Top 3 Elements of a Roadmap to R
Let's put validation to the side for a moment (we'll come to that later) and consider the top elements when creating your roadmap to R. The end is only part of the journey while the beginning can often dictate whether the journey will be successful.
Many groups delivering clinical trials are looking to expand their open source footprint to take advantage of ready-trained resources from university and to reduce the cost of their analytic platforms. But where can you even start the journey to adoption of open source, and R in particular?
Think about the main programming tasks that you perform every day, they would typically include the top 3 elements required to start that journey. They consist of data access, data processing and data visualization. In a like-for-like approach, to give flexibility of using differing languages for specific tasks, you should know the packages required to perform those same tasks.
For example, consider using base R packages, plus Haven for accessing data. Use Tidyverse and dplyr for processing data, and GGPlot2 and Survival for visualizing data. Not to mention putting together Rshiny applications for interactive visualizations.
There are these packages, and many more, available through differing sources. The immediate critical challenge in clinical trials involves validation! Not just the validation of the implementation of an R platform, but the validation of each of the packages - of which there are thousands available. Where do we start in making the use of R viable?
At d-wise, we advise our clients to start with a pragmatic approach to validation of R, identifying both common and edge use cases.There's a difference in using R for purely exploratory, development work versus using it for production deliverables. Perhaps an environment with no restrictions would work for development? But when it comes to production, the same rigorous validation principles apply as always have done.
If you are considering deploying an R platform, d-wise can help you with your approach to getting started so everyday tasks can be performed in R. This includes validation and advice on the 89 packages we recommend for that kick-start. Regardless of where your organization is on your adoption journey, d-wise is leading the way in modernizing statistical computing environments that leverage open source analytic tools.