snapKrig (version 0.0.2)

sk_nLL: Negative log-likelihood for parameter vector p

Description

Returns the negative log-likelihood of covariance model pars_fix, given the observations in data grid g_obs. Parameter values are copied from the first argument, vector p, so that the function can be passed to numerical optimizers (etc).

Usage

sk_nLL(p, g_obs, pars_fix, X = 0, iso = FALSE, quiet = TRUE, log_scale = FALSE)

Value

numeric, the negative log-likelihood of p given data g_obs

Arguments

p

numeric vector of covariance parameters accepted by sk_pars_update

g_obs

sk object or list with entries 'gdim', 'gres', 'gval'

pars_fix

list of form returned by sk_pars (with entries 'y', 'x', 'eps', 'psill')

X

numeric, vector, matrix, or NA, the mean or its linear predictors, passed to sk_LL

iso

logical, indicates to use identical kernels for x and y (pars$x is ignored)

quiet

logical indicating to suppress console output

log_scale

logical, indicates that pars_fix contains log-parameter values

Details

This is a wrapper for sk_LL (times -1) that allows parameters to be passed as a numeric vector instead of a list. Parameters in p are copied to pars_fix and passed to the likelihood computer.

p is the vector of covariance parameters to test. Names in p are ignored; Its length and order should correspond with the pattern of NAs in pars_fix. Users should check that the desired parameter list is being constructed correctly by testing with: sk_pars_update(pars_fix, p, iso=iso, na_omit=TRUE).

See Also

sk sk_GLS sk_var sk_pars_update

Other likelihood functions: sk_LL()

Other variance-related functions: sk_GLS(), sk_LL(), sk_cmean(), sk_sim(), sk_var()

Examples

# set up example grid and data
g = sk(gdim=10, gval=stats::rnorm(10^2))

# get some default parameters and vectorize them
pars = sk_pars(g, 'gau')
p = sk_pars_update(pars)
sk_nLL(p, g, pars)

# change a parameter in the numeric vector and re-evaluate
p_compare = p
p_compare[1] = 2 * p_compare[1]
sk_nLL(p_compare, g, pars)

# repeat by calling sk_LL directly with modified parameters list
pars_compare = pars
pars_compare[['eps']] = 2 * pars_compare[['eps']]
-sk_LL(pars_compare, g)

# set up a subset of parameters to replace - eg when fitting those parameters
pars_fix = pars
pars_fix[['eps']] = NA
pars_fix[['y']][['kp']] = NA

# names in p_fit are for illustration only (only the order matters)
p_fit = c(eps=1, y.rho=1)

# replace NA parameter values in pars_fix to get completed parameters list
sk_pars_update(pars_fix, p_fit, na_omit=TRUE)

# make the replacement and evaluate likelihood in one call
sk_nLL(p_fit, g, pars_fix)

# equivalently:
pars_fit = pars
pars_fit[['eps']] = p_fit[1]
pars_fit[['y']][['kp']] = p_fit[2]
-sk_LL(pars_fit, g)