Accelerating Science with Open Source, sponsored by OSSci

In lieu of our July main meeting, ChiPy is pleased to co-host a special event with Open-Source Science (OSSci) and PyData Chicago.

 

Our world today is facing many challenges – from climate change, to global food security, to diseases like Alzheimer’s and cancer. These challenges require us to work together to quickly find new solutions. However, while open source has revolutionized the world of computing in general, and industry in particular, open source in science is less mature and still facing a variety of particular hurdles that make collaboration difficult and keep open source from reaching its full potential.

 

Meet Open-Source Science (OSSci), a new NumFOCUS initiative that promotes open-source software development to accelerate scientific research and discovery. OSSci brings together scientists and OSS developers to identify the needs of researchers contributing to and using open-source science projects, to fill in gaps in existing software, and to improve the OSS in science ecosystem overall. The OSSci interest groups – focused initially on chemistry/material science, life sciences/healthcare, and climate/sustainability – have been getting under way.

 

Join us as we hear from scientists and OSS developers about how they are advancing science through open source.

 

Want to give a talk? Please get in touch!

 

Agenda

  • 6.00pm – Doors open
  • 6.30pm – Talks
  • 7.30pm – Socializing
  • 9:00pm – The end

 

 

When: July 6, 2023, 6 p.m.

Where: Sully's House

Attendance:
In Person Pythonistas: 15

Topics

  • Repeatable Repeats with rpeat
    By: Jeff Ryan
    Description:

    We made a job scheduler and reporting tool for ourselves. And then it turned out to be too good to keep internal. So here it is. Named rpeat, because jobs repeat but it also repeats important details about your job's to you wherever you are. It is dead simple but feature rich, intuitive but powerful.

  • SGX3: Accelerated Services and Community Interactions Supporting Open-Source Science
    By: Sandra Gesing
    Description:

    Software has become a major driver for research with over 90% of researchers answering surveys that they use software for their research and over 65% expressing that they even could not do their research without software. Science gateways are defined as collaborative environments that allow science and engineering communities to access shared data, software, computing services, instruments, educational materials, and other resources specific to their disciplines. Their goal is to remove the barriers to online content, computing and data infrastructures. SGX3 is the newly funded NSF Center of Excellence for Science Gateways serving the science gateway community from users to providers to developers. Mature science gateway frameworks enable developers to re-use building blocks for typical tasks such as invoking simulations or sharing data. This way, a ramp up of a science gateway can be more efficient and developers can focus on the unique aspects of a science gateway that is tailored to a specific community. Many frameworks such as Hubzero and Tapis are open source and can be further developed by the community. SGX3 offers services to the community from UX design to technical gap analysis to internship opportunities.  The talk will go into detail for SGX3 and its services and examples for science gateways openly available.

  • QMCPy: Accelerating Computational Sciences through an Open-Source (Quasi) Monte Carlo Framework
    By: Sou-Cheng Choi
    Description:

    Over several decades, (Quasi-)Monte Carlo ((Q)MC) methods have become indispensable in computational sciences. However, the field still lacks comprehensive, user-friendly software that fully harnesses the power of (Q)MC algorithms in research, development, and production environments. In response to this gap, we introduce QMCPy, an open-source software framework designed to bring together global (Q)MC researchers and practitioners. Through this presentation, we will unpack the unique features that make QMCPy a critical tool in advancing computational sciences, including its extensibility, robustness, and integration with existing (Q)MC libraries. By illustrating the criteria and practices taken in the development of QMCPy, we aim to not only showcase our work but also to invite you to contribute to and co-create in this open-source software and accelerate scientific discovery through improved (Q)MC algorithms.

  • Building tools to support FAIR ML research at Globus Labs
    By: Will Engler
    Description:

    A lightning talk to discuss Globus Labs's ongoing work developing tools to support Findable, Accessible, Interoperable, and Reusable (FAIR) applied AI research in the natural sciences. Including Foundry, a platform for sharing and accessing AI-ready data for training scientific models. And Garden, a platform for making model discovery and hosted inference easier for scientific workflows.