Unfortunately, many researchers who could benefit from the use of R will never be good programmers. Graphical user interfaces such as R Commander, Rattle, and Deducer allow non-programmers to use R with minimal training. However, those tools store what they do in the form of R programs that their users might not understand. This makes re-use and repeatable research challenging for them. The free and open source KNIME software addresses this problem through the use of a workflow (flowchart) style of user interface. Users can simply choose from a list of common models to use. If a model is needed that does not come with the software, an R programming icon can be used in which any R code can be typed. Such custom nodes can be easily shared by other users. This approach allows for analyses that can range from quite simple to extremely complex, even blending in other programming languages such as Python or MATLAB. For handling big data, work can easily be passed to a Hadoop Map Reduce or Apache Spark system. This presentation will demonstrate how R integrates into KNIME.
