acts-project / acts-project/acts
Matching data of multiple files in Examples
- 主要語言
- C++
- 星號
- 131
- 分支
- 276
- 平均合併
- 3 天 13 小時
- 30 天內合併 PR
- 112
描述
### Problem statement
Matching data across multiple files produced by the IO writers in `Examples` is not straightforward. As an example, with v44.0.0 (d5d65c794d3676034f37d89e555c131b5b7ad807), an update of the particle barcode has been integrated with #4600. Before v44.0.0, the particle barcodes were represented as a single number (`particle_id`) in the files. While it has been kept as a single number in `Examples/Io/Root/src/RootParticleWriter.cpp` with a hash, this single number became a vector in most of the other files hence making the matching harder than with a single key. In particular, the writing of `particleIds` in `Examples/Io/Root/src/RootTrackStatesWriter.cpp` seems to not work anymore and not even appearing as a key in the .root file.
In other files, like `spacepoints.root` and `measurement-simhit-map.csv` IDs of external objects are used to refer to those but those particular IDs are not found in the corresponding files (`measurement_id` not in `measurements.root` and `hit_id` not in `hits.csv`) while, for what I understood, those IDs correspond to the row number when reading the file, hence not explicit.
As adding IDs to every file might increase the file size, I can understand that a discussion is needed. However I think ensuring that data can be matched efficiently between different files ensure that all information can be retrieved and could be added as an integration test.
貢獻指南
評估
這個 Issue 還沒有評估資料。