Home

Welcome to the InstaRec wiki!

The aim of this repository is to provide software for the CLAS12 online reconstruction framework: InstaRec.

The software is written in Java, and compiled with Maven. The top level of the repository is organised as follows:

• pom.xml - file used to compile the code
• documents/ - written resources such as CHEP proceedings or publications.
• etc/networks - neural networks used in InstaRec
• src - source code
• plots - Empty directory, present by default for the output of validation tools (do not commit your plots)
• training_data - Empty directory, present by default for the output of training data providers (do not commit your data)

The source code found in src/main/java/j4np/instarec is organised as follows:

• workers - code used to the process data in HIPO or evio format. These implement InstaRec.
• data - data providers used to parse HIPO files produced by the workers into training data.
• networks - code used to train the AI/ML algorithms in InstaRec.
• validation - code used to validate the online reconstruction.