[R-pkg-devel] Appropriate keyword in help file.

Uwe Ligges ||gge@ @end|ng |rom @t@t|@t|k@tu-dortmund@de
Fri Apr 17 10:41:24 CEST 2020

keywords are not mandatory.

You can invent your own keywords and use them in


Uwe Ligges

On 17.04.2020 02:57, Jeff Newmiller wrote:
> You can get away with a lot if you are not distributing your package. But I usually try to satisfy R CMD check at least.
> On April 16, 2020 5:50:04 PM PDT, Rolf Turner <r.turner using auckland.ac.nz> wrote:
>> On 17/04/20 12:14 pm, Duncan Murdoch wrote:
>>> On 16/04/2020 8:12 p.m., Rolf Turner wrote:
>>>> I'm writing a package (just for my own use, for the time being at
>> least)
>>>> that contains a function for estimating the parameters of a
>>>> distribution.  The function is essentially a wrapper for fitdistr()
>> from
>>>> the MASS package.
>>>> When I looked at RShowDoc("KEYWORDS"), I could not find an
>> appropriate
>>>> keyword to use in the help file for this function.  There's "htest &
>>>> Statistical Inference", but that's not right, since this function is
>>>> about estimation, not hypothesis testing.  I'd hoped that there'd be
>> a
>>>> keyword "estimation" (or "point estimation") or something like that,
>> but
>>>> there isn't.
>>>> I guess I can use "utilities" (???) or "misc", but these seem a bit
>>>> unsatisfactory.
>>>> Can anyone suggest a better idea?
>>> Don't use any keyword. When was the last time you searched on one?
>> I have *never* in my life searched on a keyword! :-)
>> I just thought that it was an Immutable Law of the Universe that a help
>> file had to have at least one keyword, and that the R Gods would rain
>> fire and brimstone down upon one's head if no keyword was supplied.
>> In a similar vein:  would the R Gods allow me to make up my own keyword
>> rather than choosing from those listed by RShowDoc("KEYWORDS")?
>> Thanks.
>> cheers,
>> Rolf

More information about the R-package-devel mailing list