Utility to create efficient quadtree spatial index of any set of lat,lon

Index a list of points (e.g., schools) so getpointsnearby() can find them very quickly

Usage

indexpoints(pts, indexname = "custom_index", envir = globalenv())

Arguments

pts: a data.frame or data.table with columns name lat and lon, one row per location (point), and any other columns are ignored.
indexname: optional name to give the index
envir: optional environment - default is to assign index object to globalenv()

Value

Just returns TRUE when done. Side effect is to put into the globalenv() (or specified envir) that spatial index with name defined by indexname, as created by indexpoints().

Details

This creates a spatial index to be used by getpointsnearby() to support proxistat(), to create a new proximity score for every block and block group in the US. It relies on create_quaddata() for one step, then SearchTrees::createTree()

Examples

# \donttest{
  # EXAMPLES NOT TESTED YET ***
  pts <- testpoints_10
  tempenv <- new.env()
  index10 <- indexpoints(pts, "index10", envir = tempenv)
#> cleaning/checking latitude and longitude values 
#> Building index of  10 points, to be called  index10  and loaded in specified environment e.g., globalenv() 
#> Warning: only supports idcolname NULL, id, or blockid - ignoring the name provided and creating unique id column called pointid
#> creating quaddata format table before indexing
#>   Done building index of  10 points. 
  x <- getpointsnearby(pts, quadtree = get(index10, envir = tempenv))
#> Error in getpointsnearby(pts, quadtree = get(index10, envir = tempenv)): must provide topoints
  # y <- proxistat(pts)
  # rm(custom_index)
 # }