29.01.2015 Views

Analysing spatial point patterns in R - CSIRO

Analysing spatial point patterns in R - CSIRO

Analysing spatial point patterns in R - CSIRO

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

32.1 Numeric marks: distribution and trend 201<br />

To assess <strong>spatial</strong> trend <strong>in</strong> the marks, one way is to form a kernel regression smoother. The<br />

smoothed mark value at location u ∈ R 2 is<br />

∑<br />

i<br />

̂m(u) =<br />

m iκ(u − x i )<br />

∑<br />

i κ(u − x i)<br />

where k is the smooth<strong>in</strong>g kernel, and m i is the mark value at data <strong>po<strong>in</strong>t</strong> x i . This is computed<br />

by smooth.ppp:<br />

> plot(smooth.ppp(longleaf))<br />

smooth.ppp(longleaf)<br />

15 20 25 30 35 40<br />

The plot shows that there is a region of younger trees <strong>in</strong> the northeast of the study region.<br />

If the marks are a data frame, the result of smooth.ppp will be a list of pixel images, one for<br />

each mark variable.<br />

You can also use cut.ppp followed by split.ppp to look for <strong>spatial</strong> <strong>in</strong>homogeneity of the<br />

marks:<br />

> data(spruces)<br />

> plot(split(cut(spruces, breaks = 3)))<br />

split(cut(spruces, breaks = 3))<br />

(0.16,0.23] (0.23,0.3] (0.3,0.37]<br />

Other facilities <strong>in</strong>clude markvar which calculates a smoothed estimate of the local variance<br />

of the mark values.<br />

Copyright<strong>CSIRO</strong> 2010

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!