seclark / rht

The Rolling Hough Transform

The Rolling Hough Transform (RHT)

This is the Rolling Hough Transform, described in Clark, Peek, & Putman 2014, ApJ 789, 82 (arXiv:1312.1338). If use of the RHT results in a publication, please cite this work.

For instructions on install, use, and more, please see http://seclark.github.io/RHT/.

The RHT is written and maintained by Susan E Clark. Please feel free to get in touch with questions or submit pull requests with improvements.

Update 2/2021, S.E. Clark: I recently pushed convRHT.py. This is a version of the RHT with the core architecture rewritten to be based on a series of convolutions. It is much faster than the original RHT. This should eventually be integrated into the main code base, but I am releasing it for now so that more people can use it. If you use convRHT in a publication, please cite the main RHT paper as usual, as well as BICEP/Keck Collaboration et al. 2022 (arXiv:2210.05684), corresponding author George Halal, where the convRHT is first introduced. As always, feel free to reach out with comments and questions, or submit pull requests. Many thanks to convRHT beta-testers G. Halal and Y.K. Ma.

Update 8/2023, S.E. Clark: For a version of the RHT that runs directly on the sphere (HealPix), see Halal, Clark et al. 2023.

rht's Issues

Multiprocessing

If it possible to run it using multiprocessing? Only one cpu will be used in current version.

array input rather than filename

Hi,

I was wondering if you had any inclination to add the capability of using an array as input to rht.main() rather than a filename. For me that'd be useful because I have an image that I want to crop before processing. I might also want to do some pre-processing to the data (e.g. a small bit of smoothing to reduce the noisiness). Of course I could write the data out to a fits file and then use that, but that would be a bit less convenient and would produce an unnecessary file.
I might do that myself, and of course would share it with you if you'd like.

Jon

Header of the backprojection

Pass information from the header of the input file (e.g. positional information with CRVAL1, CRPIX1, CDELT1,...) to the header of the backprojection in the output FITS file.

Global parameter OUTPUT not used?

The global parameter OUTPUT, from which you can define an output directory, doesn't do anything. I'm decently sure it's not been implemented, but if it is, it does not work for me.

Force tag not working?

The force argument seems to be not working. I have set it to True in main() and rht(). It still skips doing the RHT. Is this not functional yet?

Documenting the second object from the output FITS file

I did not find any documentation regarding the way in which the X-Y-Theta power array is stored. What is the data format for the second object from the output FITS file?

For instance, in the function RHT_tools.get_RHT_data, what do ipoints, jpoints and hthets represent?