Inner mpgd barrel geo update

[ybedfer]

Briefly, what does this PR introduce?

Update the MPGD inner barrel, making it a set of cylindrical, as opposed to flat, tiles.

What kind of change does this PR introduce?

• New feature (issue Geometry Matrix Changes #730 )

Please check if this PR fulfills the following:

• Changes have been communicated to collaborators

Does this PR introduce breaking changes? What changes might users need to make to their code?

Yes. Users need to update their DD4hep private repository to revision #c3b305ce of the AIDASoft/DD4hep github, see "AIDASoft/DD4hep@c3b305ce".

Does this PR change default behavior?

Yes.

[ybedfer]

Plots showing the distributions of:

• Simulated Hits,
• Reconstructed Hits,
• Residuals (Rec-Sim).

Setup = ddsim with a gun of µ^- in the range 10 GeV, 15 < theta < 165 deg.

cCyMBaLSim.pdf
cCyMBaLRec.pdf
cCyMBaLRes.pdf

Remarks:

• The different colors correspond to sub-structures: red/blue = tiles along φ, else 4 sectors along Z; black is the overall.
• On the Simulated plot, some stray hits show up in the Rr (radius) distribution. This is non physical, but seems to be a common feature also present in the Outer MPGD detector, hence not connected to this PR.
• When going from Sim to Rec, there's some inefficiency. Feature also seen in all barrel trackers.
• Std Dev of residuals are as expected for a 150 $\times\sqrt{12}$ µm grid size in Z and 1~= $\sqrt{12}\times$.0.29 mrd in φ.

eicrecon

Executed on the ddsim data mentioned supra, with debugging printout ( -Ptracking:CentralTrackerMeasurements:LogLevel=trace), hits from CyMBaL are seen to be associated to tracks.

[wdconinc]

FYI, https://eicweb.phy.anl.gov/containers/eic_container/-/merge_requests/963 for inclusion of AIDASoft/DD4hep#1294.

[wdconinc]

This looks good. Thanks! I added some comments. I will wait to approve until later this when the pipelines should be able to run.

[wdconinc]

Updated branch to restart CI pipelines with updated DD4hep installation.

[ybedfer]

Hello,
What could I do concerning the failing check "eicweb/reconstruction_benchmark"? I can see that some (I haven't checked them all) jobs are crashing w/:

#5  0x00007fb9ab939b58 in draw_surfaces(std::shared_ptr<Acts::TrackingGeometry const>, Acts::ContextType, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) () from /opt/local/lib/libJugTrackPlugins.so

Does this mean that it's connected w/ the specification of the sensitive surface? or the <layer_material/> which I modified lately?
The benchmarks seem to be available on github ("[email protected]:EIC/benchmarks/reconstruction_benchmarks.git"). "https://eicweb.phy.anl.gov/EIC/benchmarks". Can I run them, with a limited amount of training and effort?

[wdconinc]

I noticed the failing reconstruction_benchmarks check yesterday and looked into it. The issue is in legacy code, and stems from an incomplete list of supported surfaces in our export to a 3D graphics file for visualization and debugging. We have fixed the equivalent code in our up-to-date code, and yesterday I propagated a fix to the legacy code as well. It takes a day or so for this to propagate to the containers in which we run the checks, so I suspect this will be resolved by Monday at the latest.

[ybedfer]

Hello,
There is no progress.
Is it because something went wrong in your fix this business of "incomplete list of supported surfaces"? Or is it simply that the job has not been restarted?

[wdconinc]

I re-triggered the job and it now succeeded. All checks now pass.