fvariogram: Computes empirical trace-variograms and...
In ogru/fdagstat: A functional geostatistics package.

Usage Arguments Details Author(s)

fvariogram(formula = NULL, g = NULL, Nlags = NULL, LagMax = NULL,
  LagTolerance = NULL, ArgStep = NULL, directions = "omni",
  angTolerance = 0, crossMode = "conventional", useResidual = FALSE,
  comments = TRUE)

`formula`	- Formula specifying which coordinates to use in variogram computation.
`g`	- Object of class `fstat` as produced by `fstat` function
`Nlags`	- Number of variogram lags (required)
`LagMax`	- Maximum lag distance (also required)
`LagTolerance`	- Optional. If not provided the code will set it to LagStep/2
`ArgStep`	- Discretization step of the functional data. Used for numerical integration. Note that the code assumes the same sampling for all functional data in contained within g.
`directions`	- It can be: 'omni' when it will compute omni-directional variogram (default), 'all' when it computes variograms along each input dimension, an `nPxd` matrix where `nP` = the number of coordinates, `d` = the number of directions. The code will normalize all vectors!
`angTolerance`	- Tolerance for computing directional variograms
`crossMode`	- Either `conventional` or `pseudo`, for conventional or pseudo trace-cross-variogram computation, set this value to "NA" if you want to skip trace-cross-variogram computation all together.
`useResidual`	- This is applicable only when the drift term was already estimated with `estimateDrift` function.
`comments`	- Whether or not to provide runtime comments!

Returns an augmented fstat version of g by inserting a gstatVariogram slot. If there are NA's in functional data, then the code will not compute regular trace variograms. Instead, it will use an approximation proposed by Gromenko and Kokoszka.

Ogy Grujic (ogyg@stanford.edu) and Alessandra Menafoglio (alessandra.menafoglio@poilimi.it).

ogru/fdagstat documentation built on May 29, 2019, 7:19 a.m.