cope: covariance parameter estimates with fast methods for big data...
In jayverhoef/SSNbd: Spatial Modeling for Big Data Sets on Stream Networks

Description Usage Arguments Value Author(s)

covariance parameter estimates with fast methods for big data with SSN

1
2
3

cope(formula, ssn.object, CorModels = c("Exponential.tailup",
  "Exponential.taildown", "Exponential.Euclid"), use.nugget = TRUE,
  addfunccol = NULL, EstMeth = "REML", partIndxCol, parallel = FALSE)

`formula`	an object of class "formula" (or one that can be coerced to that class): a symbolic description of the model to be
`ssn.object`	an object of class `SpatialStreamNetwork`, representing a spatial stream network. This contains the variables used in the model.
`CorModels`	a vector of spatial autocorrelation models for stream networks.
`use.nugget`	add a nugget effect, default is TRUE. This can be thought of as a variance component for independent errors, adding a variance component only along the diagonal of the covariance matrix.
`addfunccol`	the name of the variable in the SpatialStreamNetwork object that is used to define spatial weights. For the tailup models, weights need to be used for branching. This is an additive function and is described in Ver Hoef and Peterson (2010). See example below.
`EstMeth`	Estimation method; either "ML" for maximum likelihood, or "REML" for restricted maximum likelihood (default).
`subSampIndxCol`	the column in the `points data.frame` within the `SpatialStreamNetwork` object that indexes the grouping to be used when subsampling.

an object of class "estCovParSSNbd", which is a list, where estCovPar is a vector of estimated covariance parameters, optimOut is the output from optim used to estimate the parameters.

Jay Ver Hoef

jayverhoef/SSNbd documentation built on April 1, 2020, 8:06 a.m.