8

I have a SpatialLinesDataFrame with streets and I have a list of GPS coordinates. What I need to do is to get out the 10 closest street names for each individual GPS coordinate.

Is there a function/package in R that would calculate the distance between a line and a point for a SpatialLinesDataFrame? I can't see anything that would help in 'sp'.

There is a related question: Calculating the distance between polygon and point in R, but I want to find the distance between a line object and a point, not polygon/point, point/point.

Community
  • 1
  • 1
Geza
  • 464
  • 6
  • 16
  • Search: http://stackoverflow.com/questions/15294343/calculating-the-distance-between-polygon-and-point-in-r – IRTFM Oct 01 '13 at 23:14
  • The linked answer is useful, but only shows how to calculate the distance between one point and one polygon. – Josh O'Brien Oct 02 '13 at 01:17

1 Answers1

14

You could use rgeos::gDistance() with byid=TRUE to get a matrix of distances from each point to each line. From there, it's relatively easy to extract the ids of the 10 lines nearest to each point:

library(sp)
library(rgeos)

## Create a SpatialPoints and a SpatialLines object
example("SpatialPoints-class", ask=FALSE, echo=FALSE)
example("SpatialLines-class", ask=FALSE, echo=FALSE)

## Compute the matrix of distances between them.
(m <- gDistance(S, Sl, byid=TRUE))
#          1   2        3        4        5
# a 0.000000 0.0 2.757716 1.414214 2.757716
# b 1.788854 0.5 3.640055 1.000000 3.605551

## And then use it to extract the ids of the 10 (or in this case 1) lines
## closest to each point.
## apply(m, 2, function(X) rownames(m)[order(X)][1:10]) ## Finds 10 closest
apply(m, 2, function(X) rownames(m)[order(X)][1])       ## Finds single closest
#   1   2   3   4   5 
# "a" "a" "a" "b" "a"
Josh O'Brien
  • 159,210
  • 26
  • 366
  • 455