varMat function

create a semivariogram matrix between a set of locations, or semivariogram matrices between and within two sets of locations

varMat will create a semivariogram matrix between all the supports in a set of locations (observations or prediction locations) or semivariogram matrices between all the supports in one or two sets of locations, and also between them.

## S3 method for class 'rtop'
varMat(object, varMatUpdate = FALSE, fullPred = FALSE, params = list(), ...) 
## S3 method for class 'SpatialPolygonsDataFrame'
varMat(object, object2 = NULL,...) 
## S3 method for class 'SpatialPolygons'
varMat(object, object2 = NULL, variogramModel,
     overlapObs, overlapPredObs, ...) 

## S3 method for class 'list'
varMat(object, object2 = NULL, coor1, coor2, maxdist = Inf, 
              variogramModel, diag = FALSE, sub1, sub2, 
              debug.level = ifelse(interactive(), 1, 0), ...)

Arguments

• object: either: 1) an object of class rtop (see rtop-package ) or 2) a

  SpatialPolygonsDataFrame, or SpatialPolygons, or 3) a

  matrix with geostatistical distances (see gDist

  or 4) a list with discretized supports

• varMatUpdate: logical; if TRUE, also existing variance matrices will be recomputed, if FALSE, only missing variance matrices will be computed

• fullPred: logical; whether to create the full covariance matrix also for the predictions, mainly used for simulations

• params: a set of parameters, used to modify the default parameters for the rtop package, set in getRtopParams.

• object2: if object is not an object of class rtop; an object of the same class as object with a possible second set of locations with support

• variogramModel: variogramModel to be used in calculation of the semivariogram matrix (matrices)

• ...: typical parameters to modify from the default parameters of the rtop-package (or modifications of the previously set parameters for the rtop-object), see also getRtopParams. These can also be passed in a list named params, as for the rtop-method. Typical parameters to modify for this function:

  • rresol = 100: miminum number of discretization points, in call to rtopDisc if necessary
  • rstype = "rtop": sampling type from areas, in call to rtopDisc if necessary
  • gDistPred = FALSE: use geostatistical distance for semivariogram matrices
  • gDist: parameter to set jointly gDistEst = gDistPred = gDist

• overlapObs: matrix with observations that overlap each other

• overlapPredObs: matrix with observations and predictionLocations

  that overlap each other

• coor1: coordinates of centroids of object

• coor2: coordinates of centre-of-gravity of object2

• maxdist: maximum distance between areas for inclusion in semivariogrma matrix

• diag: logical; if TRUE only the semivariogram values along the diagonal will be calculated, typical for semivariogram matrix of prediction locations

• sub1: semivariogram array for subtraction of inner variances of areas

• sub2: semivariogram array for subtraction of inner variances of areas

• debug.level: debug.level >= 1 will give output for every element

Returns

The lower level versions of the function calculates a semivariogram matrix between locations in object or between the locations in object

and the locations in object2. The method for object of type rtop

calculates semivariogram matrices between observation locations, between prediction locations, and between observation locations and prediction locations, and adds these to object.

Note

The argument varMatUpdate is typically used to avoid repeated computations of the same variance matrices. Default is FALSE, which will avoid recomputation of the variance matrix for the observations if the procedure is cross-validation before interpolation. Should be set to TRUE if the variogram Model has been changed, or if observation and/or prediction locations have been changed.

If an rtop-object contains observations and/or predictionLocations of type STSDF, the covariance matrix is computed based on the spatial properties of the object.

References

Skoien J. O., R. Merz, and G. Bloschl. Top-kriging - geostatistics on stream networks. Hydrology and Earth System Sciences, 10:277-287, 2006.

Skoien, J. O., Bloschl, G., Laaha, G., Pebesma, E., Parajka, J., Viglione, A., 2014. Rtop: An R package for interpolation of data with a variable spatial support, with an example from river networks. Computers & Geosciences, 67.

Author(s)

Jon Olav Skoien

See Also

gDist

Examples

## Not run:

library(sf)
rpath = system.file("extdata",package="rtop")
observations = st_read(rpath,"observations")
vmod = list(model = "Ex1", params = c(0.00001,0.007,350000,0.9,1000))
vm = varMat(observations, variogramModel = vmod)
## End(Not run)