Question about R software and output

Charity

There is OFTEN confusion what we mean when we say "R".

R is effectively a single bit of software with a ton of other bits of
software as optional extras.  You might think of some of those optional
extras like apps on a phone. You'd say you have a phone when you can open
the box and power it on. But you might add say "Spotify" app to it. That
will then be downloading the songs. And inevitably sending some data back
(I not do Spotify - no idea what - but it must have your identification and
what songs to play as a minimum).

R packages are like apps. They interact with the R core and do sometimes
amazing things. I think the majority do all that entirely on the machine
running R. BUT some may be using external services. There is for instance a
package that can interact with the Twitter API.  It would be impossible to
do that without sending as a minimum some login data and downloading data
back.  But you'd not normally send your entire research database to Twitter
during the login!

There will be SOME R packages that may be doing some odd stuff which might
need the dataset sent to the web. I can't think of any! Then there are
packages that could accidentally send data that you may or may not consider
sensitive. Say for instance you have a postcode list for every patient in
your research study. And you wanted to geocode that to get their location
as a long/lat... you can do that in an r package - which will have to send
the post code to get the answer. You haven't identified why you want the
postcode... but it is patient identifiable data...

That's R.

Add to that R Studio (or possibly other software solutions) - which is a
development environment which sits on top of R. It's often what people mean
by programming in R but it's effectively a glorified text editor. R Studio
desktop doesn't send data anywhere. Although it has database connectors
that could (but so can excel so if IG are flipping out feel free to remind
them of that!).

R Studio can work with git which is a version management system and can
upload code to a service like GitHub. You COULD upload data in that, or you
could chose not to.

And finally...

R Studio has a cloud version. It looks 98% the same as R Studio desktop.
Now called Posit.Cloud. That's a cloud service. While it may be secure etc
-- you are processing on their server not your PC.  And if this was your
question - it's a whole different set of answers!!

You may also want to Google NHSRcommunity where you might find help if
there are IG question to address as many of us will have wrestled with an
NHS IG team at some point.

Question about R software and output

Thread (9 messages)