Open source R software development: Translating methodological research into global impact
Dr Mai Vu, 2025
Purpose
The Department of Econometrics and Business Statistics at Monash University recognized a critical challenge in academic research: the gap between theoretical methodological advances and their practical application by researchers and practitioners worldwide. Traditional academic publishing, while essential for peer review and knowledge validation, often leaves innovative statistical methods trapped in academic papers, inaccessible to the broader community who could benefit from these advances. We aim to democratize access to cutting-edge statistical methodologies by translating our research into accessible open-source R packages.
Practice
Over the past decade, our research group has systematically developed nearly 100 open-source R packages, each designed to make complex statistical methodologies accessible to practitioners. Our approach involves parallel development of theoretical research and practical implementation, ensuring that major methodological advances are accompanied by user-friendly software. Key packages include the ‘forecast’ package for time series forecasting and analysis, ‘Ggally’ for data visualization, ‘naniar’ providing tools for handling missing data, and the ‘fable’ framework for tidy time series forecasting. Our development process emphasizes rigorous testing, comprehensive documentation, and adherence to tidy data principles, making our tools intuitive for both novice and expert users.
Output
Our R packages have achieved remarkable reach. The ‘forecast’ and ‘Ggally’ packages each average millions of downloads per year, while another 16 packages each have tens of thousands of monthly downloads. We have created comprehensive ecosystems of interconnected packages, such as the ‘fable’ framework, which provides end-to-end solutions for time series forecasting. Additionally, we have developed educational resources, including data packages that accompany our textbooks, ensuring seamless integration between learning materials and practical tools.
Outcome
Primary stakeholders -- researchers, data scientists, and students -- have gained unprecedented access to advanced statistical methods. Users can now implement sophisticated forecasting models, visualize complex data patterns, and conduct robust statistical analyses without requiring deep programming expertise. The packages have transformed how time series analysis and exploratory data analysis are conducted across disciplines, from economics and finance to environmental science and public health.
Impact
The broader impact extends far beyond our immediate user base. Our open-source approach has catalyzed a multiplier effect: the accessibility of our implementations has led to thousands of additional citations of our original research papers, amplifying the academic impact of our theoretical contributions.
The packages have become foundational tools in university curricula worldwide, shaping how the next generation of statisticians and data scientists approach their work.
Furthermore, by enabling reproducible research practices, our software has contributed to the broader movement toward open science and transparent methodology in quantitative research.