gbm.loop: Calculate Coefficient Of Variation surfaces for gbm.auto...

In SimonDedman/gbm.auto: Automated Boosted Regression Tree Modelling and Mapping Suite

Calculate Coefficient Of Variation surfaces for gbm.auto predictions

Description

Bagging introduces stochasticity which can result in sizeable variance in output predictions by gbm.auto for small datasets. This function runs a user- specified number of loops through the same gbm.auto parameter combinations and calculates the Coefficient Of Variation in the predicted abundance scores for each site aka cell. This can be mapped, to spatially demonstrate the output variance range.

Usage

gbm.loop(
  loops = 10,
  savedir = tempdir(),
  savecsv = TRUE,
  calcpreds = TRUE,
  varmap = TRUE,
  measure = "CPUE",
  cleanup = FALSE,
  grids = NULL,
  samples,
  expvar,
  resvar,
  randomvar = FALSE,
  tc = c(2),
  lr = c(0.01),
  bf = 0.5,
  n.trees = 50,
  ZI = "CHECK",
  fam1 = c("bernoulli", "binomial", "poisson", "laplace", "gaussian"),
  fam2 = c("gaussian", "bernoulli", "binomial", "poisson", "laplace"),
  simp = TRUE,
  gridslat = 2,
  gridslon = 1,
  multiplot = FALSE,
  cols = grey.colors(1, 1, 1),
  linesfiles = TRUE,
  smooth = FALSE,
  savegbm = FALSE,
  loadgbm = NULL,
  varint = FALSE,
  map = TRUE,
  shape = NULL,
  RSB = FALSE,
  BnW = FALSE,
  alerts = FALSE,
  pngtype = c("cairo-png", "quartz", "Xlib"),
  gaus = TRUE,
  MLEvaluate = TRUE,
  runautos = TRUE,
  Min.Inf = NULL,
  ...
)

Arguments

loops	The number of loops required, integer.
savedir	Save outputs to a temporary directory (default) else change to current directory e.g. "/home/me/folder". Do not use getwd() here.
savecsv	Save coefficients of variation in simple & extended format.
calcpreds	Calculate coefficients of variation of predicted abundance?
varmap	Create a map of the coefficients of variation outputs?
measure	Map legend, coefficients of variation of what? Default CPUE.
cleanup	Remove gbm.auto-generated directory each loop? Default FALSE.
grids	See gbm.auto help for all subsequent params.
samples	See gbm.auto help.
expvar	See gbm.auto help.
resvar	See gbm.auto help.
randomvar	See gbm.auto help.
tc	See gbm.auto help.
lr	See gbm.auto help.
bf	See gbm.auto help.
n.trees	See gbm.auto help.
ZI	See gbm.auto help. Choose one.
fam1	See gbm.auto help. Choose one.
fam2	See gbm.auto help. Choose one.
simp	See gbm.auto help.
gridslat	See gbm.auto help.
gridslon	See gbm.auto help.
multiplot	See gbm.auto help. Default False
cols	See gbm.auto help.
linesfiles	See gbm.auto help; TRUE or linesfiles calculations fail.
smooth	See gbm.auto help.
savegbm	See gbm.auto help.
loadgbm	See gbm.auto help.
varint	See gbm.auto help.
map	See gbm.auto help.
shape	See gbm.auto help.
RSB	See gbm.auto help.
BnW	See gbm.auto help.
alerts	See gbm.auto help; default FALSE as frequent use can crash RStudio.
pngtype	See gbm.auto help. Choose one.
gaus	See gbm.auto help.
MLEvaluate	See gbm.auto help.
runautos	Run gbm.autos, default TRUE, turn off to only collate numbered-folder results.
Min.Inf	Dummy param for package testing for CRAN, ignore.
...	Additional params for gbm.auto sub-functions including gbm.step.

Details

Thanks to a 2023 improvement to gbm.auto and gbm.loop,

Value

Returns a data frame of lat, long, 1 predicted abundance per loop, and a final variance score per cell.

Author(s)

Simon Dedman, [e-mail]

Examples

# Not run: downloads and saves external data.
library("gbm.auto")
data(grids) # load grids
data(samples) # load samples
gbmloopexample <- gbm.loop(loops = 2, samples = samples,
grids = grids, expvar = c(4:10), resvar = 11, simp = F)

SimonDedman/gbm.auto documentation built on Feb. 9, 2024, 5:12 p.m.