The benchmark aims to verify that the optics used in the dd4hep epic simulation match those expected by the eic machine. Proposed components of the finalized benchmark: - [x] Plots showing the position, momentum and correlations for beam electrons passing through the beamline magnets. https://github.com/eic/detector_benchmarks/pull/154 - [ ] Link directly to machine distributions. - [ ] Generate events using the afterburner. - [x] Energy-theta acceptance at each virtual detector. #167 - [ ] Create check and feedback comment to epic PRs which don't pass or significantly change acceptance - [ ] Labelled beamline geometry visualization output. - [ ] Extend to include far-forward protons - [ ] Create report pdf, tidying up all plots. - [x] Train onnx neural network to reconstruct initial particle and produce resolutions. #171 - [x] Create pipeline to automatically update the onnx network used in the reconstruction. #201
The benchmark aims to verify that the optics used in the dd4hep epic simulation match those expected by the eic machine.
Proposed components of the finalized benchmark:
Beamline acceptance benchmark #167
Beamline neural network reconstruction training #171
Close the loop on retraining Low-Q2 onnx models #201