interpolationError: Interpolate many maps at once, compare them to the original...
In sensors4plumes: Test and Optimise Sampling Designs Based on Plume Simulations

Description Usage Arguments Value Author(s) Examples

This function calls interpolate, then it compares the result plume-and-location-wise to the original and summarises the resulting error values.

interpolationError(simulations, locations, kinds, 
  fun_interpolation = NA, fun_error = NA, 
  fun_Rpl = NA, fun_Rpl_cellStats = "mean", fun_l = NA, 
  tmpfile = "tmp_interpolationError", overwrite = FALSE, chunksize = 1e+7)

`simulations`	`Simulations`
`locations`	indices of `locations` to be used as input; multiple and invalid values are ignored
`kinds`	layer of the `values` of `simulations` to be used; interpolation can only be applied to one layer: if `kinds` is a vector, only the first entry is used
`fun_interpolation`	interpolation `function`, must have the parameters `y`, `data`, `newdata` (form as for `krige0`). All other parameters need default values, e.g. the `model` of `krige0` that can be set by `replaceDefault` with `type = "interpolation_fun.interpolate"`.

`fun_error`	function to compare original and interpolated map location-and-plume-wise; must have a parameter `x`, then `x[1]` is the original and `x[2]` the interpolated value; it is forwarded to `simulationsApply` as `fun_pl`, therefore it has to fulfil all requirements for such functions; `interpolationErrorFunctions` provides some common examples
`fun_Rpl`	function to summarise the location-and-plume-wise errors, is forwarded to `simulationsApply` as `fun_Rpl`; if input does not fit into memory, it cannot be applied (causes warning)
`fun_Rpl_cellStats`	alternative function to summarise the location-and-plume-wise errors, is forwarded to `simulationsApply` as `fun_Rpl_cellStats`
`fun_l`	function to compare original and interpolated location-wise, i.e. generate one global map that takes into account all plumes; must have parameter `x`, then `x[,1]` refers to the original and `x[,2]` to the interpolated values; it is forwarded to `simulationsApply` as `fun_pl`, therefore it has to fulfil all requirements for such functions; `interpolationErrorFunctions` provides some common examples
`tmpfile`	filename for the raster file in case the result does not fit into memory; if `FALSE` the function stops with a warning and does not create a file
`overwrite`	boolean, if the file at `tmpfile` may be overwritten
`chunksize`	maximal number of cells to be processed at once – forwarded to `blockSize` inside

List of values and rasters (of same dimension as the values of the simulations):

"cost": result of fun_Rpl if available (if not, warning), else result of fun_Rpl_cellStats (to guarantee that there is always a value)

"cost_cellStats": result of fun_Rpl_cellStats (if this is not in "cost")

"error_locationsplumes": raster, result of fun_error

"interpolated": result of the interpolation with fun_interpolation

"costLocations": result of fun_l

kristina.helle@uni-muenster.de

  data(radioactivePlumes)
  ## preparation
  idw0z = replaceDefault(idw0, newDefaults = list(
    formula = z ~ 1))[[1]]
  sampleLocations100 = sample.int(nLocations(radioactivePlumes), 100)
  fun_Rpl_mean = function(x, nout = 1){ 
    mean(x[,1], na.rm = TRUE)
  }
  ## compute interpolation error 
  ## Not run: 
  ## takes some seconds
    interpolationError_delineation <- interpolationError(
      simulations = radioactivePlumes,
      locations = sampleLocations100,
      kinds = 2,
      fun_interpolation = idw0z,
      fun_error = delineationError,
      fun_Rpl = fun_Rpl_mean,
      fun_Rpl_cellStats = "mean",
      fun_l = delineationErrorMap
    )    
  # cost  
  interpolationError_delineation[["cost_cellStats"]]  
  ## plot error map
  interpolationErrorMaps = radioactivePlumes
  interpolationErrorMaps@values = 
    stack(radioactivePlumes@values[[2]],
          interpolationError_delineation[["interpolated"]],
          interpolationError_delineation[["error_locationsplumes"]][[1]])
  interpolationErrorMapsSDF = extractSpatialDataFrame(interpolationErrorMaps, plumes = 1:5) 
  interpolationErrorMapsSDF@data$costMap = interpolationError_delineation[["costLocations"]]
  # original, interpolated, error (1: overestimation, 5: underestimation)
  spplotLog(interpolationErrorMapsSDF, zcol = 1:15)
  # error summary - mean error of all plumes
  spplot(interpolationErrorMapsSDF, zcol = "costMap")

  
## End(Not run)