fitCReml: Fit a spatial model by Composite REML
In kohleth/spcreml: Spatial Modelling using Composite REML

Fit a spatial model by Composite REML

1 2	fitCReml(form, data, dS, coordsMat, init, pointsj, pointsjpair, pointsk, cov.model = cov.model, lower = NULL)

`form`	a formula that specifies the response and the fixed effect variables.
`data`	a data.frame that contains the response variable and the fixed effect variables.
`dS`	a list of functions to compute the first derivative of Sigma (dSigma/dtheta).
`coordsMat`	a N by D coordinate matrix. D being the spatial dimension, typically 2 or 3.
`init`	named vector of initial value for optimiation of `recl`. Must (at least) contain named elements sigma2 and phi. Can also contain nugget and kappa.
`pointsjpair`	A list of pair integers from `1` to `length(pointsj)`. Each pair indicates which pairs of `pointjs` components should be bundled together in the variance calculation. See Details.
`pointsk`	not used.
`cov.model`	character string that specifies the spatial covariance model. See `?geoR::cov.spatial`
`lower`	Numeric vector of lower bound for the estimating parameters. Default to NULL (no lower bound).
`poinstj`	A list of row numbers that indicates which row belongs to block j (block j is itself a union of block k and block l). See Details.