[R] Size of R user base

[Patrick Burns]

I disagree with Martin.  I think his confidence interval is sufficient for
management decisions.  There are three basic questions that should
be answered:

1)  Does it do what we want?

2) Will it disappear?

3) Is the quality sufficient?

Question 1 can only be answered within the organization.

Question 2 is partly answered by Martin's and other's estimates of
the user base.  The volume of messages on R-help is another clue --
this is probably more correlated with the number of new users than
the total number of users.  

I think there is sufficient evidence that R will survive at least as long
as the expected duration of a manager's tenure, and that there will
be a reasonable supply of potential employees with knowledge of
it.

As for Question 3, I -- who am a magnet for bugs-- have been using
R intensely for 2 years now, and have only found a couple of esoteric
bugs (one of which is clearly not down to R) plus two or three minor
bugs which have been fixed.  This is a level of excellence that I
never would have imagined.

In contrast, for 5 weeks I've been using a now obsolete version of
another language not unlike R on an operating system that ought to be
obsolete. In that time I've found 2 serious bugs (i.e., system 
terminating --
only one of which is esoteric), and a couple other moderately annoying
bugs.

Patrick Burns

Burns Statistics
patrick at burns-stat.com
+44 (0)20 8525 0696
http://www.burns-stat.com
(home of S Poetry and "A Guide for the Unwilling S User")

Martin Maechler wrote:

>
>About a `confidence interval': If we log transform it becomes
>feasible to give an interval I'm ``quite confident'' about:
>
>  log10(N) \in [3.5, 5.3]
>
>but then that's probably not informative enough for management
>decisions.
>
>Martin
>
>______________________________________________
>R-help at stat.math.ethz.ch mailing list
>https://www.stat.math.ethz.ch/mailman/listinfo/r-help
>PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
>
>
>  
>