dot-HVcalc: HYPERVOLUME COMPUTATION WRAPPER
In CeresBarros/ToolsCB: Ceres's Useful R Functions

.HVcalc

R Documentation

HYPERVOLUME COMPUTATION WRAPPER

Description

Calculates two hypervolumes and compares them in terms of overlap and distance. When appropriate (i.e. HVmethod == "box"), it also calculates the disjunct factor and repeats the calculations with larger bandwidths (if freeBW == TRUE)

Usage

.HVcalc(
  big.table,
  HVpoints,
  noAxes,
  HVmethod,
  ordination,
  HVnames,
  bwHV1,
  bwHV2,
  verbose,
  ...
)

Arguments

`big.table`	is a `data.table` contained the raw data for the two hypervolumes. Used to subset `HVpoints` according to HVnames, so rows have to correspond
`HVpoints`	`data.table` of points used to build hypervolumes and final number of axes.
`noAxes`	determines the number of axes/columns to use for hypervolume calculation
`HVmethod`	determines the method used to calculate hypervolumes - passed to `hypervolume::hypervolume` method argument
`ordination`	determines whether hypervolumes will be calculated on the raw data ("none") of on ordination factor scores (one of "PCA", "HillSmith", "dudi.mix"). Defaults to using a PCA.
`HVnames`	`character` vector of length 2 of hypervolume names.
`bwHV1`	determines the bandwidth value for first hypervolume. Only used if `freeBW` is FALSE. see `hypervolume::hypervolume` Only used if `HVmethod` is "box" or "gaussian".
`bwHV2`	determines the bandwidth value for second hypervolume. Only used if `freeBW` is FALSE. see `hypervolume::hypervolume` Only used if `HVmethod` is "box" or "gaussian".
`verbose`	is passed to `⁠hypervolume::*⁠` functions to control printing of diagnostic outputs
`...`	further arguments passed to `hypervolume::hypervolume`.

Value

a list with entries:

'HV1'
'HV2'
'volume.set' - the result of hypervolume::hypervolume_set
'hv.centroid.dist' and 'hv.min.dist' - the result of hypervolume::hypervolume_distance(..., type = "centroid") and hypervolume::hypervolume_distance(..., type = "minimum")
'Bandwidth' - NA if HVmethod is not "box" or "gaussian")
'HV1.disjfact' and 'HV2.disjfact' (NA if HVmethod is not "box")
'SVM_nu' and 'SVM_gama' (NA if HVmethod is "box" or "gaussian")