July Project Updates

NumFOCUS
9 min readJul 31, 2024

--

Sponsored Project Updates

TARDIS

We have successfully wrapped up another TARDIS Con! TARDIS Con 2024 was a unique two-week un-conference where we worked on improving the code’s efficiency, implementing new features, and discussing the latest scientific insights related to supernovae. We built samplers for the new TARDIS physics processes along with improving old ones and also built a new logger widget for TARDIS Simulation runs! For more information please visit- https://tardis-sn.github.io/news/

  • We are also excited to announce that all three of our GSoC students have passed the midterm evaluations. So far, we have a visualization tool that plots photon packet interactions in the ejecta, improved benchmarks, and newly restructured trackers, all of which will significantly aid in the study and interpretation of supernova spectra.

rOpenSci

  • Announcing New Software Peer Review Editors: Beatriz Milz and Margaret Siple
  • A fresh new look for R-universe
  • Resources from the rOpenSci community at useR! 2024
  • Upcoming coworking sessions: “Building your first R package” in August, TBA in September
  • New package {osmapiR}
  • Package news
  • 3 use cases
  • Blog posts
  • Calls for contributions
  • Package Development corner
  • rOpenSci News Digest, July 2024
  • New editors, new look for R-universe, useR! resources, coworking

Upcoming events

Join us for social coworking and office hours monthly on the first Tuesdays! Hosted by Steffi LaZerte and various community hosts, everyone is welcome. No RSVP is needed. Consult our Events page to find your local time and how to join.

Tuesday, August 6th, 9:00 America Pacific (16:00 UTC), Building your first R package with cohost Carolina Pradier and Steffi LaZerte.

  • Explore how to make R packages
  • Plan out that package you’ve always wanted to create
  • Chat with our cohost about tips and tricks for making your first R package

LatinR 2024 Call for Papers is open. Due Date: August 12, 2024. Online event. You can sent the proposal in English, Spanish and Portuguese. Link: https://openreview.net/group?id=LATIN-R.com/2024/Conference

You can check all the events rOpenSci will participate in here: https://ropensci.org/events/

Recent releases

A fresh new look for R-universe!

You might have noticed that R-universe got a big refresh. Read all about this big overhaul of the interface.

Calls for participation

Calls for maintainers

If you’re interested in maintaining any of the R packages below, you might enjoy reading our blog post What Does It Mean to Maintain a Package?

Calls for contributions

Also, refer to our help wanted page — before opening a PR, we recommend asking in the issue whether help is still needed.

OpenFHE

OpenFHE (a NumFOCUS Sponsored Project) is looking for contributors for a Numpy-like matrix arithmetic library that will support privacy-preserving computations based on fully homomorphic encryption (FHE). Details are available here: https://bit.ly/openfhecfc. At the end of the slides, you will find information to drop in your application/contact.

DASK

Next Dask Demo Day on Aug. 1. Join us for demos and updates from the Dask community. Have something you’d like to share? Let us know.

New to Dask?

Check out our tutorials on using Dask DataFrames, parallelizing your Python code, plus more advanced use cases.

napari

napari recently released versions 0.5.0 and 0.5.1! 0.5.0 is a *huge* milestone for the project because it involved a significant architectural overhaul and so it was the first release from the main branch in 18 months!

But the dividends paid off immediately with an easy follow-up bugfix release in 0.5.1. See the release notes for 0.5.0 here and those for 0.5.1 here. The napari team is excited to return to a fast-release cadence in the coming weeks and months! Related: Lucy Liu led most of the architecture efforts over that time, with help from team member Draga Doncila Pop, and has now joined the core team!

napari also has received a small grant from the Chan Zuckerberg Imaging Institute to implement instanced mesh rendering for fast rendering of repeating units in a 3D volume. If you have experience with VisPy and/or OpenGL and some room to earn a bit of extra cash in the coming months, please get in touch by writing to napari-core-devs@googlegroups.com! Or just say hello in our Zulip chat room.

mlpack

Great month for mlpack… we won a NASA ROSES grant to fund two mlpack contributors to work on improving the library for spaceflight machine learning applications!

Spyder

  • Released 5.5.6, which allows Spyder to run without QtWebEngine
  • Also released 6.0 Beta 3, which is the last beta before the next major stable release and very close to what users will see in Spyder 6.0
  • Presented a tools plenary talk, a poster, and a Birds of a Feather session at SciPy 2024

ITK

The third Get Your Brain Together Hackathon, held from July 26th to July 28th, 2024, was a dynamic event that brought together over 30 participants from academic, industry, and private research institutions. The hackathon, hosted in a hybrid format at the University of North Carolina-Chapel Hill and online, focused on advancing OME-Zarr spatial transformations. The event featured tutorial sessions on the first day, collaborative review and proposal sessions on the second day, and hands-on implementation activities on the third day. Participants engaged in discussions and activities aimed at enhancing the current coordinate transformations draft and incorporating relevant neuroimaging additions.

One of the notable outcomes of the hackathon was the development of an OME-Zarr Request for Comments (RFC) on Coordinate Transformations and Axis Anatomical Orientation. This RFC aims to standardize spatial transformations in OME-Zarr, promoting reproducibility, integration with analysis workflows, and efficiency in handling large-scale bioimages. The collaborative efforts during the event underscored the importance of having standardized formats and tools to ensure consistency and accuracy across different platforms and applications in the neuroimaging community.

The hackathon’s tutorial sessions, recordings, and materials are now available on the event’s website, providing valuable resources for ongoing and future projects. The hackathon not only fostered innovation and collaboration but also highlighted the community’s commitment to advancing open-source resources and standards that facilitate the discovery of brain structure and function.

Affiliated Project Updates

Crystal

The Crystal programming language is proud to announce the release of Crystal 1.13. This release follows the previous one, improving support for multi-threaded applications.

STUMPY
STUMPY 1.13.0 has been released!

Highlights:

  • Easier-to-Use Matrix Profile (Array) Data Structure
  • NumPy 2.0 Support
  • pyproject.toml Adoption
  • Improved Documentation and Testing
  • Python 3.12 Support
  • And Much More!

Optuna

  • We released beta version of Optuna v4.0/OptunaHub.
  • With OptunaHub, you can get/share state-of-the-art algorithms/useful visualizations very easily. Check out the blog post for more information.
  • The new release also includes the stabilization of artifacts and JournalStorage and the removal of old multi-objective optimization interfaces, CLI interfaces, etc. If you use MOTPESampler, please migrate to TPESampler.
  • Hideaki Imamura and Shuhei Watanabe, our core developers, will give presentations at the AutoML Conference in September. Stay tuned! Presentation 1 Link | Presentation 2 Link

pvlib python

The 11th major version of pvlib python v0.11.0 was released on June 21, 2024. This release is the product of contributions from 16 people, including three participants in the Google Summer of Code program!

v0.11.0 Highlights:

  • Two new spectral correction factor models
  • A water albedo model for floating PV applications
  • Functions to calculate shaded fraction and electrical loss due to row-to-row shading
  • A simple transformer efficiency model

v0.11.0 also contains several breaking changes relative to the 0.10.* series. Please check the documentation for details: https://pvlib-python.readthedocs.io/en/stable/whatsnew.html

Releases are available from PyPI and the conda-forge channel:

Report issues & contribute:

Thank you for using pvlib python!

Polars

Below are links to recently published updates from Polars.

Mesa

Mesa Google Summer of Code

Mesa’s currently has three contributors through Google Summer of Code and they are performing exceptionally. Our contributors are working on Mesa Frames to vectorize operations, Mesa RL to link Mesa to reinforcement learning libraries and Cacheable Mesa so Users can replay critical replay their simulation without requiring the compute overhead.

Mesa Builds to 3.0

Mesa continues to build to our 3.0 release pre-release. At this point there are two major breaking changes at this point:

  • The old visualization is removed in favor of the new, Solara-based Jupyter Viz. This was already available in the 2.3.x release series but is now stabilized. Check out our new Visualization Tutorial. More examples and a migration guide will follow later in the Mesa 3.0 development.
  • The `mesa.flat` namespace is removed since it was not used very often.
  • Mesa 3.0 will require Python 3.10+.

Mesa-Geo Builds to 1.0

We are doing some major upgrades to Mesa-Geo to ensure it is compatible with the current version of Mesa, particularly the visualization.

Join us on Matrix

Join us on GitHub

aeon

Some of the developers at the aeon project will be giving tutorials on time series machine learning at the KDD 2024 (https://aeon-tutorials.github.io/KDD-2024/) and ECML 2024 (https://github.com/aeon-tutorials/ECML-2024) conferences in August and September.

igraph

skforecast

Skforecast new release 0.13.0

  • The Global Forecasters ForecasterAutoregMultiSeries and ForecasterAutoregMultiSeriesCustom can forecast series not seen during training. This is useful when new products are introduced but no historical data is available yet.
  • A new create_predict_X method has been added to all recursive and direct forecasters, allowing the user to inspect the matrices passed to the regressor’s predict method.
  • New metrics module with functions to compute time series prediction metrics such as mean_absolute_scaled_error and root_mean_squared_scaled_error.
  • New argument add_aggregated_metric in backtesting_forecaster_multiseries to include, in addition to the metrics for each level, the aggregated metric of all levels using the average (arithmetic mean), weighted average (weighted by the number of predicted values of each level), or pooling (the values of all levels are pooled and then the metric is calculated).
  • Added the skip_folds argument to the model_selection and model_selection_multiseries functions. It allows the user to skip some folds during backtesting, which can be useful to speed up the backtesting process and thus the hyperparameter search.

--

--

NumFOCUS

Our open source scientific software projects are changing the world. Learn more on our website: https://numfocus.org