summaryPlumes: Summarise values of Simulations by plumes
In sensors4plumes: Test and Optimise Sampling Designs Based on Plume Simulations

Description Usage Arguments Details Value Author(s) Examples

This function is similar to apply for the values of Simulations plume-wise. It takes a function and returns a vector with a value for each plume: the result of the function applied to all values belonging to this plume. As values may be too big to keep in memory, the function is applied chunk-wise; therefore it only works for associative functions. A (slower) method without this restriction is It is combined with a global summary function and thus provides a general basic form of a cost function.

summaryPlumes(simulations,
  locations = 1:nLocations(simulations), 
  plumes = 1:nPlumes(simulations), 
  kinds = 1, 
  fun, summaryFun = weightedMean,
  weight = 1, na.rm = FALSE, ...)
weightedMean(x, weight, na.rm = FALSE)

`simulations`	`Simulations`
`locations`	indices of `locations` to be taken into account (to apply the function to a subset of `simulations`); for invalid indices `NA` values are used; multiple are ignored
`plumes`	indices of `plumes` to be taken into account; for invalid indices `NA` values are used
`kinds`	index of kind of `values` to be taken into account; only one kind can be used, all but the first given index are ignored
`fun`	`function` to be applied; it must have a single parameter (vector), be associative, i.e. `fun(c(a,b)) = fun(c(fun(a),b))`, work also on vectors of length 0 and return a single value; in addition it needs a parameter `na.rm` or `...`; possible choices are `sum`, `prod`, `max`, `min`

`summaryFun`	`function` to summarise the plume-wise results (as first parameter) to a global value; by default it is a weighted mean; in addition it can have the parameters `weight`, or use `...`
`weight`	to be used by `summaryFun`: either a numeric value of length 1 or same length as `plumes` or a `character` indicating a column of the `data` associated with the `plumes` of `simulations`
`na.rm`	`logical` how to treat missing values by `fun` (has no influence on `summaryFun`)
`...`	further arguments to be forwarded to `fun` or `summaryFun`
`x`	vector

To use a different summaryFun you may use replaceDefault with
kind = "summaryFun.summaryPlumes" to check the function beforehand or to generate it by resetting the default parameters of an existing function.

A list of

`summary`	the result of `summaryFun`
`summaryPlumes`	the plume-wise results of `fun`, one value for each selected `plume`

Kristina B. Helle, kristina.helle@uni-muenster.de

data(radioactivePlumes) 
# sample of locations, e.g. proposed sensor set
locSample = sample.int(nLocations(radioactivePlumes), 10) 

# plume detection 
# (number of plumes not detected at any of locSample at a level of more than 1e-7)
plumeMin_radioactivePlumes = 
  summaryPlumes(radioactivePlumes, locations = locSample, kinds = 2, 
  fun = "min",
  summaryFun = function(x, weight = 1, na.rm = TRUE){sum(x < 1e-7)})