Venables Award for New Developers of Open Source Software for Data Analytics

Venables Award for New Developers of Open Source Software for Data Analytics sponsored by the ARDC

The goal of this award is to encourage new open source software development from the Australian community with a view to support efforts to develop and share data science and statistics methodology. Here’s Bill Venables talking about his experiences, and why this award is important.

2023 Winners

Congratulations to all applicants! The final scores were very close!

In August 2023, the winners will present their software in a public talk. Watch the SSA events web site for the schedule.

Winner

• FRK: Fixed Rank Kriging
• Matthew Sainsbury-Dale, Andrew Zammit-Mangion

FRK is a framework for spatial/spatiotemporal modelling and prediction in which a set of basis functions is used to model the underlying (latent) process of interest. The fixed-rank basis-function representation facilitates the modelling of big data, and the method naturally allows for non-stationary, anisotropic covariance functions. Discretisation of the spatial domain into so-called basic areal units (BAUs) facilitates the integration of observations with varying support (i.e., both point-referenced and areal supports, potentially simultaneously), and prediction over arbitrary user-specified regions. ‘FRK’ also supports inference over various manifolds, including the 2D plane and 3D sphere, and it provides helper functions to model, fit, predict, and plot with relative ease. Version 2.0.0 and above also supports the modelling of non-Gaussian data (e.g., Poisson, binomial, negative-binomial, gamma, and inverse-Gaussian) by employing a generalised linear mixed model (GLMM) framework. Zammit-Mangion and Cressie describe ‘FRK’ in a Gaussian setting, and detail its use of basis functions and BAUs, while Sainsbury-Dale et al. describe ‘FRK’ in a non-Gaussian setting; two vignettes are available that summarise these papers and provide additional examples.

Runner-up

• predictNMB
• Rex Parsons, Robin Blythe, Adrian Barnett, Susannna Cramb, Steven McPhail

predictNMB is a tool to evaluate (hypothetical) clinical prediction models based on their estimated Net Monetary Benefit (NMB). It may be used by prediction model developers who intend to find out how performant their model needs to be clinically useful or by those in health services deciding whether or not to implement an existing model.

Honorable Mentions

• gnnad by Katie Buchhorn, Lukasz Mentel
• daggle by Mark Hanly

Call for Submission

Developers need to submit their work for consideration, along with a nomination from an SSA member.

• A brief description of the importance of the software to data science or statistical practice and discuss the advantages and disadvantages of the new contribution compared to existing implementations. Up to 400 words.
• An installable software product (e.g. package or web app) with its source code for use by the judging panel. It should be accompanied by enough information to allow the judges to effectively use and evaluate the software (including its design considerations).
• Supporting letters from up to three users, indicating how they use the software and benefits it has provided for their work.

For inspiration, see the information about the 2022 winners here.

Eligibility

We acknowledge our diverse developer community and especially encourage submissions from under-represented individuals and groups, and new software developers from academia, industry or government.

• Open to individual or a team developers based in Australia, who are relatively new to software development.
• Software must be distributed with an open source license, as described at https://opensource.org/licenses, actively developed and maintained. It should be noted also, that this means that users would not need to buy a license for any software in order to run the the submitted code.
• Software needs to have been developed primarily in Australia.
• The submission member(s) do not need to be an SSA members, but it does need to be accompanied by a nomination from an SSA member. SSA members can nominate more than one candidate.
• There is no specific programming language required.

Criteria

• Contribution makes a clear impact for the practice of data science or statistics.
• Software tackles an important problem, and has a clear purpose.
• Ease of installation and use.
• Code is well-structured, and clearly documented in the code.
• Documentation should also be publicly available, for example, via a dedicated web site containing instructions and example usage.
• Details clarifying how to cite or otherwise acknowledge the software must be available.
• Software supports and adheres to principles of reproducibility and interoperability. That is, analyses performed can be readily shared and reproduced by others, and software systems.
• The work must be:
  • funded or supported by Australian Funders or Institutions, or
  • conducted by researchers residing in Australia, or affiliated with an Australian Institution while they wrote the software
• First release of the software is within the past three years.
• Individual or (50% of) team is relatively new to software development, that is, this is the first or second statistical software product developed.

Judges & Assessors

The award will be managed by the SSA SCV committee, who will recruit a judging panel in any year.

Submit an Application

To submit, please complete this Application form

Important Dates

Date	What
Feb 6, 2023	Opening of submissions
Fri Apr 22, 2023	Close of submissions
Jun 2023	Announcement of the award winners

Prize

First prize will be $4000, with a runner-up prize of $1000. If the judges deem appropriate there may be equal first or runners-up, resulting in a division of the prize money.

Bill Venables

Bill Venables pioneered the use of the S and consequently R languages in Australia. His book “Modern Applied Statistics with S” co-authored with Brian Ripley was the primary manner that many analysts learned their trade, across the globe. It is now in its fourth edition. For many people, Bill’s tutorials on data analysis with S and R was the first entry point to working with data. Throughout his many years in academia and with CSIRO, Bill has contributed to the analysis of data from many fields but primarily in ecology, environment, and climate change.

About the Sponsor

The ARDC is a federally funded digital research infrastructure facility enabled by NCRIS. The ARDC Research Software program is pursuing a vision of recognition of research software as a first-class output of research. The sponsorship of this award is in aid of recognising the valuable contribution made to research when new ideas, methods and models are captured in this uniquely actionable form of knowledge representation.