CTLearn: Deep Learning for IACT Event Reconstruction

CTLearn is a package under active development to run deep learning models to analyze data from all major current and future arrays of imaging atmospheric Cherenkov telescopes (IACTs). CTLearn can load R1/DL0/DL1 data from CTAO (Cherenkov Telescope Array Observatory), FACT, H.E.S.S., LST-1, MAGIC, and VERITAS telescopes reduced by ctapipe and processed by DL1DataHandler.

• Code, feature requests, bug reports, pull requests: https://github.com/ctlearn-project/ctlearn
• Documentation: https://ctlearn.readthedocs.io
• License: BSD-3

Installation for users

Download and install Anaconda, or, for a minimal installation, Miniconda.

The following command will set up a conda virtual environment, add the necessary package channels, and install CTLearn specified version and its dependencies:

CTLEARN_VER=0.10.0
wget https://raw.githubusercontent.com/ctlearn-project/ctlearn/v$CTLEARN_VER/environment.yml
conda env create -n [ENVIRONMENT_NAME] -f environment.yml
conda activate [ENVIRONMENT_NAME]
pip install ctlearn==$CTLEARN_VER
ctlearn -h

This should automatically install all dependencies (NOTE: this may take some time, as by default MKL is included as a dependency of NumPy and it is very large).

For working on the IT-cluster, please do not forget to update your LD_LIBRARY_PATH to include the necessary paths. For example, you can add the following line to your :bash:.bashrc`` file: :bash:`export LD_LIBRARY_PATH=/to/your/.conda/envs/ctlearn/lib:/fefs/aswg/workspace/tjark.miener/cudnn-linux-x86_64-8.9.2.26_cuda11-archive/lib:/fefs/aswg/workspace/tjark.miener/cudnn-linux-x86_64-8.9.7.29_cuda12-archive/lib:$LD_LIBRARY_PATH` Note: You would need to replace the /to/your/.conda/envs/ctlearn/lib with the path to your conda environment where ctlearn is installed. cudnn-linux-x86_64-8.9.2.26_cuda11-archive and cudnn-linux-x86_64-8.9.7.29_cuda12-archive are the paths to the cuDNN libraries for CUDA 11 and CUDA 12, respectively.

See the documentation for further information like installation instructions for developers, package usage, and dependencies among other topics.

Citing this software

Please cite the corresponding version using the DOIs below if this software package is used to produce results for any publication:

• 0.8.0 :

Team

Ari Brill	Bryan Kim	Tjark Miener	Daniel Nieto

Collaborators

Qi Feng	Ruben Lopez-Coto

Alumni

Jaime Sevilla	Héctor Rueda	Juan Redondo Pizarro	Luca Romanato	Sahil Yadav	Sergio García Heredia