MC: Ability to process empty timeframes (Part 1) - #15705
Open
sawenzel wants to merge 5 commits into
Open
Conversation
sawenzel
requested review from
a team,
afurs,
andreasmolander,
arossi81,
bazinski,
coppedis,
davidrohr,
f3sch,
fcolamar,
fgrosa,
fprino,
gvolpe79,
jotwinow,
kharlov,
mconcas,
mcoquet642,
njacazio,
noferini,
peressounko,
sahilupadhyaya92,
wiechula and
wille10
as code owners
August 21, 2026 16:46
sawenzel
force-pushed
the
sawenzel/colcontext-empty-tf
branch
from
August 24, 2026 11:15
6d91a54 to
41757d6
Compare
sawenzel
changed the title
This fixes a problem in the timeframe index structure of the collision context and adds a unit test. - getTimeFrameBoundaries closed only one timeframe per collision, so a timeframe without collisions was left out of the index structure entirely and the collisions after it were assigned to the wrong timeframe. - The number of extracted per-timeframe contexts was therefore the number of non-empty timeframes, not the number of timeframes asked for, and the last tf<N>/collisioncontext.root could be missing. - The scan now closes every timeframe a collision skips over and pads the result to the number of timeframes the caller asks for, so entry i always describes orbits [start + i*orbitsPerTF, start + (i+1)*orbitsPerTF). - applyMaxCollisionFilter keeps an empty timeframe empty when it re-indexes, and extractSingleTimeframe returns a valid empty context for it. - o2-steer-colcontexttool passes the number of timeframes it asked for, reports timeframes that came out empty together with the mean number of collisions per timeframe implied by the interaction rate, and refuses to continue when --noEmptyTF was requested. https://its.cern.ch/jira/browse/O2-7132 Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
This fixes a problem in GenTPCLoopers::setFlatGas when the collision context of a timeframe is empty. - A timeframe holds no collision whenever the interaction rate is low enough, and the generator called exit(1) on it. - The extent of the timeframe now comes from HBFUtils in that case, which is where it is defined, instead of from the last collision. - With a single collision in the timeframe the mean interaction spacing was divided by zero; it is now taken from the interaction rate stored in the collision context. https://its.cern.ch/jira/browse/O2-7132 Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
This fixes a segmentation fault in MCHDPLDigitizerTask when the collision context of a timeframe is empty. - The noise-only signal range was taken from eventRecords.front() and eventRecords.back(), which is undefined behaviour on an empty vector. - A timeframe holds no collision whenever the interaction rate is low. - The range now comes from HBFUtils in that case, so the noise covers the timeframe that is actually being digitised. https://its.cern.ch/jira/browse/O2-7132 Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
This fixes the TPC digit writer producing a file without a tree when a timeframe holds no collision. - The custom close callback only called TFile::Close inside "if (entries > 0)", and never called TFile::Write, so with nothing to write the tree never reached the file. - The result was a 942 byte file with no o2sim tree, and every reader of it failed on a missing branch rather than on an empty tree. - Each branch is now filled once with the empty default object it is bound to, so the file is an ordinary timeframe that happens to contain no digit and the readers downstream stay on their normal path. - The tree is written explicitly, the way RootTreeWriter's own close does. https://its.cern.ch/jira/browse/O2-7132 Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
This fixes the digit file of a timeframe without collisions carrying a label branch of the wrong type. - The MC label branch is declared as std::vector<char> and only becomes an IOMCTruthContainerView when a fill remaps it. - With no collision nothing is filled, the branch keeps the raw type, and a reader binding IOMCTruthContainerView gets a class mismatch from SetBranchAddress rather than an empty tree. - The close callback now writes one empty entry in every branch, remapping the label branch on the way, so the file has the same shape as an ordinary timeframe that happens to contain nothing. https://its.cern.ch/jira/browse/O2-7132 Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
sawenzel
force-pushed
the
sawenzel/colcontext-empty-tf
branch
from
August 24, 2026 11:28
41757d6 to
3f49e82
Compare
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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.
Fix of problems with empty timeframe handling in
It it is one step towards successful completion of the compute chain when we have an empty timeframe. Currently we fault for one or another reason and the GRID jobs crash. This shouldn't be the case: In cases when one timeframe is empty, an empty AOD should be produced instead.