エピソード

  • Modernizing Weather Forecasting with Julia – Milan Klöwer on SpeedyWeather.jl
    2024/10/03
    #20: Milan Klöwer joins Arfon and Abby to discuss SpeedyWeather.jl, its development, and its role in climate science. Milan is a Schmidt AI in Science Fellow at the University of Oxford. You can follow Milan on X/Twitter: https://twitter.com/milankloewer bsky: https://bsky.app/profile/milank.bsky.social and at his website milankl.github.io Episode highlights: [01:10] - Introducing Milan Clover[01:56] - Understanding Atmospheric General Circulation Models[05:15] - The Evolution of Weather Prediction[07:23] - SpeedyWeather.jl: A New Approach[20:05] - The Legacy of Fortran in Climate Science[20:51] - Why Julia?[25:35] - Building SpeedyWeather.jl[26:11] - Interactive Modeling with SpeedyWeather.jl[34:07] - The Importance of Open Source in Climate Science[35:29] - Call for Contributors[40:53] - Final Thoughts and Future Plans Links: JOSS paper: https://joss.theoj.org/papers/10.21105/joss.06323 SpeedyWeather.jl repository: https://github.com/SpeedyWeather/SpeedyWeather.jlMilan’s website: milankl.github.io Milan social media: X/Twitter: https://twitter.com/milankloewer bsky: https://bsky.app/profile/milank.bsky.socialThe Journal of Open Source Software (Twitter/X, blog)@arfon on (fosstodon, Linkedin, GitHub, website)@abbycabs on (Twitter/X, hachyderm, bsky, Linkedin, GitHub, website)Donate to JOSS Supercharge your research with the latest scientific software showcased in the Journal of Open Source Software (JOSS). Hear directly from authors on their work, their motivations, and new ways open source software can accelerate your work. Hosted by editor-in-chief Arfon Smith and founding editor Abby Cabunoc Mayes, each episode features an interview with different authors of published papers in JOSS. Tune in to learn about the latest developments in research software engineering and open science, and how they are changing the way research is conducted. New episodes every other Thursday.
    続きを読む 一部表示
    42 分
  • Tackling the Machine Learning Reproducibility Crisis – KJ Schmidt & Aristana Scourtas on Foundry-ML
    2024/09/19

    #19: KJ Schmidt & Aristana Scourtas join Arfon and Abby to discuss Foundry-ML, a platform designed to simplify using machine learning datasets, highlighting its development, impacts, and their career advancements post-project.

    KJ just started a position at the Institute for Genomic Medicine within Nationwide Children's Hospital. Aristana is a Product and Research Manager at The Library Innovation Lab at Harvard Law School. KJ and Aristana both worked on Foundry-ML during their time working at UChicago and Globus.

    You can follow Aristana on LinkedIn https://www.linkedin.com/in/aristana/ and Twitter/X @aristana_s. You can follow KJ on Twitter/X @kj_schmidt or LinkedIn https://www.linkedin.com/in/schmidtkj/.

    Episode highlights:

    [01:54] Beginning of Interview with KJ Schmidt and Aristana Scourtas [02:02] What is Foundry-ML? [04:02] The Role of Globus in Foundry-ML [05:29] Reproducibility in Machine Learning [06:45] Applications and Collaborations [09:15] New Roles and Future Plans [11:01] Maintaining Foundry-ML [12:37] Sustainability in Open Source [13:12] Community Building in Open Source [21:49] Challenges and Lessons Learned [24:28] Publishing in JOSS [25:38] Closing Remarks and Contact Information

    Links:
    • JOSS paper: https://joss.theoj.org/papers/10.21105/joss.05467
    • Foundry repository: https://github.com/MLMI2-CSSI/foundry
    • Website: https://foundry-ml.org/
    • Movement Building from Home
    • KJ on Twitter/X @kj_schmidt or LinkedIn https://www.linkedin.com/in/schmidtkj/
    • Aristana on LinkedIn https://www.linkedin.com/in/aristana/ and Twitter/X @aristana_s
    • The Journal of Open Source Software (Twitter/X, blog)
    • @arfon on (fosstodon, Linkedin, GitHub, website)
    • @abbycabs on (Twitter/X, hachyderm, bsky, Linkedin, GitHub, website)
    • Donate to JOSS

    Supercharge your research with the latest scientific software showcased in the Journal of Open Source Software (JOSS). Hear directly from authors on their work, their motivations, and new ways open source software can accelerate your work.

    Hosted by editor-in-chief Arfon Smith and founding editor Abby Cabunoc Mayes, each episode features an interview with different authors of published papers in JOSS. Tune in to learn about the latest developments in research software engineering and open science, and how they are changing the way research is conducted.

    New episodes every other Thursday.

    続きを読む 一部表示
    28 分
  • Visualize Hypergraphs with Open Source – Brenda Praggastis on HyperNetX
    2024/09/05

    #18: Brenda Praggastis joins Arfon and Abby to discuss HyperNetX, a Python package for modeling complex network data as hypergraphs. Brenda discusses the distinctions between graphs and hypergraphs, their real-world applications, and the collaborative nature of the hypergraph community. She also delves into the open-source development process, the capabilities and limitations of HyperNetX, and the importance of community contribution in advancing hypergraph research. Tune in to learn more about hypergraphs and their growing significance in various research fields.

    Brenda is a mathematician and data scientist at the Pacific Northwest National Laboratory.

    You can follow HyperNetX on GitHub https://github.com/pnnl/HyperNetX.

    Episode highlights:

    [01:59] - Interview Begins: Welcome Brenda!

    [02:05] - Understanding Hypergraphs

    [04:57] - Brenda's Journey to Hypergraphs

    [06:03] - Visualizing Hypergraphs

    [11:31] - Deep Dive into HyperNetX

    [19:58] - Applications of Hypergraphs

    [22:25] - Scalability and Performance of HyperNetX

    [24:54] - Publishing in JOSS

    [26:16] - Contributing to HyperNetX

    Links:

      • JOSS paper: https://joss.theoj.org/papers/10.21105/joss.06016

      • HyperNetX repository: https://github.com/pnnl/HyperNetX

      • Docs: https://hypernetx.readthedocs.io/en/latest/

      • The Journal of Open Source Software (Twitter/X, blog)

      • @arfon on (fosstodon, Linkedin, GitHub, website)

      • @abbycabs on (Twitter/X, hachyderm, bsky, Linkedin, GitHub, website)

      • Donate to JOSS

    Supercharge your research with the latest scientific software showcased in the Journal of Open Source Software (JOSS). Hear directly from authors on their work, their motivations, and new ways open source software can accelerate your work.

    Hosted by editor-in-chief Arfon Smith and founding editor Abby Cabunoc Mayes, each episode features an interview with different authors of published papers in JOSS. Tune in to learn about the latest developments in research software engineering and open science, and how they are changing the way research is conducted.

    New episodes every other Thursday.


    続きを読む 一部表示
    31 分
  • Setting the Standard for Audio Data Management – Magdalena Fuentes on Soundata
    2024/08/22

    #17: Magdalena Fuentes joins Arfon and Abby to discuss her project Soundata. Magdalena shares insights into audio datasets, standardization, and community-driven open-source tools, offering valuable takeaways on reproducibility, dataset management, and open source in machine listening.

    Magdalena is an Assistant Professor of Music Technology and Integrated Design & Media at New York University (NYU).

    You can follow Magdalena on Twitter/X @mfu3ntes, GitHub @magdalenafuentes, LinkedIn @mfu3ntes https://www.linkedin.com/in/mfu3ntes/ or at her website magdalenafuentes.com.

    Episode highlights:

    [02:02] - Magdalena's Background and Motivation [02:54] - Challenges in Machine Listening [03:48] - Standardizing Audio Data with Soundata [06:58] - Applications and Benefits of Soundata [12:33] - Industry and Academic Use Cases [13:27] - The Broader Landscape of Audio Data Tools [16:37] - Encouraging Community Contributions [21:48] - Maintaining and Evolving Soundata [29:40] - Future Goals and Vision for Soundata

    Links:
    • JOSS paper: https://joss.theoj.org/papers/10.21105/joss.06634
    • Soundata repository: https://github.com/soundata/soundata
    • Magdalena on Twitter/X @mfu3ntes https://x.com/mfu3ntes, GitHub @magdalenafuentes https://github.com/magdalenafuentes, LinkedIn @mfu3ntes https://www.linkedin.com/in/mfu3ntes/, website: http://magdalenafuentes.com/
    • The Journal of Open Source Software (Twitter/X, blog)
    • @arfon on (fosstodon, Linkedin, GitHub, website)
    • @abbycabs on (Twitter/X, hachyderm, bsky, Linkedin, GitHub, website)
    • Donate to JOSS

    Supercharge your research with the latest scientific software showcased in the Journal of Open Source Software (JOSS). Hear directly from authors on their work, their motivations, and new ways open source software can accelerate your work.

    Hosted by editor-in-chief Arfon Smith and founding editor Abby Cabunoc Mayes, each episode features an interview with different authors of published papers in JOSS. Tune in to learn about the latest developments in research software engineering and open science, and how they are changing the way research is conducted.

    New episodes every other Thursday.

    続きを読む 一部表示
    33 分
  • Greenhouse Gas Flux Analysis Made Easy – Karelle Rheault on goFlux
    2024/08/08

    #16: Karelle Rheault joins Arfon and Abby to discuss the development of goFlux, a user-friendly software designed to calculate greenhouse gas (GHG) fluxes. The conversation covers Karelle's programming journey, challenges faced, and the role of open-source practices in advancing research.

    Karelle is a PhD student and the creator of goFlux.

    You can follow Karelle on X/Twitter @KarelleRheault and on LinkedIn https://www.linkedin.com/in/karelle-rheault-063842105/

    Episode highlights:

      • [02:03] - Karelle's Journey to Creating goFlux

      • [02:56] - Understanding Greenhouse Gas Fluxes

      • [05:19] - Measurement Techniques and Tools

      • [09:30] - Challenges and Advantages of goFlux

      • [20:20] - Open Source Contributions and Community Support

      • [23:57] - Karelle's Experience with JOSS

      • [27:00] - Future Plans and Contributions

      • [28:59] - Organizing the Community and Setting Standards

      • [31:27] - Conclusion and Farewell

    Links:

      • JOSS paper: https://joss.theoj.org/papers/10.21105/joss.06393

      • goFlux repository: https://github.com/Qepanna/goFlux

      • goFlux webpage: https://qepanna.quarto.pub/goflux/

      • Karelle on X/Twitter https://x.com/KarelleRheault, LinkedIn: https://www.linkedin.com/in/karelle-rheault-063842105/, email: karh [at] ign.ku.dk

      • The Journal of Open Source Software (Twitter/X, blog)

      • @arfon on (fosstodon, Linkedin, GitHub, website)

      • @abbycabs on (Twitter/X, hachyderm, bsky, Linkedin, GitHub, website)

      • Donate to JOSS

    Supercharge your research with the latest scientific software showcased in the Journal of Open Source Software (JOSS). Hear directly from authors on their work, their motivations, and new ways open source software can accelerate your work.

    Hosted by editor-in-chief Arfon Smith and founding editor Abby Cabunoc Mayes, each episode features an interview with different authors of published papers in JOSS. Tune in to learn about the latest developments in research software engineering and open science, and how they are changing the way research is conducted.

    New episodes every other Thursday.

    続きを読む 一部表示
    32 分
  • Data Quality Crafted with Passion – Herminio Vazquez & Virginie Grosboillot on cuallee
    2024/07/25

    #15: Herminio Vazquez & Virginie Grosboillot join Arfon and Abby to discuss cuallee, their Python package designed for data quality checks across multiple data framed APIs. The discussion covers the importance of data quality, the passion behind the project, and the intersection of open-source software in both academia and industry.

    Herminio is the Director Data Strategy and Analytics at Copado. Virginie is a postdoctoral researcher at the University of Ljubljana.

    You can follow them both on LinkedIn: Hermino, Virginie

    Episode highlights:

    • [02:02] - The Story Behind cuallee
    • [03:32] - Guest Backgrounds and Expertise
    • [05:08] - Applications in Bioinformatics
    • [09:27] - Industry Applications and Inspirations
    • [11:21] - Comparing cuallee with Other Tools
    • [23:26] - Open Source Contributions and Community
    • [27:12] - Challenges in Building cuallee
    • [32:17] - Conclusion and How to Get Involved

    Links:

    • JOSS paper: https://joss.theoj.org/papers/10.21105/joss.06684
    • cuallee repository: https://github.com/canimus/cuallee
    • Herminio on LinkedIn: https://www.linkedin.com/in/herminio-vazquez-701bb0/
    • Virginie on LinkedIn: https://www.linkedin.com/in/virginie-grosboillot-7a7b8b23a/
    • The Journal of Open Source Software (Twitter/X, blog)
    • @arfon on (fosstodon, Linkedin, GitHub, website)
    • @abbycabs on (Twitter/X, hachyderm, bsky, Linkedin, GitHub, website)
    • Donate to JOSS

    ---

    Supercharge your research with the latest scientific software showcased in the Journal of Open Source Software (JOSS). Hear directly from authors on their work, their motivations, and new ways open source software can accelerate your work.

    Hosted by editor-in-chief Arfon Smith and founding editor Abby Cabunoc Mayes, each episode features an interview with different authors of published papers in JOSS. Tune in to learn about the latest developments in research software engineering and open science, and how they are changing the way research is conducted.

    New episodes every other Thursday.

    続きを読む 一部表示
    35 分
  • Open Source Revolution in Radiotherapy – James Kerns on Pylinac
    2024/07/11

    #14: James Kerns joins Arfon and Abby to discuss how Pylinac automates quality assurance for radiotherapy, running an open source project while teaching yourself to code, and the growing open source ecosystem for medical physics.

    James is a technical lead at Radformation, focusing on image analysis.

    You can follow James on LinkedIn and GitHub.

    Episode highlights:

    • [00:37] - The Importance of Quality Assurance in Radiotherapy
    • [05:19] - James's Journey into Coding
    • [07:48] - How Pylinac Works
    • [09:56] - Community and Support
    • [16:43] - Commercial Use and Collaboration
    • [19:53] - Transitioning from MATLAB to Python
    • [21:27] - Maintaining Pylinac Over the Years
    • [23:21] - Challenges in Software Development
    • [26:34] - The Importance of Documentation
    • [29:01] - Publishing in JOSS
    • [35:23] - Future of Pylinac and New Projects

    Links:

    • JOSS paper: https://joss.theoj.org/papers/10.21105/joss.06001
    • Pylinac repository: https://github.com/jrkerns/pylinac
    • Awesome Medical Physics: https://github.com/jrkerns/awesome-medphys
    • quacc: https://github.com/jrkerns/quaac
    • Radformation: https://www.radformation.com/
    • James on LinkedIn: https://www.linkedin.com/in/james-kerns/, GitHub: https://github.com/jrkerns
    • The Journal of Open Source Software (Twitter/X, blog)
    • @arfon on (fosstodon, Linkedin, GitHub, website)
    • @abbycabs on (Twitter/X, hachyderm, bsky, Linkedin, GitHub, website)
    • Donate to JOSS

    ---


    Supercharge your research with the latest scientific software showcased in the Journal of Open Source Software (JOSS). Hear directly from authors on their work, their motivations, and new ways open source software can accelerate your work.

    Hosted by editor-in-chief Arfon Smith and founding editor Abby Cabunoc Mayes, each episode features an interview with different authors of published papers in JOSS. Tune in to learn about the latest developments in research software engineering and open science, and how they are changing the way research is conducted.

    New episodes every other Thursday.

    続きを読む 一部表示
    40 分
  • Humans Think in Graphy Ways – Naomi Arnold and Ben Steer on Raphtory
    2024/06/27

    #13: Naomi Arnold and Ben Steer join Arfon and Abby to discuss their work on Raphtory, THE temporal graph engine for Rust and Python. Other topics include: pomegranates are the graphiest fruit, tracking ship to ship trading, and funding for open source projects through the Tools, Practices and Systems program at the Alan Turing Institute.

    Naomi is a Postdoctoral Research Associate at Northeastern University London. Ben is the CTO and Co-founder of Pometry.

    You can follow Naomi on Twitter/X @narnolddd. You can follow Ben on Twitter/X @Raphtory, raphtory.com, and pometry.com.


    Episode Highlights:

    • [00:00] - Introduction to Open Source for Researchers
    • [00:19] - Meet the Guests: Naomi Arnold and Ben Steer
    • [01:49] - Understanding Temporal Graphs with Raphtory
    • [04:40] - The Origin Story of Raphtory
    • [08:05] - Naming and Evolution of Raphtory and Pometry
    • [12:47] - Raphtory's Applications and Use Cases
    • [19:00] - Transition from Scala to Rust
    • [24:47] - Open Source Journey and Contributions
    • [35:01] - Conclusion and Final Thoughts

    Links:

    • JOSS paper: https://joss.theoj.org/papers/10.21105/joss.05940
    • Raphtory repository: https://github.com/Pometry/Raphtory
    • Naomi on Twitter/X: https://twitter.com/narnolddd
    • Raphtory on Twitter/X https://twitter.com/raphtory
    • Raphtory website: https://www.raphtory.com/
    • Pometry website: https://www.pometry.com/
    • The Journal of Open Source Software (Twitter/X, blog)
    • @arfon on (fosstodon, Linkedin, GitHub, website)
    • @abbycabs on (Twitter/X, hachyderm, bsky, Linkedin, GitHub, website)
    • Donate to JOSS

    ---

    Supercharge your research with the latest scientific software showcased in the Journal of Open Source Software (JOSS). Hear directly from authors on their work, their motivations, and new ways open source software can accelerate your work.


    Hosted by editor-in-chief Arfon Smith and founding editor Abby Cabunoc Mayes, each episode features an interview with different authors of published papers in JOSS. Tune in to learn about the latest developments in research software engineering and open science, and how they are changing the way research is conducted.


    New episodes every other Thursday.

    続きを読む 一部表示
    36 分