How to control for the amount of noise added to x, y spatial points in R?

Question

Fairly novice R user, and I'm sure there's an easy solution for this - but I can't find it. I have a data frame with a series of spatial coordinates, as well as a host of other attributes. Many of the spatial coordinates are the exact same - and I would like to add a set amount of noise to them, so that I can keep them within a certain radius - in this case 0.4 meters, or 40 centimeters while also keeping track of their associated attributes.

I am essentially looking for an R version of this question: https://gis.stackexchange.com/questions/35479/adding-noise-to-overlapping-x-y-coordinates-so-no-longer-in-exact-same-place

... because when I follow the directions for the answer to this question using ArcGIS - I get a series of random points but I lose the attributes associated with those points, and I can't easily find a way to match them back up.

Is there a way to use the jitter function in R or something similar and specify the radius in a spatial context (e.g. 40 cm) that the spatial coordinates are randomly distributed within that range? I don't understand how to manipulate the factor and amount argument in order to get my desired output.

Edit:

Here's an example df with fake coordinates. As you can see the coordinates in the first and third column are the same, because the animal was under the same rock twice. I would like to be able to add jitter to these coordinates so they are slightly different, but I want to control the jitter to be within 40 centimeters (no more than the size of the rock)

mydf <- data.frame("point_id" = 1:6, "date_time" = c("6/5/2018 10:57","6/5/2018 14:30","6/6/2018 10:06","6/6/2018 11:06","6/7/2018 10:35","6/7/2018 15:50"), "Animal_ID" = c(4,5,4,5,4,6), "Rock_ID" = c(1,2,1,3,4,5), x_proj = c(831120.3759,831441.0415,831120.3759,831433.4414,831128.4778,831422.0822), y_proj = c(5877582.998,5875337.074,5877582.998,5875328.897,5877575.360,5875338.216))

#make a separate object for the coordinates#
xy <- mydf[,c(5,6)]

#Convert to a spatialpoints data frame (insert own epsg)
sp.mydf <- SpatialPointsDataFrame(coords = xy, data = tumbling_test, proj4string = CRS("+init=epsg:xxxxx"))

I want my newly generated coordinates to still contain the attribute data from other columns (i.e. Animal_ID, date, etc) because other methods I have used in ArcGIS generate a series of new random pts but I can't match them back to the attributes.

Also if there is a way to only add the jitter to points that have more than one occurrence at a rock, that would be neat. For example here I only need to add the jitter to rows 1 and 3, because the other coordinates do not repeat. Once the jitter is added, I want to convert the results back to a regular dataframe that I will export to a .csv

Answer 1

There is, and it is much easier if you're working with an sf object.

You'll need to dig into coordinate reference systems (CRS) a little as well in order to jitter the points the correct distance.

If you're starting with a spatialPointsdataframe, use st_as_sf() to return an sf object.

Below is a reproducible example with points jittered around 5km. The jittering is somewhat random and in this example varies from ~2-5.5 kilometers.

library(sf)
#> Linking to GEOS 3.6.2, GDAL 2.2.3, PROJ 4.9.3
library(tidyverse)

# load example data
nc <- read_sf(system.file('gpkg/nc.gpkg', package = 'sf'))

#make single points from polygons
nc_points <- st_centroid(nc)
#> Warning in st_centroid.sf(nc): st_centroid assumes attributes are constant over
#> geometries of x
#> Warning in st_centroid.sfc(st_geometry(x), of_largest_polygon =
#> of_largest_polygon): st_centroid does not give correct centroids for longitude/
#> latitude data

# Transform to a crs that uses meters as the distance
nc_points <- st_transform(nc_points, 3358)

nc_points_jittered <- st_jitter(nc_points, amount = 5000)

p1 <- ggplot(nc_points) + 
  geom_sf() + 
  ggtitle('Original')

p2 <- ggplot(nc_points_jittered) + 
  geom_sf() + 
  ggtitle('Jittered')

p3 <- ggplot() + 
  geom_sf(data = nc_points, color = 'red') + 
  geom_sf(data = nc_points_jittered, color = 'black') + 
  ggtitle('Both')

cowplot::plot_grid(p1, p2, p3, ncol = 1)

^{Created on 2020-01-12 by the reprex package (v0.3.0)}