[R] [FORGED] Re: pair correlation function of 3D points

[Ege Rubak]

Hi Abby,

Once again I must say your form of communication puzzles me. Do you
believe it benefits the open source community to communicate like this?
What about contacting the package maintainer in this way:

"Hi, I was helping someone else using your package to estimate a pair
correlation function. I was really puzzled that the generel help page
for `pcf` said it was a generic, but it doesn't seem to be generic for
the three dimensional case we needed, and no reference was given from
`pcf` to `pcf3est` which I had to search around to find. Maybe you
should consider throwing in a few cross references. Have a nice day."

I think this all would have played out much nicer then. None of us were
ignoring the original poster. We simply weren't aware of the OPs
problem until Rolf discovered it. I think Adrian and I focus on more
spatstat devoted channels to maximize the benefit of our time.

Regarding splitting the package into smaller pieces the thought has
occurred to us without your helpful suggestion. We started the process
of doing so years ago with the introduction of `spatstat.utils` and
`spatstat.data`. This is well documented on the spatstat GitHub page
referenced in the DESCRIPTION file. However, we have to prioritize our
time and it is quite a complex task to split out the relevant parts in
standalone packages due to all the interdependencies of the existing
code.

On a final note, to give you another example of a way to bring
attention to possible improvements of documentation: Let's say I was
looking at the function `marginal.set` in the package `probhat`, then I
would contact the author and say:

"Hi Abby, I was looking at the help for `marginal.set` and I couldn't
quite figure out what the argument constructor was supposed to be. The
help doesn't mention any class and contains no cross references
whatsoever and searching for 'constructor' in the R help system didn't
lead anywhere. Maybe a few cross references and actual usage of the
section 'See Also' would be beneficial for new users of your package.
It would be much better if instead of making assumptions about
assumptions you just use hyperlinks. Have a nice day. Kind regards,
Ege"

Where of course I would never use the sentence about hyperlinks in a
real example. I guess at this point I dug myself into the hole of bad
communication which is so killing for good software projects so
apologies for that. I do wish you a good day and hope you will consider
helping us all improve the state of open source software by
constructive suggestions rather than us claiming that each others
packages are unfit for CRAN.

Kind regards,
Ege

On Thu, 2020-04-30 at 06:53 +1200, Abby Spurdle wrote:
> NOTE CITATIONS USE BRIEF EXCERPTS
> 
> > It's a complaint that no-one responded to your query within 24
> > hours.
> 
> Correction, it wasn't my query.
> I was replying to someone else's query.
> 
> > Finally, you cast doubt on whether the function pcf3est actually
> > does calculate...
> 
> You've taken what I said out of context.
> As I said in my response to Rolf, I didn't run the code.
> 
> > email from CRAN requesting me to please cut down the size of the
> > spatstat package
> 
> I can't speak on behalf on CRAN.
> But reiterating what I said earlier, you'd be better to create
> smaller
> packages, each with a more specific focus.
> Each package could have an unofficial co-maintainer to help you out.
> 
> > You complain that the documentation is insufficient, but at the
> > same time, you complain that the manual is too long (> 1700 pages)
> > and you seem unwilling to search the documentation or follow cross-
> > references.
> 
> Did you read my posts...
> 
> The functions DO NOT cross-reference.
> The pcf generic/methods DO NOT reference the pcf3est function, at
> all.
> The pcf3est function DOES NOT reference the pp3 function, but rather
> references the class of the object.
> 
> Also, the seealso sections are minimalist.
> 
> If you are not going to create smaller packages, then the least you
> could do it improve the standard of the seealso sections and
> cross-referencing.
> 
> Furthermore, if functions are not cross-referenced, the reader (I
> this
> case me) doesn't necessarily know what package those functions are
> defined in.
> You assume that readers will assume the functions are defined in your
> package, and then do a search for them.
> 
> Instead of making assumptions about assumptions, just use
> hyperlinks...
> 
> > > I was somewhat disappointed ... not one (after approx 24 hours)
> > > had tried to
> > > help answer the OP's question.
> > 
> > Seriously?
> 
> Yes, seriously.
> From my experiences, if people don't reply to an R-help question in
> the first 24 hours, chances that they will reply later are low.
> Also, I've found a bias on this forum, with few replies to
> physics-related questions.
> 
> Personally, I like physicists, one of the most under-valued
> professions on the planet.
> 
> > I found the pcf function via Googling the subject.
> > I don't want to be rude, but this seems pretty lazy.
> > Try Googling  'spatstat pair correlation function three dimensions'
> 
> That sounds like a contradiction.
> 
> Also, it's off the mark, given that I was the only one who attempted
> to answer the OPs question.
-- 
Ege Rubak, Associate Professor,
Department of Mathematical Sciences, Aalborg University
Skjernvej 4A, 9220 Aalborg East, Denmark
Phone: (+45)99408861
Mobile: (+45)30230252
Email: rubak using math.aau.dk