R help - Oct 2012 - Help with programming a tricky algorithm

Hi All,

I'm a little stumped by the following problem.  I've got a dataset with 
the following structure:

idxy    ix    iy    country    (other variables)
1        1    1    c1            x1
2        1    2    c1            x2
3        1    3    c1            x3
.        .        .       .            .

3739    55    67    c7        x3739
3740    55    68    c7        x3740

where ix and iy are interger-valued indices of the actual x and y 
coordinates for the gridded data

I want to define a "border" variable that equals 1 if the cell north, 
east, west, or south of it has a different value of the country 
variable.  So, for the row with idxy = 1, border would equal 1 if there 
is any idxy with country !=c1 and ix = 2 (or zero) or iy = 2 (or zero).

Any thoughts?

Thanks!
Andrew

Hello,

You should post a data example with ?dput. If your dataset is named 
MyData, use

dput( head(MyData, 30) )  # paste the output of this in a post

Anyway, I believe the following function might do what you want. It's 
untested, though. (Your example dataset is usefull but could be better)

is.border <- function(idx, DF){
     ix <- DF$ix %in% DF$ix[idx] + c(-1, 1)
     iy <- DF$iy %in% DF$iy[idx] + c(-1, 1)
     any(DF$country != DF$country[ix & iy])
}

sapply(MyData$idxy, fun, MyData)


It returns a logical value, so if you want 0/1 use as.integer to do the 
conversion.

Hope this helps,

Rui Barradas
Em 20-10-2012 08:36, Andrew Crane-Droesch escreveu:
> Hi All,
>
> I'm a little stumped by the following problem.  I've got a dataset 
> with the following structure:
>
> idxy    ix    iy    country    (other variables)
> 1        1    1    c1            x1
> 2        1    2    c1            x2
> 3        1    3    c1            x3
> .        .        .       .            .
>
> 3739    55    67    c7        x3739
> 3740    55    68    c7        x3740
>
> where ix and iy are interger-valued indices of the actual x and y 
> coordinates for the gridded data
>
> I want to define a "border" variable that equals 1 if the cell
north,
> east, west, or south of it has a different value of the country 
> variable.  So, for the row with idxy = 1, border would equal 1 if 
> there is any idxy with country !=c1 and ix = 2 (or zero) or iy = 2 (or 
> zero).
>
> Any thoughts?
>
> Thanks!
> Andrew
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide 
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.

HI,
I am not sure whether this is what you want.
Mydata<-read.table(text="
?idxy??? ix??? iy??? country??? col5 
?1??????? 1??? 1??? c1??????????? x1 
?2??????? 1??? 2??? c1??????????? x2? 
?3??????? 1??? 3??? c1??????????? x3?? 
?4??????? 2??? 4??? c1??????????? x4 
?5??????? 2??? 4??? c2??????????? x5? 
?6??????? 2??? 5??? c2??????????? x6 
?7??????? 3??? 5??? c3??????????? x7??? 
?8??????? 3??? 5??? c3??????????? x8? 
?9??????? 3??? 5??? c3??????????? x9 
",sep="",stringsAsFactors=FALSE,header=TRUE)
country1<-as.numeric(gsub(".*(\\d+)","\\1",Mydata$country))
?Mydata1<-data.frame(ix=abs(diff(c(Mydata$ix[2],Mydata$ix))),iy=abs(diff(c(Mydata$iy[2],Mydata$iy))),country=abs(diff(c(country1[2],country1))))
?Mydata2<- Mydata[apply(Mydata1,1,function(x)
all(!duplicated(x)|!duplicated(x,fromLast=TRUE))),]
?Mydata2$border<-1
res<-merge(Mydata,Mydata2,by.x=c("idxy","ix","iy","country","col5"),by.y=c("idxy","ix","iy","country","col5"),all.x=TRUE)
?res
#? idxy ix iy country col5 border
#1??? 1? 1? 1????? c1?? x1????? 1
#2??? 2? 1? 2????? c1?? x2????? 1
#3??? 3? 1? 3????? c1?? x3????? 1
#4??? 4? 2? 4????? c1?? x4????? 1
#5??? 5? 2? 4????? c2?? x5????? 1
#6??? 6? 2? 5????? c2?? x6????? 1
#7??? 7? 3? 5????? c3?? x7????? 1
#8??? 8? 3? 5????? c3?? x8???? NA
#9??? 9? 3? 5????? c3?? x9???? NA





----- Original Message -----
From: Andrew Crane-Droesch <andrewcd at gmail.com>
To: r-help at r-project.org
Cc: 
Sent: Saturday, October 20, 2012 3:36 AM
Subject: [R] Help with programming a tricky algorithm

Hi All,

I'm a little stumped by the following problem.? I've got a dataset with
the following structure:

idxy? ? ix? ? iy? ? country? ? (other variables)
1? ? ? ? 1? ? 1? ? c1? ? ? ? ? ? x1
2? ? ? ? 1? ? 2? ? c1? ? ? ? ? ? x2
3? ? ? ? 1? ? 3? ? c1? ? ? ? ? ? x3
.? ? ? ? .? ? ? ? .? ? ?  .? ? ? ? ? ? .

3739? ? 55? ? 67? ? c7? ? ? ? x3739
3740? ? 55? ? 68? ? c7? ? ? ? x3740

where ix and iy are interger-valued indices of the actual x and y coordinates
for the gridded data

I want to define a "border" variable that equals 1 if the cell north,
east, west, or south of it has a different value of the country variable.? So,
for the row with idxy = 1, border would equal 1 if there is any idxy with
country !=c1 and ix = 2 (or zero) or iy = 2 (or zero).

Any thoughts?

Thanks!
Andrew

______________________________________________
R-help at r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.