[Rd] organisation of packages & CRAN

Spencer Graves spencer.graves at prodsyse.com
Mon Nov 10 03:36:21 CET 2014


On 11/9/2014 4:50 PM, Ben Bolker wrote:
> Spencer Graves <spencer.graves <at> prodsyse.com> writes:
>
>>         Might it be appropriate to add "http://metacran.github.io/search"
>> and the "sos" package to the official list of R search capabilities at
>> "www.r-project.org/search.html"?  [Disclaimer:  I'm the lead author of
>> "sos".]
>>
>>         Best Wishes,
>>         Spencer Graves
>>
>    I would go farther and nominate the sos package for inclusion as
> a recommended package (I don't have any conflict of interest in
> making the suggestion).  When I tell students about it many of them
> are puzzled that it's not part of the core R framework. I understand
> the arguments against migrating new functionality into core R, but
> this functionality is super-useful especially for beginners, and
> it seems worth it to lower the bar for finding help, at the cost
> of not a huge amount of extra code to maintain.


       Thanks.  For me, "sos" has provided the fastest literature search 
I've found for anything statistical (including using "installPackages" 
and "writeFindFn2xls", which could be overlooked by people using only 
"findFn").  Gone are the days when I'd puzzle for hours over a single 
line or page of mathematics.  If I don't understand something, I give 
examples to the code, and see what I get -- possibly using debug to 
trace what it does line by line.  If I don't find what I want with 
"sos", then I can look elsewhere.


       I only became aware of "http://metacran.github.io/search/", 
"http://www.rdocumentation.org/", and 
"http://www.bioconductor.org/packages/release/BiocViews.html" through 
responses earlier in this thread, so I can't compare "sos" to these 
other capabilities.


       Spencer Graves
>
>    Ben Bolker
>
> ______________________________________________
> R-devel at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel
>



More information about the R-devel mailing list