runBootstrap: Run a bootstrap in StoX

In Sea2Data/Rstox: Running Stox functionality independently in R

Description

runBootstrap is a wrapper function for the bootstrap functions below.

runBootstrap_1.6 runs the bootstrap of Rstox 1.6. The bootstrap changed from Rstox 1.6 to 1.7, by applying sorting prior to sampling, and different results (but with the same expected value) should be expected. Using runBootstrap_1.6 identical results from previous runs should be expected.

runBootstrap_AcousticTrawl runs a simple bootstrap of biotic PSUs within strata.

runBootstrap_SweptAreaLength runs a simple bootstrap of biotic PSUs within strata.

runBootstrap_SweptAreaTotal runs a simple bootstrap of biotic PSUs within strata.

Usage

runBootstrap(projectName, bootstrapMethod = "AcousticTrawl",
  acousticMethod = PSU ~ Stratum, bioticMethod = PSU ~ Stratum,
  nboot = 5, startProcess = "TotalLengthDist",
  endProcess = "SuperIndAbundance", seed = 1, cores = 1,
  msg = TRUE, sorted = TRUE, JavaMem = getRstoxDef("JavaMem"), ...)

runBootstrap_1.6(projectName, bootstrapMethod = "AcousticTrawl",
  acousticMethod = PSU ~ Stratum, bioticMethod = PSU ~ Stratum,
  nboot = 5, startProcess = "TotalLengthDist",
  endProcess = "SuperIndAbundance", seed = 1, cores = 1,
  msg = TRUE, ...)

runBootstrap_AcousticTrawl(projectName, acousticMethod = PSU ~ Stratum,
  bioticMethod = PSU ~ Stratum, nboot = 5,
  startProcess = "TotalLengthDist", endProcess = "SuperIndAbundance",
  seed = 1, cores = 1, msg = TRUE, sorted = TRUE,
  JavaMem = getRstoxDef("JavaMem"), ...)

runBootstrap_SweptAreaLength(projectName, acousticMethod = NULL,
  bioticMethod = PSU ~ Stratum, nboot = 5,
  startProcess = "TotalLengthDist", endProcess = "SuperIndAbundance",
  seed = 1, cores = 1, msg = TRUE, sorted = TRUE,
  JavaMem = getRstoxDef("JavaMem"), ...)

runBootstrap_SweptAreaTotal(projectName, acousticMethod = NULL,
  bioticMethod = PSU ~ Stratum, startProcess = "SweptAreaDensity",
  endProcess = NULL, nboot = 5, seed = 1, cores = 1,
  ignore.case = TRUE, sorted = TRUE, ...)

Arguments

projectName	The name or full path of the project, a baseline object (as returned from getBaseline or runBaseline, og a project object (as returned from openProject).
bootstrapMethod	The method to use for the bootstrap. Currently implemented are given in the following table: bootstrapMethod Description AcousticTrawl Bootstrap of acoustic tralw surveys, where both acoustic and biotic data are resampled SweptAreaLength Bootstrap only biotic data with length information SweptAreaTotal For surveys with information only about total catch (count or weight), bootstrap biotic stations
acousticMethod, bioticMethod	Specification of the method to use for bootstrapping the acoustic and biotic data. Currently only one method is available for acoustic and one for biotic data: acousticMethod = PSU~Stratum, bioticMethod = PSU~Stratum. Other methods are planned in later versions, involving the levels of the data given in the below table. Level Acoustic Biotic Survey Survey Survey Stratum Stratum Stratum Assignment Not relevant Assignment of biotic station groups to acoustic PSUs PSU Acoustic data averaged over e.g. one tansect Biotic station group EDSU Acoustic data averaged over e.g. one nmi Biotic station Sample Ping Individal catch sample
nboot	Number of bootstrap replicates.
startProcess	The start process of the bootstrapping, e.g., the last process before bio-stations have been assigned and NASC values have been calculated.
endProcess	The end process of the bootstrapping, e.g., the process returning a matrix containing the following columns: "Stratum", "Abundance", "weight", and grouping variables such as "age", "SpecCat", "sex" (typically "SuperIndAbundance").
seed	The seed for the random number generator (used for reproducibility).
cores	An integer giving the number of cores to run the bootstrapping over.
msg	Logical: if TRUE print messages from runBaseline().
sorted	Should the data be sorted prior to sampling?
JavaMem	The memory occupied by the Java virtual machine. Default is returned by getRstoxDef("JavaMem"). Reducing this may be usefull when using mutiple cores.
...	Used for backwards compatibility.
ignore.case	Logical: If TRUE ignore case in species category SpecCat.

Details

Resample (bootstrap) trawl stations based on swept area data and possibly also acoustic data to estimate uncertainty in estimates. By the default method (bootstrapMethod="AcousticTrawl"), the acoustic transect values (mean NASC along transects) and biotic stations (trawls) are resampled with replacement within each stratum for each bootstrap replicate, and the StoX project rerun and super individual abundance recalculated (or the output from a different process given by endProcess).

Value

list with (1) the abundance by length in the orginal model, (2) the abundance by length in the bootstrap run, (3) the abundance by super individuals in the orginal model, (4) the abundance by super individuals in the bootstrap run

Examples

## Not run: 
projectName <- "Test_Rstox"
boot <- runBootstrap(projectName, nboot=10, seed=1, bootstrapMethod="AcousticTrawl")

## End(Not run)

Sea2Data/Rstox documentation built on May 14, 2019, 8:58 a.m.