Matt Craig, Professor of Physics and Astronomy at Minnesota State University Moorhead, has created this list of Astropy affiliated packages to help improve your experience exploring astronomy using Python. This post was inspired by Ole Moeller-Nilsson’s recent blog post on how to get started doing astronomy with open data and Python.
—
One of the things I both love and hate about Python is that there is almost always more than one way to get a job done. When I first switched to Python I found that really frustrating. Of the three (or four or five or six) astronomy packages, which one was “the” package to use? Or was I supposed to use the more fundamental tools, like Numpy or SciPy? Now that I’ve been using Python for a few years, I’ve come to appreciate the diversity of tools.
I was reminded of this by a recent NumFOCUS blog post about doing astronomy with Python and open data which outlined how to use several fundamental packages to do image reduction and photometry. Here I’ll give an overview of several high-level packages for doing astronomy that are Astropy affiliated packages; these are packages that follow the same coding, documentation and testing conventions as the core astropy package. The full list of affiliated packages is on the Astropy project’s website.
I’ll give a quick overview of the packages and then wrap up with some installation instructions. I’ve only mentioned each package once in the description below, even though several really cross categories.
Some good news/bad news before we start: there are almost thirty affiliated packages, and there are many very good astronomy-related packages that are not affiliated packages. The menu of interesting Python packages only seems to increase.
Optical/IR image processing
Reduced, reprojected, stacked image with overlay of sources detected by photutils (gray circles) and two catalogs from Vizier. Courtesy of ccdproc, astropscrappy, reproject, photutils, and astroquery.
More specialized, but in the same genre, is python-cpl, a python interface to recipes of the ESO data reduction pipeline.
Once you’ve calibrated your images, photutils can be used to do source detection and aperture and PSF photometry. If you need to stack those images first, try reproject which can align them for you if the images have WCS headers.
You probably want to look at all of those nice images you are working with, so up next is…
Image viewers and plotting helpers
ginga provides by a “reference” viewer and the libraries for making your own image viewer or for customizing the reference viewer. If you use IRAF you should check out imexam, which duplicates the functionality of the IRAF tool of the same name.
Coming soon to a notebook near you: ginga-based image viewer widget…
APLpy (pronounced “apple pie”) makes it easy to prepare publication-quality plots of images with lots of options for annotating them or overlaying additional information. Finally, montage-wrapper provides a python interface to the Montage image mosaic system (Montage itself is developed and maintained by IPAC and is not an affiliated package).
There are two packages for interacting with regions on the sky: pyregion, which handles ds9 and ciao region files and helps you overlay them on matplotlib images, and the more general regions, which is still in the very early stages of development. Rounding out this category is spherical-goemetry for representing spherical polygons on the sky.
Glue
Though it can display images, glueviz is really a much more general tool. It allows you create links between plots so that you can, say, select stars on part of a color-magnitude diagram and see those points highlighted on an image of the region containing those stars.
Optical spectroscopy
The specutils package provides tools for working with 1-D spectra; it is undergoing rapid development, described in detail in a proposal for the future of spectroscopy in astropy, but currently provides tools for reading and writing spectra in a variety of formats, and for performing common operations. A couple of packages in the early stages of developments (they are not affiliated packages yet) are specreduce, for reducing spectra, and specviz for viewing those spectra.
omnifit performs spectroscopic fitting of interstellar ices.
Finally, gwcs is being developed both to handle celestial for the James Webb Space Telescope and to provide a framework for spectral coordinates. It provides a framework for carrying out a sequence of transforms from one set of coordinates to another.
Theory
As a former theorist myself, it is exciting to see so many theory and simulation packages. If you want to do galactic dynamics in python, galpy is for you; it includes a wide variety of potentials and sampling of distribution functions.
asro-gala, which also does gravitational dynamics, contains several general-purpose integrators and tools for carrying out several coordinate and velocity transforms in our galaxy.
cluster-lensing computes galaxy cluster halo properties and weak lensing profiles.
If a single cluster isn’t enough for you, use halotools to construct your own mock universe. Given a set of dark matter halos from a simulation, it can populate those halos with galaxies and make measurements on them.
Crossover appeal
A few packages don’t fit neatly into a single category.
One is astroML which contains a variety of astronomy-related machine learning tools, and even has an extensive user’s manual on Amazon.
Another is sncosmo, which makes supernova light curve models, including fitting photmetric data to a model. There are a variety of models and passbands to choose from. This one has been extremely useful to an undergraduate I’ve been working with this summer to (finally) reduce and analyze the data we have on SN 2011fe.
Those coming from IDL may be interested in PyDL, which contains Python implementations of some widely-used astronomical routines from IDL, with an emphasis on access to Sloan Digital Sky Survey (SDSS) data, and tools from the IDL Astronomy User’s Library, also known as the “Goddard Library”.
Spectral extremes
The long and short ends of the electromagnetic spectrum get some special attention.
On the high energy side are gammapy, providing a quite complete set of tools for both simulation and analysis of gamma ray data. At a slightly lower energy, maltpynt has some general-purpose tools for x-ray astronomy, but its primary purpose is for timing analysis of NuSTAR data. Finally, naima calculates non-thermal spectra of relativistic particle distributions.
Radio astronomers should check out spectral-cube, which supports working with data cubes with two spatial and one spectral dimension, including tools for masking, extracting subarrrays, and working with data sets too large for memory.
Planning and data access
Need some visibility or airmass plots? Or maybe some finder charts? Take a look at astroplan, which does those things with minimal effort.
With pyvo, everything available through the Virtual Observatory is available via python.
The hips Python package lets you fetch HiPS image tiles from astronomical data centers for any astronomical survey and sky region you like. It then stitches the tiles together and reprojects them into a sky image that you can save to FITS, PNG or JPEG, or analyse directly from Python.
Which reminds me, there is…
One more thing…
I’ve saved one of the best for last. No matter what your area of interest is in astronomy you need access to online databases, and astroquery has you covered. Simbad lookups? Yep. Data from a Vizier table? Can do. Gaia catalog? Sure! ALMA? No problem. Take a look at the gallery of examples.
Installation
Each of the packages above is on the astropy conda channel and on PyPI, the Python package index. Navigating to the package on the conda channel will get you to specific installation instructions, but they boil down to conda install -c astropy package_name. With pip (i.e. if you do not use conda), install with pip install package_name. I’ve tried to make sure the package names above match the names on PyPI, i.e. the name given in the text is the name you use to install the package with pip.
Support Astropy by becoming a NumFOCUS member or making a donation to the project.
