Raster R package, points outside the grid

Question

I am posting my issue in this community as I did not get satisfying answers in stackoverflow, hoping to get a solution here.

I am working on a spatial simulation and I created spatial points, but when I plot them with raster R package I can see that some points are outside the grid in the plot (example 1). Even when I try to define raster base on min and max values of XY coordinates(example 2).

I would like to know how I could be sure that all my points are inside the grid as I need to use polygon ID as localisation for the next steps of my simulations.

This is my code for reproducibility:

Example 1:

library(raster)
set.seed(100)
tmp = tmp2 = data.frame("x" = rnorm(1000, 10, 10),"y" = rnorm(1000, 10, 10))
coordinates(tmp) <- c("x", "y")
r1 <- raster(tmp, resolution = 10)
r1<-rasterToPolygons(r1)
plot(r1)
for (a in 1:nrow(tmp2)) {
  d = tmp2[a,]
  pts<- d
  coordinates(pts)<-~x+y
  plot(pts,add=T,col=a)
}

Example 2:

library(raster)
set.seed(100)
tmp = tmp2 = data.frame("x" = rnorm(1000, 10, 10),"y" = rnorm(1000, 10, 10))
coordinates(tmp) <- c("x", "y")

r1 <- raster(xmn = min(tmp2$x), xmx = max(tmp2$x), 
          ymn = min(tmp2$y), ymx = max(tmp2$y), 
          resolution = 10)

r1<-rasterToPolygons(r1)
plot(r1)
for (a in 1:nrow(tmp2)) {
  d = tmp2[a,]
  pts<- d
  coordinates(pts)<-~x+y
  plot(pts,add=T,col=a)
}

Answer 1

Since you set a custom resolution (10), the output raster doesn't have the same boundary box as the points layer, because the raster approximates to the closest value for defining the number of rows/columns:

(max(tmp2$y)-min(tmp2$y))/10
# 6.161062
# and your raster has 6 rows

You can add some extra space if you want to preserve the resolution or define a custom number of rows/columns, but this will create rectangular cells instead of squares.

Adding some extra space:

library(raster)
set.seed(100)
tmp = tmp2 = data.frame("x" = rnorm(1000, 10, 10),"y" = rnorm(1000, 10, 10))
coordinates(tmp) <- c("x", "y")
r1 <- raster(xmn = min(tmp2$x)-2.5, xmx = max(tmp2$x)+2.5, 
             ymn = min(tmp2$y)-2.5, ymx = max(tmp2$y)+2.5, 
             resolution = 10)

r1<-rasterToPolygons(r1)
plot(r1)
for (a in 1:nrow(tmp2)) {
  d = tmp2[a,]
  pts<- d
  coordinates(pts)<-~x+y
  plot(pts,add=T,col=a)
}