Nicholas Barnett

Advisor: Austin Alan Baty

Mentor: Philippe Canal

Undergraduate: California State University, Chico (Physics and Chemistry (minor in Mathematics))

Graduate: University of Illinois Chicago (Physics)

Project: Lossy Compression of RNTuple

One of the major bottlenecks of data collection at the LHC is size of incoming data and data storage. Compression algorithms are a practical way of starting to address this issue, but historically lossless algorithms have been used. When the High-Luminosity LHC is operational the data size will significantly increase, making compression algorithms even more critical for data collection and storage. Investigating, identifying, and understanding lossy compression algorithms for use within the ROOT infrastructure is the aim of this project.

This Investigation includes the use of lossy compression algorithms in TTree and RNTuple data types using augmented ROOT frameworks. This is intended for general use by physics analyzers, and potentially during data taking. Determining ideal stages to implement lossy compression algorithms as well as quantifying the loss occurring during compression for different categories of physics observables are key aspects of this study.