ITS: Add second iteration for the seeding vertexer#13306
ITS: Add second iteration for the seeding vertexer#13306shahor02 merged 11 commits intoAliceO2Group:devAliceO2Group/AliceO2:devfrom mconcas:its-second-vertexer-iterationmconcas/AliceO2:its-second-vertexer-iterationCopy head branch name to clipboard
Conversation
|
REQUEST FOR PRODUCTION RELEASES: This will add The following labels are available |
95daea3 to
d8e439a
Compare
d8e439a to
f4ec6a6
Compare
8be9ac2 to
ff2dd46
Compare
|
@shahor02 this should now be the best implementation so far. In the first comment of the PR there is the latest rationale of the PR. |
|
@mconcas thanks, I've reconstructed 1 chunk of LHC23zzh/544124 at 8kHz with the settings of
The increase of low mult. PVs look reasonable and at least there is no loss of K0 (though I don't see an increase either). Unless you protest, I'll squash/merge it as soon as it passes fullCI, then we can have a central test on the grid which UPC people can analyse. |
* Add iteration structure * Overrule size of vector param with specific config value, for pp/PbPb * Propagate CKVals only to the first iteration * Add UPC vertices to the output * MC: assign a single flag to each vertex + purity * Improve logging * Tag vertices with iteration * Add DeltaROF skipping to the tracker * Add in-iteration deltaRof filtering * Add 4th iteration for UPCs * Skip second iteration vertices in first three iterations of the tracking (cherry picked from commit 710d1c8)
* Add iteration structure * Overrule size of vector param with specific config value, for pp/PbPb * Propagate CKVals only to the first iteration * Add UPC vertices to the output * MC: assign a single flag to each vertex + purity * Improve logging * Tag vertices with iteration * Add DeltaROF skipping to the tracker * Add in-iteration deltaRof filtering * Add 4th iteration for UPCs * Skip second iteration vertices in first three iterations of the tracking (cherry picked from commit 710d1c8)
* Add iteration structure * Overrule size of vector param with specific config value, for pp/PbPb * Propagate CKVals only to the first iteration * Add UPC vertices to the output * MC: assign a single flag to each vertex + purity * Improve logging * Tag vertices with iteration * Add DeltaROF skipping to the tracker * Add in-iteration deltaRof filtering * Add 4th iteration for UPCs * Skip second iteration vertices in first three iterations of the tracking (cherry picked from commit 710d1c8)

ITSVertexerParam.nIterations=2: performs the second iteration for the vertexer with hard-coded UPC configurations.ITSCATrackerParam.doUPCIteration=true: performs a 4th iteration on the ROFs tagged by containing tagged vertices.EDIT:
The tracking completely ignores UPC vertices in its first three iterations but will not ignore any ROF (so that clusters can be considered also from ones that are without a vertex or do have an UPC).
Conversely, the fourth iteration will ignore all the rofs that do not contain a UPC vertex.
This will maintain consistency when comparing "hadronic" reconstruction with and without UPC settings.