Event by event random choice of one helicity #570
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…omenta' (prepare to add RndNum for color and helicity)
…mented out for now) There are seven other types that are equivalent to this one and could be removed...
…atrixElements was used instead of PinnedHostBufferGs HOWEVER this was harmless because the two types are the same! This is an example why BufferOneFp might simplify the code...
…ndomNumberKernels)
…and device Also improve the existing copies of gs and mes
…menta are different in F and C)
… (and reorder other arguments)
…lls with/without shared memory...
…ons as done in sigmaKin
./tmad/teeMadX.sh -ggtt +10x -makeclean ./tmad/teeMadX.sh -ggtt +10x -makeclean -fltonly ./tmad/teeMadX.sh -ggtt +10x -makeclean -mixonly Note that the bridge throughputs are a bit lower as there are many more copies of rnd/sel hel/col
…ing calculate_wavefunction in sigmakin
… to simplify helicity choice (madgraph5#403) Disable OMP loops for the moment
./tmad/teeMadX.sh -ggtt +10x -makeclean -mix
…n tput, looks good (a bit faster?)
…adgraph5#568) - will revert it and do it separately
… and new code madgraph5#568 - ok but cuda is 30% slower I guess the assert is not helping? Anyway, will revert and do it later, maybe.
…as "new" mulrichannel enabled madgraph5#568 Will revert the last changes anyway and move madgraph5#568 to a separate MR if any...
…st 4 commits Revert "[lhe] rerun tput tee ggtt with multichannel enabled, all ok, as slow as "new" mulrichannel enabled madgraph5#568" This reverts commit 445b321. Revert "[lhe] rerun tput ggtt mad with "#undef MGONGPU_SUPPORTS_MULTICHANNEL" and new code madgraph5#568 - ok but cuda is 30% slower" This reverts commit 3458706. Revert "[lhe] in ggtt.mad prototype a simplification of multichannel ifdefs (madgraph5#568) - will revert it and do it separately" This reverts commit 63e04b9. Revert "[lhe] rebuild ggtt with #undef MGONGPU_SUPPORTS_MULTICHANNEL and rerun tput, looks good (a bit faster?)" This reverts commit b9f534b.
./tmad/teeMadX.sh -ggtt +10x -makeclean -mix
./CODEGEN/generateAndCompare.sh gg_tt --mad --nopatch git diff --no-ext-diff -R gg_tt.mad/Source/dsample.f gg_tt.mad/Source/genps.inc gg_tt.mad/Source/vector.inc gg_tt.mad/SubProcesses/makefile > CODEGEN/MG5aMC_patches/PROD/patch.common git diff --no-ext-diff -R gg_tt.mad/SubProcesses/P1_gg_ttx/auto_dsig1.f gg_tt.mad/SubProcesses/P1_gg_ttx/driver.f gg_tt.mad/SubProcesses/P1_gg_ttx/matrix1.f > CODEGEN/MG5aMC_patches/PROD/patch.P1 git checkout gg_tt.mad
… CODEGEN (CPPProcess.cc fragments)
… CODEGEN (CPPProcess.h fragments)
… gg_tt.mad to CODEGEN The Fortran codegen was using allow_reverse=True and the cpp codegen allow_reverse=False. Now moved to allow_reverse=True also in cudacpp.
…remain 0 (no lhe files! madgraph5#403 madgraph5#402)
|
A few comments
While writing this I see that some tests failed in the CI, so there is more work to be done... |
This was
linked to
issues
Dec 14, 2022
|
Peculiar, it is some tests that fail, I can reproduce this interactively
The failures are that the events with given momenta give MEs which are not those in the refernce files?.... |
|
Ah voila, as I thought I had forgotten some hardcoded parameters in the codegen... |
|
OK I fixed the bug and regenerated all processes. It looks better in ggttgg interactively |
This was referenced Dec 14, 2022
… (slow) Also reorder the arguments to keep makeclean and sa last (easier to copy/paste)
…es short (skip 6 ggttggg to gain time) ./tput/allTees.sh -short This completes the random helicity choice madgraph5#403
|
This MR is now complete. The CI tests are succeeding, I will self merge |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Hi @oliviermattelaer @roiser I have finally ~completed the patch for the random choice of helicity #403
This MR does NOT include the random choice of color, which I have not yet done (and may be a bit more complex).
The MR is essentially complete with code regeneration of all processes. I am now rerunning the full tests to check performance and see functionalities.