Group cases into neighborhoods based on walking distance.
neighborhoodWalking(pump.select = NULL, vestry = FALSE, weighted = TRUE,
case.set = "observed", multi.core = FALSE)
Numeric. Default is NULL: all pumps are used. Otherwise, selection by a vector of numeric IDs: 1 to 13 for pumps
; 1 to 14 for pumps.vestry
. Exclusion (negative selection) is possible (e.g., -6). Note that you can't just select the pump on Adam and Eve Court (#2): it's a technical isolate.
Logical. TRUE uses the 14 pumps from the Vestry Report. FALSE uses the 13 in the original map.
Logical. TRUE computes shortest path weighted by road length. FALSE computes shortest path in terms of the number of nodes.
Character. "observed", "expected" or "snow". "snow" captures John Snow's annotation of the Broad Street pump neighborhood printed in the Vestry report version of the map.
Logical or Numeric. TRUE uses parallel::detectCores(). FALSE uses one, single core. You can also specify the number logical cores. On Window, only "multi.core = FALSE" is available.
An R list with 7 objects:
paths
: list of paths to nearest or selected pump(s).
cases
: list of cases by pump.
vestry
: "vestry" from neighborhoodWalking().
observed
: "observed" from neighborhoodWalking().
pump.select
: "pump.select" from neighborhoodWalking().
cores
: number of cores to use for parallel implementation.
metric
: incremental metric used to find cut point on split road segments.
# NOT RUN {
# neighborhoodWalking()
# neighborhoodWalking(pump.select = -6)
# }
Run the code above in your browser using DataLab