complex functions using aggregate

Ben, perhaps this will do it:

fun <- function(x) { sum(abs(x[-length(x)] - x[-1])) }

agg<-aggregate(Elevp$Elevation,by=list(Elevp$From,Elevp$To), fun)

Robert


On Wed, Sep 14, 2011 at 4:17 PM, Ben Weinstein
<benweinstein2010 at gmail.com>wrote:

Hi all

I have a series of elevation transects between a list of pairwise
locations.
This is not specifically a spatial question, but since it is a common
spatial task, i hope its okay.

head(Elevp)
Elevation ComboID From  To
1  3408.080       1  262 138
2  3264.469       1  262 138
3  2904.685       1  262 138
4  2752.840       1  262 138
5  3408.080       2  262 129
6  3217.581       2  262 129

In the small section above, the first 'transect' has four points, and goes
from site 262 to site 138. I'm trying to find a way to get the total
elevation change over each transect. An abstract example is shown below.

h

[1]  10   5  40  30 100  60

g<-numeric()
for(x in 1:(length(h)-1))

+ g[x]<-((abs(h[x+1]-h[x])))

g

[1]  5 35 10 70 40

sum(g)

[1] 160

to perform this over the entire database (1000's of comparisons)

agg<-aggregate(Elevp$Elevation,by=list(Elevp$From,Elevp$To),function(x)

how do i supply the former function, as a product that fits this setup?
Does a cumulative difference function exist? I'm not sure what to call it
to
begin searching.

I appreciate the help.

Ben Weinstein
Ecology and Evolution
Stony Brook University

       [[alternative HTML version deleted]]

	[[alternative HTML version deleted]]