[R-sig-Geo] Fit a sine curve to raster stack

[Verbesselt, Jan]

Dear Sofea,

You could try something like this:

fun3 <- function(x){
        time <- 1:length(x)
	xc <-cos(2*pi*t/365)
	xs <-sin(2*pi*t/365)
	fit <- lm(x~xc+xs)
        return(fit$coefficients[1])
        }

s3 <- calc(s,  fun = fun3)


But check for NA's within your time series as these might trigger
errors/warnings. Maybe also check na.action options like na.approx(). Can
you provide a reproducible example in that case?

Best,
Jan Verbesselt
Wageningen University
The Netherlands



On 31/10/13 21:39, "Sofea Wright" <sofeawright at gmail.com> wrote:

>Dear list,
>
>I am trying to fit a sine curve to my data. When dealing with non-raster
>data, it's quite straight forward:
>
># Fit sine curve to describe the data
>temperature<-maxtemp$temperature
>time<-1:n
>xc<-cos(2*pi*time/365)
>xs<-sin(2*pi*time/365)
>fit<-lm(temperature~xc+xs)
>summary(fit)
>
>I am not sure how to do that with rasterstack as I keep on getting error
>message.
> > s
>class       : RasterStack
>dimensions  : 1000, 1621, 1621000, 365  (nrow, ncol, ncell, nlayers)
>resolution  : 0.04166667, 0.04166667  (x, y)
>extent      : 96.5, 165.5, -16.5, 26.5  (xmin, xmax, ymin, ymax)
>coord. ref. : +proj=longlat +datum=WGS84 +no_defs +ellps=WGS84
>+towgs84=0,0,0
>
>s <- stack(tif)
>time <- 1:nlayers(s)
>fun1 <- function(x){
>  xc<-cos(2*pi*time/365)
>  xs<-sin(2*pi*time/365)
>          }
>
>fun2 <- function (x) {
>        lm(s~xc+xs)
>        }
>Error in .calcTest(x[1:5], fun, na.rm, forcefun, forceapply) :
>  cannot use this function
>
>s1 <- calc(s,  fun = fun1)
>s2 <- calc(s,  fun = fun2)
>
>Anybody?
>
>Thank you
>Sofea
>
>	[[alternative HTML version deleted]]
>
>_______________________________________________
>R-sig-Geo mailing list
>R-sig-Geo at r-project.org
>https://stat.ethz.ch/mailman/listinfo/r-sig-geo