bb
MadSpin: auto picks joint when a single particle decays
oliviermattelaer Aug 14, 2026
6c051b6
MadSpin: say once which unweighting scheme a run used
oliviermattelaer Aug 14, 2026
74887cf
MadSpin: record two findings from the me_frame/beampol branch
oliviermattelaer Aug 15, 2026
e48efa7
MadSpin: pair decay branches to identical-PID mothers in order
oliviermattelaer Aug 15, 2026
68ce2ee
Merge pull request #334 from mg5amcnlo/claude/madspin-sequential-offs…
oliviermattelaer Aug 15, 2026
d04f6ea
Fix FourMomentum.boost for a boost momentum that is already at rest
oliviermattelaer Aug 15, 2026
68a296e
MadSpin v1: apply the beam polarisation before the helicity sum, not …
oliviermattelaer Aug 15, 2026
b231141
MadSpin: honour me_frame and beampol in the density spinmodes
oliviermattelaer Aug 15, 2026
ec811a3
MadSpin: the beampol option speaks the run_card's convention
oliviermattelaer Aug 15, 2026
576047d
MadSpin: refuse a one-sided beampol when the card is read
oliviermattelaer Aug 15, 2026
0c91195
MadSpin density modes: warn that '@' decay grouping tags are ignored
oliviermattelaer Aug 15, 2026
b6dfe25
MadSpin: parse and validate the '@' decay grouping tags
oliviermattelaer Aug 15, 2026
fc23b1f
MadSpin density modes: honour the '@' decay grouping tags
oliviermattelaer Aug 15, 2026
38151a6
MadSpin: measure the grouped sample against two dedicated runs
oliviermattelaer Aug 15, 2026
2d5f6ce
MadSpin: validate the grouping with two parents per pdg (p p > t t~ t…
oliviermattelaer Aug 15, 2026
4ba6e80
MadSpin: check the grouping under PA and sequential_decay
oliviermattelaer Aug 15, 2026
c43e5a4
MadSpin: say what _clamped_partial_width actually does, and pin it
oliviermattelaer Aug 15, 2026
3d65d68
MadSpin: pin the grouping fallback after the rebase onto the unweight…
oliviermattelaer Aug 15, 2026
ec2b1dc
MadSpin: give the PA spinmode the up-front mass draw
oliviermattelaer Aug 15, 2026
2626564
MadSpin: replicas for the remaining two PA up-front schemes
oliviermattelaer Aug 15, 2026
4f03a39
MadSpin: measure the PA unweighting schemes on a 250k sample
oliviermattelaer Aug 16, 2026
fb5e036
MadSpin: record the PA up-front-mass-draw follow-up in the plan
oliviermattelaer Aug 16, 2026
07ae82e
MadSpin: scan the unweighting schemes over the decay multiplicity
oliviermattelaer Aug 16, 2026
2ce853a
MadSpin: switch the auto unweighting to the two-branch rule
oliviermattelaer Aug 16, 2026
bce1d44
In madgraph_interface.py convert self.options['auto_update'] to float…
valassi Aug 17, 2026
141e069
Merge pull request #339 from valassi/avalassi_autoupdate
oliviermattelaer Aug 18, 2026
bf90eae
Merge pull request #338 from mg5amcnlo/claude/madspin-plan-pa-followup
oliviermattelaer Aug 18, 2026
8a188e5
Merge pull request #336 from mg5amcnlo/claude/madspin-frame-beampol-m…
oliviermattelaer Aug 18, 2026
71f3e17
Merge pull request #337 from mg5amcnlo/claude/pensive-haibt-326ba9
oliviermattelaer Aug 18, 2026
7855704
MadSpin: CI coverage for the unweighting schemes and their consistency
oliviermattelaer Aug 15, 2026
05a92e1
MadSpin: pair decay branches to identical-PID mothers in order
oliviermattelaer Aug 18, 2026
d04b39d
Merge madspin_density into the PA up-front-mass branch
oliviermattelaer Aug 18, 2026
03c9961
Merge pull request #342 from mg5amcnlo/claude/compassionate-elgamal-d…
oliviermattelaer Aug 18, 2026
d1b631f
Merge pull request #340 from mg5amcnlo/claude/trusting-chatterjee-77cc57
oliviermattelaer Aug 18, 2026
349f9be
MadSpin: rename the up-front-mass max-weight cache helpers
oliviermattelaer Aug 18, 2026
dc374a3
MadSpin: hide the two_stage unweighting scheme from the user interface
oliviermattelaer Aug 18, 2026
09c2a6b
MadSpin: document (and assert) the slot/parents pairing in _check_wei…
oliviermattelaer Aug 18, 2026
679cb29
MadSpin: name the spinmode/scheme predicates behind the unweighting d…
oliviermattelaer Aug 18, 2026
23ecb49
Merge pull request #341 from mg5amcnlo/claude/madspin-decay-branch-pa…
oliviermattelaer Aug 18, 2026
43437d6
MadSpin: count the decaying particles per event, not per pdg
oliviermattelaer Aug 18, 2026
ef07db3
Merge pull request #343 from mg5amcnlo/claude/ms-offshell-cache-rename
oliviermattelaer Aug 18, 2026
db37225
MadSpin tests: fix the stale unweighting expectations in the factory …
oliviermattelaer Aug 18, 2026
7a4f362
MadSpin: restrict the density convolution to the production polarisation
oliviermattelaer Aug 18, 2026
301fd29
Merge pull request #344 from mg5amcnlo/claude/ms-hide-two-stage
oliviermattelaer Aug 18, 2026
61c9ff4
Merge pull request #345 from mg5amcnlo/claude/ms-pa-parents-index
oliviermattelaer Aug 18, 2026
dfc8f65
MadSpin: assess a pure-interference mode, and land the cross restriction
oliviermattelaer Aug 18, 2026
9dc4215
Merge pull request #347 from mg5amcnlo/claude/ms-ci-wj-ms-decay
oliviermattelaer Aug 18, 2026
f933bd3
Merge remote-tracking branch 'origin/madspin_density' into tmp-t5
oliviermattelaer Aug 18, 2026
69a9c22
Merge pull request #348 from mg5amcnlo/claude/ms-factory-unweighting-…
oliviermattelaer Aug 18, 2026
bf5f073
MadSpin: keep_weight_for_polarization, extra LHEF v3 weights per pola…
oliviermattelaer Aug 18, 2026
e20303f
MadSpin: gather the restricted density rows by index, not by mask
oliviermattelaer Aug 18, 2026
11e07a6
MadSpin: record the end-to-end validation of the polarised convolution
oliviermattelaer Aug 18, 2026
abb7ddf
Merge madspin_density (PR #348) to fix the red unweighting_resolution…
oliviermattelaer Aug 18, 2026
ddc9017
MadSpin: same-pdg mixed polarisation (p p > z{0} z{T})
oliviermattelaer Aug 18, 2026
1adf965
MadSpin: fail loudly on an accept/reject that can never accept
oliviermattelaer Aug 18, 2026
fa6df97
MadSpin: define the production polarisation in the me_frame frame
oliviermattelaer Aug 18, 2026
6ad19d7
MadSpin: raise the intended error on an invalid '@' order restriction
oliviermattelaer Aug 18, 2026
ccc5879
allow dsqrt_shatmax to be a float
oliviermattelaer Aug 18, 2026
30e0596
Merge pull request #355 from mg5amcnlo/claude/ms-polarization-frame
oliviermattelaer Aug 18, 2026
6db6b9d
Merge pull request #346 from mg5amcnlo/claude/ms-unweighting-helpers
oliviermattelaer Aug 18, 2026
20068aa
Merge remote-tracking branch 'origin/madspin_density' into fix356
oliviermattelaer Aug 18, 2026
79f465c
MadSpin: polarisation weights as the product of the per-particle choices
oliviermattelaer Aug 18, 2026
27bfd92
MadSpin: fix get_inter_value unpacking get_pdir with the wrong arity
oliviermattelaer Aug 18, 2026
86c37e4
Merge pull request #354 from mg5amcnlo/claude/ms-zero-density-guard
oliviermattelaer Aug 18, 2026
2d01785
Merge pull request #356 from mg5amcnlo/claude/ms-madspinerror-nameerror
oliviermattelaer Aug 18, 2026
7e35f77
Merge claude/ms-density-pol-restrict into claude/ms-pure-interference
oliviermattelaer Aug 18, 2026
bc37807
Merge pull request #350 from mg5amcnlo/claude/ms-density-restriction-…
oliviermattelaer Aug 18, 2026
4e7c66a
Refactor comments in madspin_card_default.dat
oliviermattelaer Aug 18, 2026
81d5274
Merge pull request #357 from mg5amcnlo/claude/ms-polarization-weight-…
oliviermattelaer Aug 18, 2026
221cefd
Merge pull request #353 from mg5amcnlo/claude/ms-samepdg-polarization
oliviermattelaer Aug 18, 2026
421fc17
Merge pull request #352 from mg5amcnlo/claude/ms-keep-weight-polariza…
oliviermattelaer Aug 18, 2026
6456a45
Merge branch 'madspin_density' into claude/ms-get-inter-value-arity
oliviermattelaer Aug 18, 2026
e514dab
Merge pull request #358 from mg5amcnlo/claude/ms-get-inter-value-arity
oliviermattelaer Aug 18, 2026
cd4dd9d
MadSpin: remove the dead get_inter_value/get_nhel/get_mymod cluster
oliviermattelaer Aug 18, 2026
0a9ab6c
MadSpin: implement the pure-interference mode
oliviermattelaer Aug 18, 2026
1f59eb3
Merge remote-tracking branch 'origin/madspin_density' into t23fix
oliviermattelaer Aug 18, 2026
e1467f4
MadSpin: count the overweight trials of the pure-interference mode
oliviermattelaer Aug 18, 2026
5b687c4
MadSpin: record the end-to-end validation of the pure-interference mode
oliviermattelaer Aug 18, 2026
5c343f0
Merge remote-tracking branch 'origin/madspin_density' into HEAD
oliviermattelaer Aug 18, 2026
d78947d
Merge pull request #349 from mg5amcnlo/claude/ms-density-pol-restrict
oliviermattelaer Aug 18, 2026
a1c7eca
Merge pull request #359 from mg5amcnlo/claude/ms-drop-dead-f2py-cluster
oliviermattelaer Aug 18, 2026
e148f8a
MadSpin: 'auto' takes sequential when the production is polarised
oliviermattelaer Aug 18, 2026
bdf3835
Merge remote-tracking branch 'origin/madspin_density' into HEAD
oliviermattelaer Aug 18, 2026
1394688
Merge origin/claude/ms-density-pol-restrict: re-seat the polarised 'a…
oliviermattelaer Aug 18, 2026
d1a8490
MadSpin: take the polarisation weights on the me_frame axis
oliviermattelaer Aug 18, 2026
c386b82
tests: name the unweighting decision-table case fields
oliviermattelaer Aug 18, 2026
2160750
Merge pull request #361 from mg5amcnlo/claude/ms-frame-axis-wiring
oliviermattelaer Aug 18, 2026
ee3856f
Merge pull request #360 from mg5amcnlo/claude/ms-auto-sequential-pola…
oliviermattelaer Aug 18, 2026
f098cdb
MadSpin pure_interference: fully weighted output, w = sigma*BR*W/c
oliviermattelaer Aug 18, 2026
dc5f4e4
MadSpin pure_interference: card syntax, diagonal blocks, and refusals
oliviermattelaer Aug 18, 2026
c216be3
MadSpin plan: record 13.13-13.15 (fully weighted output, syntax, 9 vs…
oliviermattelaer Aug 18, 2026
c2330bd
MadSpin pure_interference: the launch warning describes the new output
oliviermattelaer Aug 18, 2026
6757c64
MadSpin pure_interference: end-to-end validation of the fully weighte…
oliviermattelaer Aug 18, 2026
d19d8a2
Merge remote-tracking branch 'origin/madspin_density' into HEAD
oliviermattelaer Aug 18, 2026
41fd25b
MadSpin: move MADSPIN_SEQUENTIAL_PLAN.md to doc/madspin_sequential_pl…
oliviermattelaer Aug 18, 2026
92928b7
Merge pull request #362 from mg5amcnlo/claude/ms-move-plan-to-doc
oliviermattelaer Aug 18, 2026
371e09a
MadSpin pure_interference: an unweighted-up-to-a-sign output option
oliviermattelaer Aug 18, 2026
569f419
MadSpin pure_interference: normalise the unweighted output on the run…
oliviermattelaer Aug 18, 2026
3accc6d
MadSpin: decay_output = weighted, for every density mode
oliviermattelaer Aug 18, 2026
9f855d6
MadSpin: the 'unweighted' interference path must not be tested twice
oliviermattelaer Aug 18, 2026
1cb7c0d
MadSpin plan: record the 13.18 regression and the final test count
oliviermattelaer Aug 18, 2026
5bc0a7a
MadSpin: reusing an ms_dir no longer zeroes the branching ratio
oliviermattelaer Aug 18, 2026
9bd00b7
MadSpin ms_dir reuse: regression tests for the zero branching ratio
oliviermattelaer Aug 18, 2026
b939412
MadSpin madspin_v1: write the decayed events where the caller reads them
oliviermattelaer Aug 18, 2026
3d36428
MadSpin ms_dir: archive the card that was run, not the one curr_dir p…
oliviermattelaer Aug 19, 2026
c471c46
Merge pull request #364 from mg5amcnlo/claude/ms-msdir-zero-br
oliviermattelaer Aug 19, 2026
1be268b
Merge pull request #366 from mg5amcnlo/claude/ms-msdir-card-archive
oliviermattelaer Aug 19, 2026
01f33ee
Merge pull request #365 from mg5amcnlo/claude/ms-v1-msdir-gzip-path
oliviermattelaer Aug 19, 2026
39b834b
MadSpin density: the accept/reject weight is a real number, and says so
oliviermattelaer Aug 19, 2026
a0b9195
Fix MadLoop standalone Source/makefile under external dependencies
oliviermattelaer Aug 19, 2026
0711f83
Merge pull request #367 from mg5amcnlo/claude/ms-complex-weight-check
oliviermattelaer Aug 19, 2026
10a963e
Stop the standalone Source 'make' fallback from hiding real failures
oliviermattelaer Aug 19, 2026
535c1db
MadSpin fixed_order: decay every member of an event group
oliviermattelaer Aug 19, 2026
fbd42ef
Clean up comments in export_v4.py
oliviermattelaer Aug 19, 2026
60c9c95
Merge pull request #369 from mg5amcnlo/claude/mg5-exporter-make-except
oliviermattelaer Aug 19, 2026
6625cd7
Merge pull request #368 from mg5amcnlo/claude/mg5-sqrvirt-cuttools
oliviermattelaer Aug 19, 2026
ca1ad45
Merge pull request #363 from mg5amcnlo/claude/ms-interference-unweighted
oliviermattelaer Aug 19, 2026
14ab076
Merge branch 'madspin_density' into claude/ms-pure-interference
oliviermattelaer Aug 19, 2026
90d6ed3
Merge pull request #351 from mg5amcnlo/claude/ms-pure-interference
oliviermattelaer Aug 19, 2026
3d399b0
Merge remote-tracking branch 'origin/madspin_density' into HEAD
oliviermattelaer Aug 19, 2026
e80c476
Merge pull request #370 from mg5amcnlo/claude/ms-fixed-order-decays
oliviermattelaer Aug 19, 2026
adb31ea
MadSpin: fold pure_interference_output into decay_output, add 'auto'
oliviermattelaer Aug 19, 2026
26dfd8f
MadSpin: remove the sequential_decay and keep_weight_for_polarization…
oliviermattelaer Aug 19, 2026
c683cdd
MadSpin: fix a stale docstring reference to pure_interference_output
oliviermattelaer Aug 19, 2026
179ebdc
Merge pull request #371 from mg5amcnlo/claude/ms-unify-decay-output
oliviermattelaer Aug 19, 2026
7be269b
doc/madspin_sequential_plan: drop the planning scaffolding for work t…
oliviermattelaer Aug 19, 2026
612708e
doc/madspin_sequential_plan: sections 10-11 in the names the schemes …
oliviermattelaer Aug 19, 2026
0535605
doc/madspin_sequential_plan: sections 12-13 against the options that …
oliviermattelaer Aug 19, 2026
152eeba
doc/madspin_decay_groups: current option names, and the lineshape bia…
oliviermattelaer Aug 19, 2026
21506c7
doc: last of the planning voice, and two dangling references
oliviermattelaer Aug 19, 2026
0c9ac1f
Merge pull request #372 from mg5amcnlo/claude/ms-doc-cleanup
oliviermattelaer Aug 19, 2026
1306d7a
MadSpin: fix the "PA (default)" label on the post-acceptance reshuffle
oliviermattelaer Aug 19, 2026
a311d2e
Merge pull request #374 from mg5amcnlo/claude/madspin-pa-comment-fix
oliviermattelaer Aug 19, 2026
c8b79c0
MadSpin: carry the overweight instead of clipping it, in every stage
oliviermattelaer Aug 19, 2026
eae4af0
MadSpin: the overweight escalation must use the bound it tested against
oliviermattelaer Aug 19, 2026
d0019d4
MadSpin: make the overweight report weight-aware, and cover negative …
oliviermattelaer Aug 19, 2026
87195c3
MadSpin: bound the mass stage per production event, for any 2 -> N
oliviermattelaer Aug 19, 2026
d2253d3
MadSpin: measure the offshell mass-stage bound, and keep the global one
oliviermattelaer Aug 19, 2026
46c1feb
MadSpin tests: parse the overweight line under its current wording
oliviermattelaer Aug 20, 2026
d6605ef
MadSpin: publish a refill generation only once its files exist
oliviermattelaer Aug 20, 2026
d6023a0
Merge pull request #375 from mg5amcnlo/claude/ms-overweight-safety-net
oliviermattelaer Aug 20, 2026
ded1b05
Merge pull request #378 from mg5amcnlo/claude/ms-refill-race-density
oliviermattelaer Aug 20, 2026
cfc248e
Merge pull request #376 from mg5amcnlo/claude/ms-maxweight-mismatch
oliviermattelaer Aug 20, 2026
7926379
MadSpin: guard that the offshell spinmodes keep the global mass bound
oliviermattelaer Aug 20, 2026
538505d
Merge pull request #377 from mg5amcnlo/claude/ms-per-event-mass-bound
oliviermattelaer Aug 20, 2026
f9ee215
MadSpin: the reported cross-section now carries the Breit-Wigner trun…
oliviermattelaer Aug 20, 2026
a3d2352
MadSpin: validate the Breit-Wigner truncation, and correct the one te…
oliviermattelaer Aug 20, 2026
178e954
Merge pull request #379 from mg5amcnlo/claude/ms-bw-truncation-norm-2
oliviermattelaer Aug 21, 2026
0f9b647
Fix "Too many open files" when collecting events from many workers
oliviermattelaer Aug 22, 2026
da27777
Bound the descriptor use of the external shuffle path
oliviermattelaer Aug 22, 2026
7b3b3ec
MadSpin: tag the overweights that sit where the NWA cannot hold
oliviermattelaer Aug 22, 2026
9c729f3
MadSpin: the 2 -> 3 joint overweights are the production's own resonance
oliviermattelaer Aug 23, 2026
77e46e6
MadSpin: test "shat near its minimum" against the 2 -> 3 overweights …
oliviermattelaer Aug 23, 2026
e866909
MadSpin: record why the production-resonance pairing accepts a status…
oliviermattelaer Aug 23, 2026
083fb7e
reduce verbosity of overweight information
oliviermattelaer Aug 23, 2026
8629983
MadSpin: make the quiet overweight line safe, and test both regimes
oliviermattelaer Aug 23, 2026
fa04b40
Merge pull request #384 from mg5amcnlo/claude/ms-joint-bound-2to3
oliviermattelaer Aug 23, 2026
02ee49a
Merge pull request #383 from mg5amcnlo/claude/ms-threshold-overweights
oliviermattelaer Aug 23, 2026
83a74f4
Size the descriptor budget from the job, and drop the bookkeeping whe…
oliviermattelaer Aug 24, 2026
fb8ece1
Merge pull request #381 from mg5amcnlo/claude/fix-collect-events-fd-l…
oliviermattelaer Aug 25, 2026
6520d74
Merge remote-tracking branch 'mg5/3.7.3' into madspin_branch
oliviermattelaer Aug 25, 2026
e5c5de9
Merge remote-tracking branch 'mg5/madspin_density' into madspin_branch
oliviermattelaer Aug 25, 2026
4903c0f
Fix the standalone libmodel rule building DHELAS instead of MODEL
oliviermattelaer Aug 25, 2026
ae60cc9
CI: drop the stdhep / onlygen acceptance jobs re-enabled by madspin_d…
oliviermattelaer Aug 25, 2026
7d488d2
MadSpin: do not call initialise() on the density mode's module list
oliviermattelaer Aug 25, 2026
82442a8
Report the real error when 'make' fails in a standalone Source directory
oliviermattelaer Aug 25, 2026
afa257e
MadSpin: keep the unweight-efficiency summary visible under the drivers
oliviermattelaer Aug 25, 2026
3a3e2c4
ALOHA: route the FLV flag through 'tags' for amplitudes in the cross-…
oliviermattelaer Aug 25, 2026
424c011
tests: expect the merged process in the p p > w+ w- check-gauge tests
oliviermattelaer Aug 25, 2026
d5c17ba
CI: give unittest_10 numpy for the madspin unit tests
oliviermattelaer Aug 25, 2026
6644aa3
tests: build the read-only gridpack with the madevent exporter
oliviermattelaer Aug 25, 2026
993d115
tests: build the parallel-Delphes process with the madevent exporter
oliviermattelaer Aug 25, 2026
e0a95d2
update IOTest
oliviermattelaer Aug 25, 2026
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
15 changes: 14 additions & 1 deletion .github/actions/restore-pip-cache/action.yml
Original file line number Diff line number Diff line change
Expand Up @@ -17,6 +17,19 @@ runs:
sudo apt install meson ninja-build
echo "PYTHONPATH=$PYTHONPATH:$HOME/.cache/pip-packages" >> $GITHUB_ENV
echo "$HOME/.cache/pip-packages/bin" >> $GITHUB_PATH
# Install meson+ninja if missing (numpy 2.x f2py requires meson as build backend).
# --upgrade is needed so pip writes scripts into the existing bin/ directory;
# it only affects meson/ninja since numpy is not listed here.
if [ ! -x "$HOME/.cache/pip-packages/bin/meson" ]; then
pip install --target "$HOME/.cache/pip-packages" --upgrade meson ninja
fi
cd $GITHUB_WORKSPACE
echo "f2py_compiler = $(which f2py)" >> input/mg5_configuration.txt
# Use direct path because GITHUB_PATH additions don't take effect in the same step
F2PY_BIN="$HOME/.cache/pip-packages/bin/f2py"
if [ -x "$F2PY_BIN" ]; then
echo "f2py_compiler = $F2PY_BIN" >> input/mg5_configuration.txt
else
pip install --target "$HOME/.cache/pip-packages" --upgrade numpy
echo "f2py_compiler = $F2PY_BIN" >> input/mg5_configuration.txt
fi
shell: bash
36 changes: 36 additions & 0 deletions .github/actions/restore_delphes/action.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,36 @@
name: Restore Delphes cache
description: Restores Delphes (and its ROOT dependency) from a dedicated cache and configures MG5

# Delphes depends only on ROOT. Both live in the dedicated ~/.cache/Delphes
# cache (kept separate from HEPtools); the delphes-<os> cache is stacked on top
# of the root-<os> one, exactly like the HEPtools sub-caches.

inputs:
setup_only:
required: false
default: "false"

runs:
using: composite
steps:
# Ensure the ROOT environment (ROOTSYS/PATH/LD_LIBRARY_PATH) is set up,
# restoring the root cache only if a higher level did not already do so.
- name: restore ROOT
uses: ./.github/actions/restore_root
with:
setup_only: ${{ inputs.setup_only }}

- name: Set cache key
run: echo "CACHE_KEY=delphes-$ImageOS" >> $GITHUB_ENV
shell: bash

- uses: actions/cache/restore@v5
if: ${{ inputs.setup_only != 'true' }}
with:
path: ~/.cache/Delphes
key: ${{ env.CACHE_KEY }}

- run: |
cd $GITHUB_WORKSPACE
echo "delphes_path = $HOME/.cache/Delphes/Delphes" >> input/mg5_configuration.txt
shell: bash
84 changes: 84 additions & 0 deletions .github/actions/restore_root/action.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,84 @@
name: Restore ROOT cache
description: Restores ROOT from its dedicated cache (separate from HEPtools) and sets ROOTSYS

# ROOT and Delphes are kept in their own cache (~/.cache/Delphes) on purpose:
# they are large and rarely change, so they should not be bundled with - nor
# invalidated together with - the HEPtools cache. They are warmed and cleaned
# independently (see the root_cache/delphes_cache jobs and the reset_delphes
# input of warm_cache.yml).

inputs:
setup_only:
required: false
default: "false"

runs:
using: composite
steps:
- name: Set cache key
run: echo "CACHE_KEY=root-$ImageOS" >> $GITHUB_ENV
shell: bash

- name: Restore ROOT cache
id: cache
if: ${{ inputs.setup_only != 'true' }}
uses: actions/cache/restore@v5
with:
path: ~/.cache/Delphes
key: ${{ env.CACHE_KEY }}

# ROOT (6.40 binaries) links against Intel TBB at build- and run-time. The
# apt packages are not part of the cache, so install the runtime dependency
# wherever ROOT is used (Delphes build and the Delphes test jobs). The
# ldconfig guard makes this a no-op on images that already ship it (e.g.
# ubuntu-22.04); ubuntu-24.04 needs it (rootcint fails on libtbb.so.12).
- name: Install ROOT runtime dependencies
run: |
if ! ldconfig -p | grep -q 'libtbb\.so\.12'; then
sudo apt-get update
sudo apt-get install -y --no-install-recommends libtbb12
fi
shell: bash

# Self-heal on a cache miss instead of silently configuring a non-existent
# ROOTSYS: when this level is the one responsible for restoring ROOT
# (setup_only != true) and the cache was not found, install ROOT inline,
# exactly like the root_cache warm job. Keep ROOT_VERSION / the tarball
# names in sync with the root_cache job in warm_cache.yml.
- name: Install ROOT (cache miss fallback)
if: ${{ inputs.setup_only != 'true' && steps.cache.outputs.cache-hit != 'true' }}
run: |
echo "::warning::ROOT cache miss for ${CACHE_KEY}; installing ROOT inline."
ROOT_VERSION=6.40.02
case "$ImageOS" in
ubuntu22) ROOT_TARBALL=root_v${ROOT_VERSION}.Linux-ubuntu22.04-x86_64-gcc11.4.tar.gz ;;
ubuntu24) ROOT_TARBALL=root_v${ROOT_VERSION}.Linux-ubuntu24.04-x86_64-gcc13.3.tar.gz ;;
*) echo "Unsupported runner image: $ImageOS"; exit 1 ;;
esac
mkdir -p "$HOME/.cache/Delphes"
cd "$HOME/.cache/Delphes"
rm -rf root
wget -q "https://root.cern/download/${ROOT_TARBALL}"
tar -xzf "${ROOT_TARBALL}"
rm -f "${ROOT_TARBALL}"
shell: bash

- name: Configure ROOT environment
run: |
echo "ROOTSYS=$HOME/.cache/Delphes/root" >> $GITHUB_ENV
echo "$HOME/.cache/Delphes/root/bin" >> $GITHUB_PATH
echo "LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$HOME/.cache/Delphes/root/lib" >> $GITHUB_ENV
echo "PYTHONPATH=$PYTHONPATH:$HOME/.cache/Delphes/root/lib" >> $GITHUB_ENV
shell: bash

# Fail fast (only when this level is responsible for ROOT) if ROOT is still
# missing after restore + fallback install, rather than letting downstream
# steps run against a non-existent ROOTSYS.
- name: Verify ROOT is available
if: ${{ inputs.setup_only != 'true' }}
run: |
if [ ! -x "$HOME/.cache/Delphes/root/bin/root-config" ]; then
echo "::error::ROOT is not available at $HOME/.cache/Delphes/root after restore/install."
exit 1
fi
shell: bash
Loading
Loading