Summary and Schedule
Benchmarks are scripts that run detector simulations and analyze the resulting data to extract quantities related to the detector performance.
Prerequisites
The following tutorial assumes basic knowledge of shell. It may be practical to use eic-shell for benchmark development, as it is the environment that will closely match the one during benchmark execution. It also assumes that you are a member of the EIC organization on GitHub and belong to the “epic-devs” team. You also need to have your local ssh key added to GitHub so that you can push.
This tutorial follows other tutorials in the EIC series:
Further information is included in other tutorials:
| Setup Instructions | Download files required for the lesson | |
| Duration: 00h 00m | 1. Exercise 1: Analysis Scripts and Snakemake | How does one set up data analysis workflows? |
| Duration: 00h 30m | 2. Exercise 2: Setting up your first benchmark with pipelines | How do we create a new pipeline with GitLab CI? |
| Duration: 01h 00m | 3. Exercise 3: Filling out your benchmark | How do we fill in each stage of the benchmark pipeline? |
| Duration: 01h 30m | 4. Exercise 4: Adding a Status Flag | How can your benchmark indicate that there were detrimental changes to software or detector design? |
| Duration: 01h 50m | 5. Exercise 5: Making Useful Figures | How does one make useful benchmark figures? |
| Duration: 02h 10m | Finish |
The actual schedule may vary slightly depending on the topics and exercises chosen by the instructor.
Tutorial participants will need
- a working eic-shell installation
- if you’ll be following along on a remote cluster through ssh, you may also need to set up ssh-keys. This will allow you to perform GitHub operations while using ssh.
This tutorial shows analyzers how to turn their analysis script into a benchmark. To do this, we will start with a ready-made bare-bones analysis. This is modeled after the u-channel rho benchmark, but is stripped down to just a few crucial parts.
To start, download the following
- this analysis script:
uchannelrho.cxx - this plotting macro:
plot_rho_physics_benchmark.C - this style header:
RiceStyle.h
We will also start by running over a file from the simulation campaign. Download it to your workspace: