R: applying a function on whole dataset to find points within a circle - r

Question

I have a difficulty with application of the data frame on my function in R. I have a data.frame with three columns ID of a point, its location on x axis and its location on y axis. All I need to do is to find for a given point IDs of points that lies in its neighborhood. I've made the function that shows whether the point lies within a circle where the center is a location of observed point and returns it's ID if true.
Here is my code:
point_id <- locationdata$point_id
x_loc <- locationdata$x_loc
y_loc <- locationdata$y_loc
locdata <- data.frame(point_id, x_loc, y_loc)
#radius set to1km
incircle3 <- function(x_loc, y_loc, center_x, center_y, pointid, r = 1000000){
dx = (x_loc-center_x)
dy = (y_loc-center_y)
if (b <- dx^2 + dy^2 <= r^2){
print(shopid)} ##else {print('')}
}
Unfortunately I don't know how to apply this function on the whole data frame. So once I enter the locations of the observed point it would return me IDs of all points that lies in the neighborhood. Ideally I would need to find this relation for all the points automatically. So it would return me the points that lies in the neighborhood of each point from the dataset. Previously I have been inserting the center_x and center_y manually.
Thank you very much for your advices in advance!

Answer 1

You can tackle this with R's dist function:
# set the random seed and create some dummy data
set.seed(101)
dummy <- data.frame(id=1:100, x=runif(100), y=runif(100))
> head(dummy)
id x y
1 1 0.37219838 0.12501937
2 2 0.04382482 0.02332669
3 3 0.70968402 0.39186128
4 4 0.65769040 0.85959857
5 5 0.24985572 0.71833452
6 6 0.30005483 0.33939503
Call the dist function which returns a dist object. The default distance metric is Euclidean which is what you have coded in your question.
dists <- dist(dummy[,2:3])
Loop over the distance matrix and return the indices for each id that are within some constant distance:
neighbors <- apply(as.matrix(dists), 1, function(x) which(x < 0.33))
> neighbors[[1]]
1 6 7 8 19 23 30 32 33 34 42 44 46 51 55 87 88 91 94 99
Here's a modification to handle volatile ids:
set.seed(101)
dummy <- data.frame(id=sample(1:100, 100), x=runif(100), y=runif(100))
> head(dummy)
id x y
1 38 0.12501937 0.60567568
2 5 0.02332669 0.56259740
3 70 0.39186128 0.27685556
4 64 0.85959857 0.22614243
5 24 0.71833452 0.98355758
6 29 0.33939503 0.09838715
dists <- dist(dummy[,2:3])
neighbors <- apply(as.matrix(dists), 1, function(x) {
dummy$id[which(x < 0.33)]
})
names(neighbors) <- dummy$id
> neighbors[['38']]
[1] 38 5 55 80 63 76 17 71 47 11 88 13 41 21 36 31 73 61 99 59 39 89 94 12 18 3