Helper scripts and configuration files for the CLAS12 RG-E Lambda analysis at Jefferson Lab.
This repository hosts utilities and configuration files used to automate and manage batch processing of CLAS12 data for the RG-E Lambda analysis.
Jobs are managed via SWIF2, and scripts generate JSON workflow definitions for import into the JLab computing farm.
rge_lambda_utils/
│
├── scripts/ # Shell scripts for workflow generation and utilities
│ ├── generate_batch_jobs.sh
│ ├── ...
│
├── yaml/ # Configuration files for recon-util
│ ├── RGE_Lambda.yaml
│ ├── ...
│
├── examples/ # Optional: example JSON files or sample outputs
│
├── README.md # This file
└── .gitignore # Ignored output files
- Jefferson Lab environment with CLAS12 software stack:
module load clas12
- SWIF2 workflow manager (/usr/bin/swif2)
- Access to the JLab /volatile and /work storage areas.
- Edit paths in scripts/generate_batch_jobs.sh if needed:
- INPUT_BASE_DIR – base directory with input .hipo files
- OUTPUT_BASE_DIR, SKIMMED_DIR, etc. – where outputs will be written
- YAML_FILE – path to your RGE_Lambda.yaml configuration
- Run the script to generate a SWIF2 JSON workflow:
cd scripts
bash generate_batch_jobs.sh- Import and submit the workflow to the JLab farm:
swif2 import -file <file_name>.json
swif2 run <workflow-name>- Monitor jobs:
swif2 status <workflow-name> -jobs- Each folder in the input directory generates four sequential jobs:
- s1_ep_* – initial electron selection (independent)
- s2_skim_* – skim relevant particles
- s3_recon_* – run recon-util
- s4_reduce_lambda_* – extract Lambda decay events
- Steps 2–4 are chained via SWIF2 antecedents, ensuring they run in order.
- Update the YAML configuration in yaml/ to tune reconstruction settings for different runs or cuts.
- Any other shell scripts that will be useful as utilities for Lambda analysis.
- Further optimizations to make this run faster on the farm.