Ensemble regression

This is the primary function for ensemble regression. It will take ensemble values in time and/or values in the predictor (X), and regress them on ensemble values in time and/or values in Y (the predictand). The function will then apply the ensemble linear model to the full length of X to create a modeled Y. Will also optionally create plots.

regressEns(
  time.x,
  values.x,
  time.y,
  values.y,
  bin.vec = NA,
  bin.step = NA,
  bin.fun = mean,
  max.ens = NA,
  percentiles = c(0.025, 0.25, 0.5, 0.75, 0.975),
  recon.bin.vec = NA,
  min.obs = 10,
  gaussianize = TRUE
)

Arguments

time.x

matrix of age/time ensembles, or single column

values.x

matrix of values ensembles, or single column

time.y

matrix of age/time ensembles, or single column

values.y

matrix of values ensembles, or single column

bin.vec

vector of bin edges for binning step

bin.step

spacing of bins, used to build bin step

bin.fun

function to use during binning (mean, sd, and sum all work)

max.ens

maximum number of ensemble members to regress

percentiles

quantiles to calculate for regression parameters

recon.bin.vec

bin vector to use for the modeled regression.

min.obs

minimum number of points required to calculate regression

gaussianize

Boolean flag indicating whether the values should be mapped to a standard Gaussian prior to analysis.

Value

list of ensemble output

Long-form example

View a full-fledged example of how to use this function.

See also

Other regress: plotRegressEns(), plotScatterEns(), plotTrendLinesEns(), regress()

Author

Nick McKay