10 things John Sall thinks scientists and engineers should know about JMP 14
Oct 23, 2018 10:58 AM
“We hear feedback clearly – and can address most of it,” says John Sall, Executive VP of SAS and chief architect of JMP. “But keep giving us your bottlenecks, your problems and your hidden frustrations."
John Sall welcomed a crowd at the world headquarters of SAS during the opening keynote for Discovery Summit 2018 in Cary, NC. The talk by the Executive VP of SAS and chief architect of JMP was a live demo JMP 14, which was released in March 2018.
Nearly all of the new capabilities in the product were added because of suggestions from users. If you want to know how you can impact product development for JMP, read my article on how features get into the product.
Here are 10 things Sall believes that scientists and engineers should know about JMP 14.
If you have pains around using multiple windows, check out the Window -> Move To/From Project to automatically build a tabbed project. This is an easy way to organize your JMP windows automatically. Projects can help you organize a JMP session and can serve as a storage container, display container or master window.
If you have ever wanted to open a folder full of text files in JMP – email or data that comes from multiple sites, devices or time periods – try the File -> Open Multiple to make quick work of this data import activity.
JMP as an analytic hub continues to expand its reach. Python joins the suite of languages that JMP Scripting Language can interface with. Also, HTTP Request lets you pull data directly from APIs, opening up JMP to a variety of data sources not easily accessed using standard query tools, like the Query Builder.
Dirty data continues to be a challenge for many data explorers. Recode adds capabilities to handle more difficult data cleanup tasks in columns containing categories.
Clickable cells, afforded by the new column property Event Handler, will let you add web commands and navigate directly to websites, among other use cases.
Graph Builder, the most widely used platform in JMP, adds a new chart type invented by Xan Gregg (@XanGregg) : packed bars. Packed Bars combines the best aspects of the bar chart and tree map. You can find this new chart type by changing Bar Style from its default to Packed in the drop-down menu. Also, if you ever have been given pre-summarized data, you can still get error bars by using the new interval drop zone.
Innovation in design of experiments (DOE) continues with an alternative optimality: A. You can weight the effects in the model that will build an experiment that are optimal for what you are interested in – interactions, for example.
Get to know and love the SHASH distribution. Where you’ve used the Johnson SU in the past, the SHASH’s parameter estimates can be much more stable. The normal distribution is a limiting case of the Johnson SU distribution, but the SHASH handles it nicely as an all-purpose fitter.
If you are involved in process monitoring, you may want to detect shifts. The new CUSUM Control Chart platform can be more sensitive to subtle shifts that the I/MR Control Chart may miss.
Functional data is added to the new types of data that JMP handles. Functional Data Explorer (in JMP Pro) lets you answer such questions as: How can I make something behave according to a particular profile? Or: How can I predict something using functional data?
Enjoy using JMP 14. If you are already using JMP 14, make sure you have upgraded to the recently released JMP 14.1. Also, ensure that you have "check for updates" turned on so you are always running the latest version. For more information about why you should keep JMP updated, read my article on JMP maintenance releases.