R/initialZoning.R

In geozoning: Zoning Methods for Spatial Data

##################################################################
#' initialZoning zoning with no correction
#'
#' @details calculates a zoning on kriged map data, based on quantiles of data attribute values. These quantiles correspond the given probabilities. Contour lines define zones, they are projected on to the map boundary in order to close them if necessary. The distance matrices intra and inter zones are calculated as well as the quality criterion.
#' @param qProb probability vector used to generate quantile values
#' @param map object returned by function genMap
#' @param pErr equality tolerance for distance calculations
#' @param simplitol tolerance for spatial polygons geometry simplification
#' @param optiCrit criterion choice
#' @param disp 0: no info, 1: some info, 2: detailed info
#' @param GridData logical value indicating if data are already on a regular grid (no kriging in that case)
#'
#' @return a list with components
#' \describe{
#' \item{resCrit}{criterion value}
#' \item{resDist}{list with components matDistance, matDistanceCorr and cost, such as returned by a call to calDistance}
#' \item{resZ}{list with components zoneN, zoneNModif, listZonePoint, meanTot, meanZone,listSurf, critSurf, zonePolygone, such as the object returned by calNei}
#' }
#'
#' @export
#'
#' @examples
#' data(mapTest)
#' ZK=initialZoning(qProb=c(0.4,0.7),mapTest)
#' plotZ(ZK$resZ$zonePolygone)
initialZoning=function(qProb, map, pErr=0.9,simplitol=1e-3,optiCrit=2,disp=0,GridData=F)
##################################################################
  {

  qProb=as.numeric(qProb)
  Z=zoneGeneration(map,qProb,GridData) #in funcZoning
  if(is.null(Z)) return(NULL) # no zones
  #
  # create comments (for holes)
  Z = crComment(Z)
  # add IDs to identify zones in zoning
  Z = setIds(Z)
  # compute neighborhood
  K=calNei(Z,map$krigData,map$krigSurfVoronoi,map$krigN,simplitol)
  # assign zone labels
  K= labZone(K,qProb,map$krigGrid)
 # per label computations - compute overall cost per label
  cL=costLab(K,map)
#
  Z=K$zonePolygone
  # reset ids
  Z = setIds(Z)
  K$zonePolygone=Z
#
#
    tabVal=map$krigData
    listZonePoint=K$listZonePoint
    zoneN=K$zoneN
    surfVoronoi=map$krigSurfVoronoi
    meanTot=K$meanTot
    meanZone=K$meanZone

    # compute distances and criteria
  resDist=calDistance(typedist=1,tabVal,listZonePoint,zoneN,surfVoronoi,meanZone,pErr)
    # compute only on neighbor zones
  resCrit=calCrit(resDist$matDistanceCorr,K$zoneNModif,optiCrit)

   if(disp==2) dispZ(map$step,map$krigGrid,zonePolygone=Z,K=K,boundary=map$boundary,nbLvl=0,mu=2)

  return(list(resCrit=resCrit,resDist=resDist,resZ=K,cL=cL,qProb=qProb))
}