Learn R Programming

spatstat.explore (version 3.3-1)

plot.scan.test: Plot Result of Scan Test

Description

Computes or plots an image showing the likelihood ratio test statistic for the scan test, or the optimal circle radius.

Usage

# S3 method for scan.test
plot(x, ..., what=c("statistic", "radius"),
               do.window = TRUE)

# S3 method for scan.test as.im(X, ..., what=c("statistic", "radius"))

Value

The value of as.im.scan.test is a pixel image (object of class "im"). The value of plot.scan.test is NULL.

Arguments

x,X

Result of a scan test. An object of class "scan.test" produced by scan.test.

...

Arguments passed to plot.im to control the appearance of the plot.

what

Character string indicating whether to produce an image of the (profile) likelihood ratio test statistic (what="statistic", the default) or an image of the optimal value of circle radius (what="radius").

do.window

Logical value indicating whether to plot the original window of the data as well.

Author

Adrian Baddeley Adrian.Baddeley@curtin.edu.au and Rolf Turner rolfturner@posteo.net

Details

These functions extract, and plot, the spatially-varying value of the likelihood ratio test statistic which forms the basis of the scan test.

If the test result X was based on circles of the same radius r, then as.im(X) is a pixel image of the likelihood ratio test statistic as a function of the position of the centre of the circle.

If the test result X was based on circles of several different radii r, then as.im(X) is a pixel image of the profile (maximum value over all radii r) likelihood ratio test statistic as a function of the position of the centre of the circle, and as.im(X, what="radius") is a pixel image giving for each location \(u\) the value of r which maximised the likelihood ratio test statistic at that location.

The plot method plots the corresponding image.

See Also

scan.test, scanLRTS

Examples

Run this code
   online <- interactive()
   Nsim <- if(online) 19 else 2
   r <- if(online) seq(0.04, 0.1, by=0.01) else c(0.05, 0.1)
   a <- scan.test(redwood, r=r, method="poisson", nsim=Nsim)
   plot(a)
   as.im(a)
   plot(a, what="radius")

Run the code above in your browser using DataLab