aNKflag

An intelligent radio frequency interference (RFI) removal tool to work in multi-dimensions radio interferometric data.

Background

Radio interferometric observations are generally affected by terrestrial radio emission, known as radio frequency interference (RFI). aNKflag is an intelligent tool developed to detect and remove these RFIs both in time-frequency as well as in the Fourier domain, popularly known as uv-domain in radio interferometry.

• aNKflag works on UVFITS files
• Before flagging one needs to convert CASA measurement set to UVFITS
• After flagging one needs to convert UVFITS to CASA measurement set and copy the flags to original measurement set.
• These features are not provided in aNKflag, as these are readily available in CASA.
• This python version uses precompiled and containersed sourcecode of aNKflag, so no need to worry about installing C/C++ libraries.

Documentation

aNKflag documentation is available at: ankflag.readthedocs.io

Quickstart

aNKflag is distributed on PyPI. To use it:

1. Create conda environment with python 3

   conda create -n ankflag_env python=3.10
   conda activate ankflag_env
   

2. Install aNKflag in conda environment

   pip install ankflag
   

3. Initiate necessary metadata and containers

   run-ankflag init --datadir </full/path/to/data/directory>
   

   Contaniers will be stored in the data directory.

4. Run aNKflag

   run-ankflag run </full/path/to/input/uvfits> </full/path/to/output/uvfits> --scratchdir </full/path/to/ankflag/workdir> --flagmode <uvbin/baseline> --npol <num_of_polarisation> --nthreads <num_of_cpu_threads> --target_type <target_type>
   

That's all. You run aNKflag for analysing flagging RFI. It will create a UVFITS file with output file location 🎉.

Acknowledgements

aNKflag is developed by Apurba Bera (ASTRON, NL) and Devojyoti Kansabanik (IAA-CSIC, Spain). If you use aNKflag for analysing your work, include the following statement in your paper

RFI flagging is perfomed using aNKflag.

1. Cite aNKflag software in zenodo: https://doi.org/10.5281/zenodo.20568784

and cite the following papers.

2. aNKflag paper: Kansabanik et al., ApJS 2023

License

This project is licensed under the MIT License.