Oral Abstract

Oral Contribution (O11.2)

Theme: Other

BoF Summaries

Oral Abstract

(.1)

Bus leaves for tour of telescopes (Departure Martiniplaza)

https://www.adass2019.nl/social/telescope-tour-page/

Oral Abstract

(11.4)

Theme: Other

Closing remarks

Oral Abstract

(.1)

Welcome reception (Academiegebouw, Broerstraat 5, Groningen)

https://www.adass2019.nl/social/conference-reception-page/

Oral Abstract

(.1)

Dinner at restaurant 'de Rietschans'

https://www.adass2019.nl/social/conference-dinner-page/

Tutorial Abstract

Tutorial (T0)

A Practical Introduction to the Advanced Scientific Data Format (ASDF)

ASDF is a language-neutral file format for serializing scientific data. It is in use by JWST, DKIST and Modelforge and under consideration for other projects. Relative to FITS, it provides a lot of flexibility for storing complex data structures, while retaining the human-readability of metadata. Relative to VOTables, it provides compact transport of binary data. Relative to HDF5, the metadata are in text, it allows for deep levels of nesting, and the overall specification is relatively simple. The structure of the data can be automatically validated using schemas, which can reference other schemas. There is built-in versioning. Native Python data types (numerical types, strings, dicts, lists) are serialized automatically, and it is straightforward to support custom data types. There is a Python library to support reading, writing, and validation, and there is work in progress on a C++ implementation.

Tutorial Abstract

Tutorial (T0)

Test-driven development for Astronomers

Test-driven development (TDD) is a software development practise where the developer writes a unit test for each piece of code before writing the code itself. This follows a simple six step cycle: 1) Write a unit test. 2) Check the test fails. 3) Write the code. 4) Check the test (and all previous tests) pass. 5) Refactor. 6) Repeat. By focusing on writing code to pass tests, the developer immediately specifies the code’s necessary functionality and therefore writes code which performs only the task required. This forces the developer to immediately consider how the code will be used, placing the emphasis on interface over implementation. This ensures simpler, cleaner and more concise code is written, leading to a better designed and fully tested codebase in the long-term. For large-scale collaborative development projects, TDD is a must, as it allows developers to be certain exactly what a piece of code written by someone else was intended to do and therefore make changes to it with confidence.

Tutorial Abstract

Tutorial (T0)

How to get started with Data Science and Machine Learning for Astrophysics

This introductory tutorial walks through how to get started with Data Science and Machine Learning for Astronomy and Astrophysics generally. The tutorial will be run primarily in Python3 and will cover statistics, python, data science process including a 'what is data science' section. The tutorial will optionally include distributed cloud computing and will orient around astrophysical datasets. The tutorial will use data from optical spectroscopic surveys but participants will be able to load in any data they are interested in to complete a machine learning pattern search locally and in the cloud in a distributed fashion. All code will be hosted on github in jupyter notebooks. Instructions will be given for software installation beforehand and all participants will need a computer.

Tutorial Abstract

Tutorial (T0)

Open Science with EGI Jupyter, DataHub, GitHub, Zenodo and Binder

This tutorial displays a practical and realistic implementation of the Open Science cycle, combining services from the European Open Science Cloud (EOSC) and from the broader open source domain. The audience will set up data analysis, execute it on a scalable cloud compute platform, share the analysis application in a permanently identifiable way (with DOI), discover the analysis code and re-execute it (imitating a fellow researcher). The tutorial will combine the following tools: • EGI Notebooks is a scalable Jupyter Notebooks service hosted on the pan-European EGI Cloud provided by academic and commercial providers for science and education. EGI notebooks will be used to setup and execute an astronomy data analysis and visualisation application. • EGI DataHub is a data access platform that can gather distributed datasets from distributed data repositories and stage them to compute sites for data analytics. DataHub will be used as the data gathering layer in Notebooks, collecting analysis input data. • GitHub is a public service on the internet, widely used to version, store and share software code. GitHub will be used as a platform to store versioned Notebooks applications together with the requirements concerning their executability (for example library dependencies) • Zenodo is a scientific data and publication repository by OpenAIRE. Zenodo will be used to generate Digital Object Identifiers (DOIs) with a permanent, sharable link to versioned astronomy Notebook application. • Binder is an online service that accepts Notebook applications and makes them re-executable with 'one-click'. We will use Binder to re-execute Notebook applications referenced by their DOIs.

Tutorial Abstract

Tutorial (T0)

DACE: hands on web platform and Python client API

The Data Analysis Center for Exoplanets (DACE) is a web platform based at the University of Geneva (CH) dedicated to extrasolar planets data visualisation, exchange and analysis. Observational data like radial velocities, light curves and imaging measurements are available as well as sophisticated analysis tools. The tutorial will start with a description of the platform : how to use it step by step with the attendees and overview of different tools especially linked to observations. Then it will focus on python-dace-client API as a live-coding session. The python-dace-client API is a python module that anybody can use to get and manipulate data from DACE. It allows you to get radial velocities data, light curves data and to download fits files.

Tutorial Abstract

Tutorial (T0)

Turning pipelines into portable pipelines

Astronomy software tools and pipelines are generally developed with functionality over portability in mind. This tends to result in tools which are hard to compile and pipelines that are hard to tweak by users. One solution to the problem of software compilation in the broad open source community is to provide the software and its dependencies as a container or in binary packages. For the flexibility of pipelines, many solutions exist, among which the Common Workflow Language (CWL) standard. Using the CWL standard separates the concerns of the application and pipeline developers while being independent of a specific pipeline execution framework. Because of this independence, using the CWL makes it easier to scale the processing up from a personal laptop to clouds, private clusters or supercomputers. In this tutorial we will focus on the usage of the CWL standard to create pipelines out of containerised software tools (using packaged from the KERN suite).

Tutorial Abstract

Tutorial (T0)

World-wide Single Sign-On and Group Management for Astronomy Projects using CILogon and eduGAIN

CILogon enables researchers to log on to cyberinfrastructure (CI) using credentials from their home university, college, or organization, and provides an integrated open source identity and access management platform for research collaborations that combines federated single sign-on (SSO) through the pan-international eduGAIN federation with collaborative organization and group management using COmanage. CILogon supports over 2500 identity providers from more than 60 countries, as well as GitHub, Google, and ORCID. CILogon is a nonprofit hosted service offering from the University of Illinois at Urbana-Champaign. See https://www.cilogon.org. Tutorial attendees will learn how to log in to the hosted CILogon COmanage Registry to configure collaboration and group management for their astronomy or astrophysics project. They will configure both self-signup with approval and invitation enrollment workflows that enable their collaborators to register with the project using their federated identities. They will create in the registry groups and roles for users to be used to manage authorization to services like project wikis, data analysis portals, and other collaboration tools. Attendees will provision user data from the registry to an LDAP directory in order to facilitate other authentication and authorization workflows, including SSH login access to computing resources. Finally tutorial attendees will provision an OpenID Connect (OIDC) client key and secret using the registry and then configure a web service to use the key and secret to facilitate federated authentication and group-based authorization to the service.

Tutorial Abstract

Tutorial (T0)

Packaging and deploying with Conda

The goal of this tutorial is to teach participants how to release software through conda. Conda (https://conda.io) is a cross-platform binary package manager, and is steadily gaining traction as the way to install data science oriented software. It is particularly popular for Python and R, but is not limited to those languages. Participants will set up conda-only development environments, write recipes, and build cross-platform packages. Special attention will be given to building modern packages for legacy C and Fortran software. For those unfamiliar with conda, we will also show how to install packages as an end-user, while the experts get to self-host their own conda channel.

Tutorial Abstract

Tutorial (T0)

ALMA Archive test data as a service

In these times of clouds and petabytes of data, the software for telescopes must continue to be tested and improved each day. This cannot happen without data. A simpler approach is to take production data as the initial one. This is the case at the ALMA observatory. In this tutorial we propose you to learn of the new way of presenting terabytes of data in minutes. We reduced the time to deliver from 2 days to minutes by changing the way we provided our data for testing. In this tutorial, you will learn how to adapt this technique for your project.

Tutorial Abstract

Tutorial (T0)

Use Firefly to do interactive visualization in your web page

Firefly is IPAC's Advanced Astronomy WEB UI Framework. It is open source code hosted at GitHub. It is for building a web-based Front end to access science archives. Firefly APIs exposes the most powerful components of Firefly in a way that can be used by any web page with no prerequisites. It allows any web developer access to Firefly's image visualizers or Table Tool with just a very few lines of JavaScript. The goal is to make these APIs very easy to use. The three main components of Firefly are: image display, table data display, and plot display.This will be a hands on session with instructions to participants to do necessary setup before tutorial session. We will step through the lists of primary learning objective (see PDF file attached). We will have one main instructor and at least two assistants during the session to work with participants. The goal is for students to have something to take back home at the end of tutorial so they can continue to work on it later.

Tutorial Abstract

Tutorial (T0)

Building a Database Optimized for Astronomical Data

See PDF