Introducing openstatsware

Who we are and what we build together

Daniel Sabanés Bové on behalf of openstatsware

2024-10-30

Table of contents

• Introducing openstatsware
• Our work
• Outlook

Introducing openstatsware

openstatsware

• Formed on 19 August 2022
• Official working group of the American Statistical Association (ASA) Biopharmaceutical section (BIOP)
• Special Interest Group (SIG) of the European Federation of Statisticians in the Pharmaceutical Industry (EFSPI).
• Cross-industry collaboration (58 members from 37 organizations)
• Homepage: openstatsware.org
• We welcome new members to join!

Motivation

• Open-source software increasingly popular in Biostatistics
  • Rapid uptake of novel statistical methods
  • Unprecedented opportunities for collaboration
  • Transparency of methods and implementation
• Variability in software quality
  • No statistical quality assurance on open-source extension package repositories, e.g. CRAN
  • No industry standard for assessing quality of R packages
• Reliable software for core statistical analyses is paramount

Our work

openstatsware objectives

• Engineer selected packages to fill in gaps in the open-source statistical software landscape, and to promote software tools designed by the working group through publications, conference presentations, workshops, and training courses.

• Develop good SWE practices for engineering high-quality statistical software and promote their use in the broader Biostatistics community via public training materials.

• Communicate and collaborate with other R software initiatives including via the R Consortium.

Workstreams in Package Development

• Mixed Models for Repeated Measures (MMRM)
  • Developed the mmrm R package for frequentist inference in MMRM
• Bayesian MMRM
  • Developed the brms.mmrm R package for Bayesian inference in MMRM
• Health Technology Assessment (HTA)
  • Developed the maicplus R package for matching-adjusted indirect comparison (MAIC)
• Bayesian Safety Signal Detection
  • Developed the SafetySignalDetection.jl Julia package

Best Practices Dissemination - Workshop

• Workshop “Good Software Engineering Practice for R Packages” on world tour
• To teach hands-on skills and tools to engineer reliable R packages
  • Topics: R package structure, engineering workflow, ensuring quality, version control, collaboration and publication, and shiny development
• 5 events in 2023 in Basel, Shanghai, San José, Rockville, and Montreal
• 4 events in 2024 in Zurich, Salzburg, Beijing, and this Monday online at R/Pharma APAC!

Best Practices Dissemination - openstatsguide

• Small and concise set of recommendations for package developers
• Opinionated, but aims to be based on experienced majority opinions
• Focus are developers, while users might find complementary “validation” frameworks valuable
• Primarily for statistical packages (not plotting, data wrangling, etc.)
• Generic principles which can be used across functional data science languages R, Python, and Julia
• Concrete tools are mentioned as examples

Outlook

Long Term Perspective

• Software engineering is a critical competence in producing high-quality statistical software
• A lot of work needs to be done regarding the establishment, dissemination and adoption of best practices for engineering open-source software
• Improving the way software engineering is done will help improve the efficiency, reliability and innovation within Biostatistics

Next Steps

• Join the Effective Statistician conference invited session organized by openstatsware
• We will relaunch the CRAN Task View on Clinical Trials
• We are considering to write a Shiny app development guide

Q&A