envelopes of simulations

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view

envelopes of simulations

Adrian Baddeley-2

Sara Mouro writes:
 > Hello!
 > I am writing you because I could not plot the confidence envelopes for
 > functions Jest, Jcross, Jdot, Jmulti, and L, using the Spatstat package.

Enquiries about a package should be sent to the package maintainer
rather than R-help.

 > I have already understood how to do that for Kest or Jest, that is:
 > JEnv <- plot(envelope(PPPData, Jest))
 > Where PPPData is my ppp object.
 > However, for Jcross I must specify the two marks I want to analyse.
 > That is, usually I would get the Jcross doing:
 > Jc <- Jcross(PPPData,"Aun","Qsu")
 > For marks "Aun" and "Qsu".

To do this, type
      envelope(PPPData, Jcross, i="Aun", j="Qsu")

        Looking at help(Jcross) we see that its formal syntax is
                Jcross(X, i, j).
        This means that when you call
        the arguments are matched as

        So our problem is, how to pass the arguments i="Aun" and j="Qsu"
        to the function Jcross each time it is called by `envelope'.

        The help entry for 'envelope' mentions that `envelope' will accept
        extra arguments "..." which are then passed to the function 'fun'.
        So, to pass the two argument values i="Aun" and j="Qsu",
        just type
             envelope(PPPData, Jcross, i="Aun", j="Qsu")

 > For L function, I can make:
 > K <- Kest (PPPData, correction="isotropic")
 > plot (K, r-sqrt(iso/pi)~r)

        You can compute the envelope for the K-function
        and transform it afterwards.

        To plot envelopes of the L function discrepancy
           LD(r) = sqrt(K(r)/pi) - r,
        just type
                E <- envelope(PPPData, Kest, correction="isotropic")
                plot(E, sqrt(./pi) -r ~ r)

Hope this helps.
Adrian Baddeley

[hidden email] mailing list
PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html