Simulations: Class "Simulations"

In sensors4plumes: Test and Optimise Sampling Designs Based on Plume Simulations

Description

Class for maps of plume simulations with spatial reference and plume properties.

Objects from the Class

Objects hold plume simulations and additional data: plume simulations are maps, all defined by values at the same set of locations; there may be several maps for each plume (e.g. the concentration of different pollutants in the same accident scenario, values at different times, etc.). As all maps have the same spatial properties, it is sufficient to hold the locations once as a SpatialDataFrame. The simulations themselves can be represented by a matrix where rows refer to location and columns refer to different plumes; in case of several maps per plume this can be extended to an array. These values arrays are actually implemented as raster objects; thus they can hold more data than it fits into memory. To keep information that refers to the single plumes like the kind of scenario they belong to or their probability to occur, a data.frame of plumes is part of each Simulations object.

Objects may be created by calls to the function Simulations(locations, plumes, values).

Slots

locations:

SpatialDataFrame; locations and location-related data, e.g. population or cost to put a sensor here (n locations)

plumes:

data.frame; data related to plumes but not to locations, e.g. average impact of the plume (aggregated over space) or likelihood of this plume to occur — rows refer to plumes, columns to different kind of information (N plumes)

values:

raster; maps of plume values; each column represents the map of one plume: raster layers do not represent maps, but contain the values of the maps of all plumes (n x N values) — spatial properties of the values are meanigless and must have the standardised form —; there can be different layers to contain different kind of data, e.g. cumulative concentrations of a pollutant and time when pollution started.

Methods

nLocations

number of locations (equals nrows(values))

nPlumes

number of plumes (equals ncols(values))

nKinds

number of kind of values (equals nlayers(values))

cbind

cbind-like method to combine the plumes of different Simulations with same locations and value types; for details see cbind

extractSpatialDataFrame

extract one plume from one layer of the values and assign it to the locations and return the resulting SpatialDataFrame; ; for details see extractSpatialDataFrame

Author(s)

Kristina B. Helle, [e-mail]

Examples

# generate Simulations object: small, artificial example
data(SPixelsDF)
plumes = data.frame(source = c("A", "A", "B", "B", "B"),
                    date = c("2000-01-01", "2000-04-01", 
                             "2000-07-01", "2000-01-01", "2000-01-03"),
                    totalCost = runif(5, min = 5, max = 15))

values1 = replicate(n = nrow(plumes), expr = rlnorm(length(SPixelsDF), sdlog = 2))
values2 = replicate(n = nrow(plumes), expr = rnorm(length(SPixelsDF), m = 10, sd = 3))
values = stack(raster(x = values1, xmn = -90, xmx = 90, ymn = -90, ymx = 90, 
                      crs = "+init=epsg:4326" ), 
               raster(x = values2, xmn = -90, xmx = 90, ymn = -90, ymx = 90, 
                      crs = "+init=epsg:4326" ))

Simulations1 = Simulations(locations = SPixelsDF, plumes, values)

# nLocations, nPlumes, nKinds
nLocations(Simulations1)
nPlumes(Simulations1)
nKinds(Simulations1)

sensors4plumes documentation built on May 1, 2019, 10:27 p.m.