Developing Custom Platforms within JMP® Using JSL and R to Enable Additional Analysis
Mark Ewing, Statistician, Eastman Chemical Company
The ability to access R via JMP Scripting Language (JSL) allows for custom platforms to be developed, leveraging the interactivity of JMP and the breadth of statistical techniques available in R. Custom platform development in JMP through JSL helps ensure that the right tool is always available to the end user. The ability to link with R simplifies the programming required to use techniques that are not available in JMP, reducing the burden on the JSL programmer. The result is that JMP is more powerful and more flexible than other statistical software packages without giving up the level of interactivity that makes it so attractive. Change-point analysis (CPA) is a powerful technique for identifying shifts in historical data. Unlike a control chart, which identifies points that exhibit special causes, CPA is used on historical data sets to identify when a shift in the process mean occurred and the magnitude of the change. This technique has many applications in a variety of environments such as use in a control lab to identify the effect of maintenance, seasonal environmental conditions and other routine events on test equipment or use on the manufacturing floor to identify shifts in product levels and correlate the shifts with other changes in the process. This presentation will show a CPA script that leverages the change-point library in R to perform the calculations and JMP to visualize the output. The issues involved in developing an appropriate user interface in JMP for input and output will be discussed and explored interactively.