FDA: Have you come across competition with newer R packages already?

Question

Hi

I saw some new R packages using "functional data analysis" either as a web based app service (http://www.statfda.com/ ) or as real R Package (https://cran.r-project.org/web/views/FunctionalData.html). They seem to go beyond nonlinear or PCA approaches and it looks like they use also spline interpolation to smooth the data and use it for further modeling, often NN or Random Forest models from what I could see.

I saw one customer using R and Spotfire in combination to embed the R analysis in an easy to use Web-Interface. Have you seen similar cases already? How does development feel about this? Will this be a threat for our (may not anymore) unique value proposition?

@RyanParker @chris_gotwalt1 JMP Pro

MikeD_Anderson · Accepted Answer

Some of the R/Python packages have been around for a while - FDA isn't new. I think the first implementation of FDA code was in R (could be wrong there).  Our advantage over the code-based implementations is the GUI.  The website you mention is a web app version of the R package put out by the University of Granada:

The only time I've had an issue with going against R is with people who feel their entire job situation depends on the company needing them to code R.  Most customers are getting the idea that with open-source, you get what you pay for.  They also know they're going to need to put significant resources behind cross-checking anything home-grown from open-source tools.

benfrancis · Accepted Answer

@saradoudt's R/Python resource is important here - R / Python Training .

A good discussion happened in the meeting and comments on the page; sometimes JMP Pro is used in lieu of the packages you highlight, sometimes it complements. We can enlighten about what JMP Pro has to offer and let the customer determine value.