OpenCon2016 Day 0

I’m attending OpenCon2016 (#OpenCon) in Washington, DC this weekend; this is a meeting of people from around the world who are improving access to data, software, and educational materials for better science, research, and education.

Tonight was an informal reception at the conference hotel. I picked up my badge from Nicole Allen and Brady Yano, and immediately got into conversation with Cody Taylor from the University of Oklahoma Libraries, where he is an Emerging Technologies Librarian. His work includes working with professors to make textbooks open access to reduce costs. Kristofferson Culmer, a PhD student in computer science at the University of Missouri and also President of the National Association of Graduate-Professional Students (NAGPS) joined us. This is Kristofferson’s third OpenCon; as Cody and I are both OpenCon newbies, our conversation included a bit of Q&A about the conference itself.

After leaving Cody and Kristofferson, I walked over to a small group of people that included Erin McKiernan, a physics professor at the National Autonomous University of Mexico and whose tweets frequently appear in my Twitter feed, and my target for walking in that direction, Peter Murray-Rust, a chemist at the University of Cambridge. Peter and I had exchanged tweets when we were both in Portland at Force2016 this past April — not that I remembered that initially — and I’ve been following him ever since. I had a reason for seeking him out tonight, as in my earlier conversation, I learned that Peter has a content mining tools that might help the ASCL: ContentMine. Yes! I cannot wait to start using it! We also talked about the state of data representation in chemistry and astrophysics and how closed chemistry is, with theses embargoed in some universities for five (!!) years. (Wow!)

I had a great conversation next with Thomas Hervé Mboa Nkoudou; he is studying Scholarly Communication at Université Laval and is from Cameroon, where he started a school. He’s very interested in science and in promoting research done in Africa and Haiti, and in the maker movement in education. We also talked about astronomy education. After that, I met Daniel Himmelstein from the University of Pennsylvania and Vinodh Ilangovan from the Max Planck Institute for Biophysical Chemistry, with whom I lamented my apparent inability to find Jon Tennant, the one person I actually expected to meet tonight!

This is a conference of mostly young people — grad students and early career — who have amazing (quite daunting!) accomplishments and are working hard to make knowledge more accessible. I am really looking forward to the next two days! I’m participating, too, with a two-minute presentation on Saturday on the ASCL, and an unconference session on Sunday on encouraging software release and citation.

October 2016 additions to the ASCL

Sixteen codes were added to the ASCL in October 2016:

BurnMan: Lower mantle mineral physics toolkit
BXA: Bayesian X-ray Analysis
C3: Command-line Catalogue Crossmatch for modern astronomical surveys
CERES: Collection of Extraction Routines for Echelle Spectra

cluster-in-a-box: Statistical model of sub-millimeter emission from embedded protostellar clusters
DSDEPROJ: Direct Spectral Deprojection
Fourierdimredn: Fourier dimensionality reduction model for interferometric imaging
Freddi: Fast Rise Exponential Decay accretion Disk model Implementation

gatspy: General tools for Astronomical Time Series in Python
GSGS: In-Focus Phase Retrieval Using Non-Redundant Mask Data
MC3: Multi-core Markov-chain Monte Carlo code
MUSE-DRP: MUSE Data Reduction Pipeline

NuPyCEE: NuGrid Python Chemical Evolution Environment
Piccard: Pulsar timing data analysis package
PyMC3: Python probabilistic programming framework
velbin: radial velocity corrected for binary orbital motions

ADASS BoF: Implementing Ideas for Improving Software Citation and Credit

On Tuesday at ADASS, ASCL Advisory Committee Chair Peter Teuben led a Birds of a Feather session intended as a working session to have people put some of the ideas for improving software citation and credit into practice.

ADS now has a doc type called software

Slide from Peter’s opening presentation

He opened the session with a few remarks about last year’s BoF, similar efforts elsewhere, and examples of progress since last year. Yes, there has been progress! He then showed a list of actionable items and asked people to work on them, adding their work to a common Google doc. His slides are here.

And they did! It was the quietest BoF ever, I believe, as Keith Shortridge, Bruce Berriman, and Jessica Mink wrote about their experiences in releasing software; Renato Callado Borges and Greg Sleap provided guidance on the types of software contributions that add value to science; Alberto Accomazzi, Nuria Lorente, and Kai Polsterer listed ways one can publish and take credit for software; Peter Teuben, Steven Crawford, and possibly others pulled together a list of organization web pages about software created at the institutions, this as a way to highlight and recognize scientific software contributions; Maurizio Tomasi added a suggestion for gathering licensing information; and Thomas Robitaille, Ole Streicher, Tim Jenness, Kimberly DuPrie, and I discussed exactly what should be in the “Preferred citation field” of the ASCL and various people listed about a dozen preferred citations to be added to the ASCL and others used the Suggest a change or addition link for several software packages to provide preferred citation information.

Though Peter had asked that people work for about 30 minutes, he monitored contributions to the Google doc and saw work was still being done so did not call us back together until only 15 minutes or so were left in the session. Instead of having people report back on what they had done as originally plan, he asked for other feedback instead, as progress made was evident in the shared document, and after a bit of discussion on licensing and a few other comments, closed the session.

Though the session is over, the next phase is to put this information to use or disseminate it in some way so it can do some good and be the changes we want to see for software!

 

ADASS XXVI poster: Decoupling the Archive

Decoupling the archive posterThe James Webb Space Telescope (JWST) archive will store numerous metadata for the various files that it contains: at the time of this writing a single FITS file can have up to 250 different metadata fields in the archive, most of which map to keywords in the primary header or header extensions. One of the goals of the archive design is to allow for changes to the fields stored in the database without having to change the ingest code. We have found this to be very helpful during the code development phase of the mission when the FITS file definitions are frequently changing. We also anticipate it will be advantageous during the lifetime of the mission as changes to processing will likely result in changes to the keywords but should not require changes to the ingest code. This poster describes the methods we use to decouple the archive from the ingest process.

Kimberly DuPrie, Space Telescope Science Institute
Lisa Gardner, Space Telescope Science Institute
Michael Gough, Space Telescope Science Institute
Richard C. Kidwell Jr., Space Telescope Science Institute

Birds of a Feather session at ADASS on software citation and credit

The Implementing Ideas for Improving Software Citation and Credit BoF is intended to be a working session to put ideas already generated into action! Everyone in the community has a role in improving it. We have listed a lot of ideas in the previous post about this BoF, have slides online here, and a Google doc to which you can contribute here.

Montage poster at ADASS 2016

We want to share some of the posters that are appearing at ADASS this week (with permission of their authors). Montage is in the ASCL; we love this poster for several reasons, but especially because it makes clear that sustainability of the software is important!
Image of paper on the software Montagle

Abstract: The Montage toolkit is finding exceptional breadth of usage, far beyond its intended application as a mosaic engine for astronomy. New uses include:
– Visualization of complex images with data overlays: e.g. as a re-projection engine integrated into the server-side architecture of a Gbit visualization system supporting investigations of 3D printing with the X3D protocol creation of sky coverage maps for missions and projects bulk creation of sub-images of multiband photometry data creation of plots in the APLPy library.
– Creation of new data products at scale: mosaics of Gemini AO images from the Gemini Multi-Conjugate Adaptive Optics System/Gemini South Adaptive Optics Imager (GEMS/GSAOI) instrument, from the VISTA VIDEO and the UKIDSS DXS surveys welding the Herschel infrared Galactic plane (Hi-GAL) far-infrared Survey into a set of large-scale mosaics, for planetarium shows at a digital as well as for research
– As a re-projection engine to support discovery of 86 Near Earth Asteroids (a U.S. congressional mandate) in the Lincoln Near-Earth Asteroid Research Program (LINEAR).
– Integration into data processing environments: integration of the 4D image cutout tool into the VO-compliant CSIRO ASKAP Science Data Archive (CASDA) as a re-projection engine for the Dark Energy Survey (DES) pipeline.
– Discovery of imaging data at scale: use of memory mapped R-tree indices to support searches for spatially extended data, in use in Spitzer and WISE image searches and in spatial and temporal searches for WISE and KOA.
It has been cited as an exemplar application for development of next generation cyber-infrastructure in 238 papers between 2014 and 2016 to date. What has enabled this broad take-up is that Montage has been built and managed as a scalable toolkit, written in C and portable across all common *nix platforms, with minimal dependencies on third-party software, such that it can be built with a simple “make” command. All the components have proven powerful general-purpose tools in their own right, even those first developed to support mosaic creation, such as discovery of images for input to the engine and for management of mosaics. We describe how Montage is managed to assure that the benefits of the architecture are retained, and how we ensure that new development is driven by the needs of the community.

ASCL poster at ADASS XXVI

ASCL poster for ADASS XXVIThe Astrophysics Source Code Library (ASCL) is a free online registry of codes used in research; it is indexed by ADS and Web of Science and has over 1300 code entries. Its entries are increasingly used to cite software; citations have been at least doubling each year since 2012, and every major astronomy journal accepts citations to the ASCL. Codes in the resource cover all aspects of astrophysics research and many programming languages are represented. In the past year, the ASCL has added dashboards for users and administrators, started minting DOIs for codes it houses, and added metadata fields requested by users. This presentation covers the ASCL’s growth in the past year and the opportunities afforded to it as one of the few domain libraries for science research codes, and will solicit ideas for new features.

Alice Allen, Astrophysics Source Code Library
G. Bruce Berriman, Infrared Processing and Analysis Center, California Institute of Technology
Kimberly DuPrie, Space Telescope Science Institute/ASCL
Jessica Mink, Harvard-Smithsonian Center for Astrophysics
Robert Nemiroff, Michigan Technological University
Thomas Robitaille, Freelance
Judy Schmidt, Astrophysics Source Code Library
Lior Shamir, Lawrence Technological University
Keith Shortridge, Australian Astronomical Observatory
Mark Taylor, University of Bristol
Peter Teuben, Astronomy Department, University of Maryland
John Wallin, Middle Tennessee State University

Download poster

ADASS!

The Astronomical Data Analysis Software and Systems (ADASS) conference starts today in Trieste, Italy with two tutorials, Everything you’ve heard about Agile development is wrong by Simon O’Toole and Multi-dimensional linked data exploration with glue by Thomas Robitaille, and then the opening reception.

The ASCL has two activities at ADASS this year. I’m presenting a poster on the ASCL’s growth over the past year; the poster will be featured in a separate blog post tomorrow. On Tuesday, Peter Teuben, chair of the ASCL Advisory Committee, and other ASCLers here are running a Birds of a Feather session on Implementing Ideas for Improving Software Citation and Credit; this is a follow-up of last year’s session on Improving Software Citation and Credit. This year, we will look at some of the ideas that came out of last year’s session along with ideas generated at the Engineering Academic Software seminar held in June, 2016 and WSSSPE4, held last month at the University of Manchester, and work to implement them. These ideas include:

  • collecting and publishing stories from people who have released their software to share their experience with doing so
  • updating software sites you own to explicit state how your software should be cited
  • emailing code authors who don’t have clear citation instructions on their repos/code sites to suggest they make clear how their codes should be cited
  • adding preferred citation methods to ASCL entries via the “Suggest a change or addition” link or sending them to associate editor Kimberly duPrie or me to add directly
  • looking at your organization’s annual research activity report and suggesting ways it can specifically request software activities
  • writing a document that provides guidance about the types of software contributions that add value to science
  • developing guidelines for recognizing software contributions in hiring and promotion
  • gathering representative research software engineering job descriptions and adding them to the AstroBetter wiki, and suggesting other ways they can be shared for use them as examples

We look forward to an interesting and productive BoF, and an interesting and productive ADASS overall. And gelato, too!

First look at software activities at AAS 229

Though we have a way to go before January’s AAS meeting (and ADASS and OpenCon on the ASCL’s schedule coming up sooner), a look at the schedule for the AAS meeting already shows multiple options for the computationally-inclined astronomer. I’m very excited about the Special Session we’ve organized with the Moore-Sloan DSE, called Perspectives in Research Software. Bruce Berriman (IPAC, Caltech/Astronomy Computing Today) will moderate the session. In keeping with previous sessions, the session will include a discussion period with the floor open for questions and comments; we want to hear what you have to say. We have a panel of seven speakers; the presenters and topics are:

Tracy Teal (Data Carpentry), Software not as a service
Michael Hucka (Caltech), Finding the right wheel when you don’t want to reinvent it
Lior Shamir (LTU), Reproducibility and reusability of scientific software
Ivelina Momcheva (STScI), Funding research software development
Heather Piwowar (ImpactStory), Capturing the impact of software
David W. Hogg (NYU), The relationships between software publications and software systems
And me, Update on research software citation efforts

I hope to see you there!

Other software events that have shown up so far on the AAS schedule are listed below. Good times coming!

Tuesday, 3 January 2017
Workshop: Introduction to Software Carpentry, 8:00 am ‐ 5:30 pm
Workshop: Using Python for Astronomical Data Analysis, 8:00 am ‐ 4:30 pm

Wednesday, 4 January 2017
Splinter Meeting: Flexible Multi‐dimensional Modeling of Complex Data in Astronomy, 9:30 am ‐ 11:30 am

Friday, 6 January 2017
Special Session: Perspectives in Research Software: Education, Funding, Reproducibility, Citation, and Impact, 10:00 am – 11:30 am

Saturday, 7 January 2017
Special Session: Statistical, Mathematical and Computational Methods for Astronomy (ASTRO): SAMSI 2016-17, 10:00 am – 11:30 am
Workshop: Hack Together Day, 10:00 am ‐ 7:00 pm

Also of likely interest is the Special Session on The Value of Astronomical Data and Long Term Preservation that will take place on Thursday, 4 January from 10:00 am – 11:30 am.

 

September 2016 additions to the ASCL

Twenty-five codes were added to the ASCL in September 2016:

21cmSense: Calculating the sensitivity of 21cm experiments to the EoR power spectrum
AdaptiveBin: Adaptive Binning
AIPY: Astronomical Interferometry in PYthon
Askaryan Module: Askaryan electric fields predictor
contbin: Contour binning and accumulative smoothing

CuBANz: Photometric redshift estimator
FISHPACK: Efficient FORTRAN Subprograms for the Solution of Separable Elliptic Partial Differential Equations
FISHPACK90: Efficient FORTRAN Subprograms for the Solution of Separable Elliptic Partial Differential Equations
FIT3D: Fitting optical spectra
GRASP: General-purpose Relativistic Atomic Structure Package

Kranc: Cactus modules from Mathematica equations
NSCool: Neutron star cooling code
Photutils: Photometry tools
PKDGRAV3: Parallel gravity code
PYESSENCE: Generalized Coupled Quintessence Linear Perturbation Python Code

PyPHER: Python-based PSF Homogenization kERnels
SCIMES: Spectral Clustering for Interstellar Molecular Emission Segmentation
SIP: Systematics-Insensitive Periodograms
Sky3D: Time-dependent Hartree-Fock equation solver
spectral-cube: Read and analyze astrophysical spectral data cubes

StarPy: Quenched star formation history parameters of a galaxy using MCMC
SuperBoL: Module for calculating the bolometric luminosities of supernovae
T-PHOT: PSF-matched, prior-based, multiwavelength extragalactic deconfusion photometry
TIDEV: Tidal Evolution package
Weighted EMPCA: Weighted Expectation Maximization Principal Component Analysis