Speaker
Description
Many data-intensive experiments, need to track large amounts of data, large code bases to process the data, and configuration data (or metadata). All of these datasets have unique requirements for version control. Furthermore, these datasets must have rules and requirements about how they can interact with the code. For this it is crucial to make the storage of data, the metadata in particular as flexible as possible so future versions of the metadata are not restricted by the current requirements of the metadata.
In the XENONnT experiment, we use git/GitHub for code, for data we have strax/straxen, and for metadata we have rframe/xedocs. Here we will focus on our handling of metadata via rframe/xedocs. As an example of this metadata we will focus on detector characterization data. We want this version data to be flexible enough to adapt to future changes, to have insertion rules and to follow time dependace requirements.
Rframe however is not tied down by any of the specific requirements of our experiment and can be used by many others as a framework for their storage of metadata. These frameworks are written in python which is used as the main programing language for the analysis in many experiments.
| Consider for long presentation | No |
|---|
