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

[Rolf Turner]

On 29/04/20 10:07 am, Abby Spurdle wrote:

> I haven't attempted this.
> (Mainly because I'm not familiar with the theory surrounding it).
> 
> However, I looked at the documentation for the spatstat package.
> There are are several functions prefixed with pcf, including one named pcf3est.
> According to its description field:
> 
>            Estimates the pair correlation function
>            from a three-dimensional point pattern.
> 
> *If* it does what it claims ...

Why would you doubt that it does what it claims?

> ... would that solve your problem?
> 
> Note (to spatstat authors):
> 
> I'm not convinced this package is well documented.
> In fact, I'm not even convinced it meets CRAN standards, which require
> functions to have their arguments documented.
> 
>       X
>      Three-dimensional point pattern (object of class "pp3").
> 
> Nowhere in the help page, does it say what a pp3 object is, or how to
> create it, or where to find that information.
> Potentially requiring a user to search through a 1766 page document
> for the answer.
> (Yes, I know there's a function named pp3, but I don't think that's
> good enough).
> 
> If people are not going to document their packages properly, they
> could try a little bit harder to answer R-help questions that involve
> their packages...

<SNIP>

Wouldn't the first thing that one would try be:

   ??"pp3"

The required information is then immediately apparent.

Moreover if one takes the trouble to look at the examples, one is led to 
the function rpoispp3() which points to the function pp3().

This reminds not a little of fortunes::fortune(9).

Of course I'm biased, but IMHO spatstat is documented not only 
"properly", but superbly well! :-)

cheers,

Rolf

-- 
Honorary Research Fellow
Department of Statistics
University of Auckland
Phone: +64-9-373-7599 ext. 88276