hpmoc: HEALPix Multi-Order Coordinate Partial Skymaps

HPMOC is an ultra high-performance, cross-platform toolset for working with multi-order coordinate (MOC) HEALPix_ images (i.e. images with multiple pixel resolutions). MOC images are used by the LIGO-Virgo-KAGRA collaboration, the Interplanerary Network and others to represent portions of the sky with variable resolution. By only including pixels in regions of interest, and only then at a resolution appropriate to how they were observed/calculated, it is possible to reduce storage and computation costs by several orders of magnitude.

HPMOC is the only library providing tools for loading partial/whole MOC skymaps (as well as standard HEALPix skymaps), taking spatial intersections, modifying resolution, plotting the skymaps, converting them to and from Astropy WCS projections, performing pointwise math, and generating PSF skymaps from point sources, all using algorithms that minimize memory, computation, and storage costs. It is based off of work on LLAMA, the world's first Gravitational Wave/High-Energy Neutrino low-latency search pipeline, which has been improved and refactored into this separate module.

If you use hpmoc in published research, we ask that you cite Stefan Countryman's thesis. hpmoc is introduced in section 4.5.13.

hpmoc is licensed under the terms of the GNU General Public License, version 2 or later

Installation

hpmoc has only a few dependencies, but they are large numerical/scientific libraries. You should therefore probably create a virtual environment of some sort before installing. The easiest and best way to do this at the moment is to use conda, which should come with an Anaconda distribution of Python:

conda create -n hpmoc
conda activate hpmoc

note that creating a new environment is optional and hpmoc can now be installed similar to any other python package.

With pip

If you just want to use hpmoc and don't need to modify the source code, you can install using pip:

pip install hpmoc

This should install all required dependencies for you.

Developers

If you want to install from source (to try the latest, unreleased version, or to make your own modifications, run tests, etc.), first clone the repository:

git clone https://github.com/markalab/hpmoc.git
cd hpmoc

Make sure the build tool, flit, is installed:

pip install flit

Then install an editable version of hpmoc with flit:

flit install --symlink

As with the pip installation method, this should install all requirements for you. You should now be able to import hpmoc. Note that you'll need to quit your python session (or restart the kernel in Jupyter) and reimport hpmoc before your changes to the source code take effect (which is true for any editable Python installation, FYI).

You can go ahead and run the tests with pytest (which should have been installed automatically by flit):

py.test --doctest-modules --cov=hpmoc