Perform principle components analysis (PCA) across an ensemble

Ensemble PCA, or Monte Carlo Empirical Orthogonal Functions

pcaEns(
  bin.list,
  method = "ppca",
  weights = NA,
  pca.type = "corr",
  gaussianize = TRUE,
  n.pcs = 8,
  n.ens = 1000,
  simulateTrendInNull = FALSE
)

Arguments

bin.list

A list of binned data, the output of binTs()

method

What method to use for PCA? pcaMethods::listPcaMethods() for options. "ppca" is default. Other options may not work in GeoChronR.

weights

Vector of weights to apply to timeseries in the bin.list

pca.type

Correlation ("corr" - default) or Covariance ("cov"), matrix

gaussianize

Map input data to a standard Gaussian distribution? This is only relevant for correlation matrices, covariance matrices will not be gaussianized. (default = TRUE)

n.pcs

number of PCs/EOFs to calculate

n.ens

how many ensemble members to calculate

simulateTrendInNull

Should the null include the trend?

Long-form example

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

See also

Other pca: ar1Surrogates(), createSyntheticTimeseries(), plotPcaEns(), plotScreeEns()