[R-sig-Geo] Breakpoint analysis with two variables

[nate_m]

Hi Robert,

Thanks for taking the time to provide suggestions. Below is what I've worked
out, and I can now retrieve variables from lm (in this case the two
coefficients), but can't figure out how to save an entire "lm object" for
each cell to feed into the "segmented" portion of my script. So basically,
I'm looking to somehow save an entire "lm object" in raster brick form to
then utilize in the next step (where library(segmented) begins). Any ideas?

library(raster)
x <- raster(nrow=10, ncol=10)
y <- raster(nrow=10, ncol=10)
r <- stack( sapply(1:11, function(i) setValues(x, rnorm(ncell(x), i, 3) )) )
s <- stack( sapply(1:11, function(i) setValues(y, rnorm(ncell(y), i, 3) )) )
z <- stack(r,s)
z[1] <- NA
fun <- function(x) { if (is.na(x)) { return(cbind(NA,NA)) } else lm(x[1:11]
~ x[12:22])$coefficients }

g2 <- calc(z, fun)
g2

library(segmented)
min_x <- calc(r, min)
fun2 <- function(x) { if (is.na(x)) { return(cbind(NA,NA,NA,NA)) } else
segmented(x, ~s, psi = min_x)$coefficients }

seg_g2 <- calc(g2, fun2)



--
View this message in context: http://r-sig-geo.2731867.n2.nabble.com/Breakpoint-analysis-with-two-variables-tp7585607p7585654.html
Sent from the R-sig-geo mailing list archive at Nabble.com.