CELEBI is an automated data processing pipeline for producing sub-arcsecond precision localisations and high-time resolution polarimetric measurements of fast radio bursts (FRBs) from voltages obtained with the Australian Square Kilometre Array Pathfinder (ASKAP).

CELEBI is a nextflow based pipeline which requires a config script to process an FRB. For each FRB, a config file needs to be created to store all the CELEBI parameters. On Ozstar/NT supercomputer in oz313 group we can easily create an config file using the writeconfig.sh script.

writeconfig.sh 220610

Where 220610 is the FRB number, e.g. 220610 for FRB 220610. This will create a config file called 220610.config that holds default CELEBI parameter values. A complete list of config parameters can be found here, the most important parameters will be explicitly defined in your newly made config file.

To use CELEBI, first make sure nextflow is installed. On Ozstar/NT you can simply ml nextflow. You can start a CELEBI run by envoking the following command:

nextflow <CELEBI dir>/pipelines/main.nf --makeimage --beamform -c <config dir>/220610.config -w <work dir>/220610 > frb220610.out

main.nf is the main nextflow script
--makeimage --beamform are --options for what kind of processing you want to do on the FRB
-c specifies what config file you want to use, in this case 220610.config
-w specifies where you want to store all the files that are created during processing
> frb220610.out specifies to log everything to a text file that can be easily read for debugging.

CELEBI can perform the following processes:

1. --makeimage Localise FRB to sub-arcsecond precision.
2. --beamform Beamform Antenna voltages and coherently dedisperse to obtain High time resolution (HTR) polarimetry (Requirements: fluxcal solutions, delays, position (JMFIT) file)
3. --shrine Structure maximisation to optimise Dispersion Measure (DM) (Requirements: HTR outputs)
4. --mfimage Matched filter imaging to optimise localisation precision (Requirements: HTR outputs, frb image outputs)

In addition, there are a handful of hidden options that can also be run:

5. --locfrb Localise FRB without processing fluxcal or polcal data (Requirements: fluxcal solutions, delays)
6. --htrfrb Only do FRB beamforming (Requirements: Everything else!!)
7. --corrcal Correlate and process only fluxcal to get solutions
8. --impcal Only image polcal (Requirements: fluxcal solutions)
9. --beampcal Only beamform polcal data (Requirements: fluxcal solutions, delays, polcal position file)

CELEBI is a modular pipeline. We can specify what main processes we want to perform as outline above. For example we can just perform arc-second localisation by just specifying --makeimage. Or with the advent of CRACO FRBs with improved precision, we can go straight to beamforming by just specifing --beamform. We can perform these processes in any order! The full list of CELEBI options can be found here.

In CELEBI, an output directory is specified to store all the processed FRB data, files and plots. on Ozstar/NT in oz313 the output directories exist in /fred/oz313/processing/output. The following image shows the contents of a typical FRB output directory.

*_summary.txt - is a summary file that compiles all the important information about the FRB into a single .txt document
binconfigs- stores all the files used for correlation
field - stores all the data/files created when imaging the field of the FRB
finder - stores all the data/files created when initially imaging the FRB
fluxcal - stores all the data/files created when processing the flux calibrator
htr - stores all the data/files created when beamforming the data, such as HTR data
loadfits - stores all the fits files
mf - stores all the data/files created when applying a matched filter to the FRB image
polcal - stores all the data/files created when processing and applying the polarisation calibrator
position - stores all the data/files created when solving for and applying RACS astrometric corrections
shrine - stores all the data/files created when performing structure maximisation on the dispersion measure.

You can dive deeper into the FRB data products using ILEX: https://github.com/tdial2000/ILEX