simList-accessors-inputs: Simulation inputs

In SpaDES.core: Core Utilities for Developing and Running Spatially Explicit Discrete Event Models

Simulation inputs

Description

Accessor functions for the inputs slots in a simList object.

Usage

inputs(sim)

## S4 method for signature 'simList'
inputs(sim)

inputs(sim) <- value

## S4 replacement method for signature 'simList'
inputs(sim) <- value

inputArgs(sim)

## S4 method for signature 'simList'
inputArgs(sim)

inputArgs(sim) <- value

## S4 replacement method for signature 'simList'
inputArgs(sim) <- value

Arguments

sim	A simList object from which to extract element(s) or in which to replace element(s).
value	The object to be stored at the slot. See Details.

Details

These functions are one of three mechanisms to add the information about which input files to load in a spades call.

1. As arguments to a simInit call. Specifically, inputs or outputs. See ?simInit.

2. With the outputs(simList) function call.

3. By adding a function called .inputObjects inside a module, which will be executed during the simInit call. This last way is the most "modular" way to create default data sets for your model.

See below for more details.

Value

Returns or sets the value(s) of the input or output slots in the simList object.

inputs function or argument in simInit

inputs accepts a data.frame, with up to 7 columns. Columns are:

file	required, a character string indicating the file path. There is no default.
objectName	optional, character string indicating the name of the object that the loaded file will be assigned to in the simList. This object can therefore be accessed with sim$xxx in any module, where objectName = "xxx". Defaults to the filename without file extension or directory information.
fun	optional, a character string indicating the function to use to load that file. Defaults to the known extensions in SpaDES (found by examining .fileExtensions()). The package and fun can be jointly specified here as "packageName::functionName", e.g., "terra::rast".
package	optional character string indicating the package in which to find the fun);
loadTime	optional numeric, indicating when in simulation time the file should be loaded. The default is the highest priority at start(sim), i.e., at the very start.
interval	optional numeric, indicating at what interval should this same exact file be reloaded from disk, e.g,. 10 would mean every 10 time units. The default is NA or no interval, i.e, load the file only once as described in loadTime
arguments	is a list of lists of named arguments, one list for each fun. For example, if fun="raster", arguments = list(native = TRUE) will pass the argument "native = TRUE" to raster. If there is only one list, then it is assumed to apply to all files and will be recycled as per normal R rules of recycling for each fun.

Currently, only file is required. All others will be filled with defaults if not specified.

See the modules vignette for more details (browseVignettes("SpaDES.core")).

.inputObjects function placed inside module

Any code placed inside a function called .inputObjects will be run during simInit() for the purpose of creating any objects required by this module, i.e., objects identified in the inputObjects element of defineModule. This is useful if there is something required before simulation to produce the module object dependencies, including such things as downloading default datasets, e.g., downloadData('LCC2005', modulePath(sim)). Nothing should be created here that does not create an named object in inputObjects. Any other initiation procedures should be put in the "init" eventType of the doEvent function. Note: the module developer can use sim$.userSuppliedObjNames inside the function to selectively skip unnecessary steps because the user has provided those inputObjects in the simInit call. e.g., the following code would look to see if the user had passed defaultColor into during simInit. If the user had done this, then this function would not override that value with 'red'. If the user has not passed in a value for defaultColor, then the module will get it here:

if (!('defaultColor' %in% sim$.userSuppliedObjNames)) { sim$defaultColor <- 'red' }

See Also

SpaDES.core-package, specifically the section 1.2.2 on loading and saving.

Other functions to access elements of a 'simList' object: .addDepends(), checkpointFile(), envir(), events(), globals(), modules(), objs(), packages(), params(), paths(), progressInterval(), times()

Examples

#######################
# inputs
#######################

# Start with a basic empty simList
sim <- simInit()

test <- 1:10
tmpdir <- file.path(tempdir(), "inputs") |> checkPath(create = TRUE)
tmpFile <- file.path(tmpdir, "test.rds")
saveRDS(test, file = tmpFile)
inputs(sim) <- data.frame(file = tmpFile) # using only required column, "file"
inputs(sim) # see that it is not yet loaded, but when it is scheduled to be loaded
simOut <- spades(sim)
inputs(simOut) # confirm it was loaded
simOut$test

# can put data.frame for inputs directly inside simInit call
allTifs <- dir(getMapPath(tempdir()), full.names = TRUE)

# next: .objectNames are taken from the filenames (without the extension)
# This will load all 5 tifs in the SpaDES sample directory, using
#   the rast fuction in the terra package, all at time = 0
sim <- simInit(
  inputs = data.frame(
    files = allTifs,
    functions = "rast",
    package = "terra",
    loadTime = 0,
    stringsAsFactors = FALSE)
)

##############################
# A fully described inputs object, including arguments:
files <- dir(getMapPath(tempdir()), full.names = TRUE)

# arguments must be a list of lists. This may require I() to keep it as a list
#   once it gets coerced into the data.frame.
# arguments = I(rep(list(native = TRUE), length(files)))
filelist <- data.frame(
  objectName = paste0("Maps", 1:5),
  files = files,
  functions = "terra::rast",
  # arguments = arguments,
  loadTime = 0,
  intervals = c(rep(NA, length(files) - 1), 10)
)
inputs(sim) <- filelist
spades(sim)

# Example showing loading multiple objects from global environment onto the
#   same object in the simList, but at different load times
a1 <- 1
a2 <- 2
# Note arguments must be a list of NROW(inputs), with each element itself being a list,
#  which is passed to do.call(fun[x], arguments[[x]]), where x is row number, one at a time
args <- lapply(1:2, function(x) {
               list(x = paste0("a", x),
                    envir = environment()) # may be necessary to specify in which envir a1, a2
                                           # are located, if not in an interactive session
               })
inputs <- data.frame(objectName = "a", loadTime = 1:2, fun = "base::get", arguments = I(args))
a <- simInit(inputs = inputs, times = list(start = 0, end = 1))
a <- spades(a)
identical(a1, a$a)

end(a) <- 3
a <- spades(a) # different object (a2) loaded onto a$a
identical(a2, a$a)

# Clean up after
unlink(tmpdir, recursive = TRUE)

SpaDES.core documentation built on June 22, 2024, 10:32 a.m.