Fast Holographic Deconvolution
FHD is an open-source imaging algorithm for radio interferometers, specifically tested on MWA Phase I, MWA Phase II, PAPER, and HERA. There are three main use-cases for FHD: efficient image deconvolution for general radio astronomy, fast-mode Epoch of Reionization analysis, and simulation. A license for IDL 8.2 or above is required.
For further details, please see Sullivan et al 2012 or Barry et al 2019a.
Download the latest versions of these libraries/repos.
a. This repository
b. The fhdps_utils repository
c. The NASA astro IDL library
d. The coyote library
e. The HEALPix library
f. Optional: the eppsilon power spectrum generation code
Add these libraries to your IDL path. Be sure to use the correct separator (; or : ) and include the ‘+’ sign at the start of +path/to/library/
. Optionally, you can create an IDL startup file to set the IDL path.
a. Windows: !PATH = !PATH + ';' + Expand_Path('+path\to\library\')
b. Unix: !PATH = Expand_Path('+path/to/library/') + ':' + !PATH
Install HEALPix
Install Imagemagick if not already present.
Suggested test commands:
a. print,cgHasImageMagick()
prints error if coyote library not installed, returns 0 if coyote library installed but Imagemagick not installed correctly, and returns 1 if both are installed correctly.
b. astrolib
prints error if astro IDL library not installed correctly, prints message “ASTROLIB: Astronomy Library system variables have been added” if installed correctly.
c. init_healpix
prints error if HEALPix not installed correctly
d. imagefast,randomN(5,256,256),file_path='some/output/path/testimage.png'
prints error if FHD not installed correctly, generates a 256x256 pixel image of poisson noise at path 'some/output/path/testimage.png' if installed correctly and read/write permissions are set correctly.
We are an open-source community that interacts and discusses issues via GitHub. We encourage collaborative development. New users are encouraged to submit issues and pull requests and to create branches for new development and exploration. Comments and suggestions are welcome.
Please cite Sullivan et al 2012 and Barry et al 2019a when publishing data reduction from FHD.
FHD was built by Ian Sullivan and the University of Washington radio astronomy team. Maintainance is a group effort split across University of Washington and Brown University, with contributions from University of Melbourne and Arizona State University.