refit: Loss for a refitted (restricted) unpenalized model

In genscore: Generalized Score Matching Estimators

Loss for a refitted (restricted) unpenalized model

Description

Refits an unpenalized model restricted to the estimated edges, with lambda1=0, lambda2=0 and diagonal_multiplier=1. Returns Inf if no unique solution exists (when the loss is unbounded from below or has infinitely many minimizers).

Usage

refit(res, elts)

Arguments

res	A list of results returned by get_results().
elts	A list, elements necessary for calculations returned by get_elts().

Details

Currently the function only returns Inf when the maximum node degree is >= the sample size, which is a sufficient and necessary condition for nonexistence of a unique solution with probability 1 if symmetric != "symmetric". In practice it is also a sufficient and necessary condition for most cases and symmetric == "symmetric".

Value

A number, the loss of the refitted model.

Examples

# Examples are shown for Gaussian truncated to R+^p only. For other distributions
#   on other types of domains, please refer to \code{gen()} or \code{get_elts()},
#   as the way to call this function (\code{refit()}) is exactly the same in those cases.
n <- 50
p <- 30
domain <- make_domain("R+", p=p)
h_hp <- get_h_hp("min_pow", 1, 3)
mu <- rep(0, p)
K <- diag(p)
dm <- 1 + (1-1/(1+4*exp(1)*max(6*log(p)/n, sqrt(6*log(p)/n))))
x <- tmvtnorm::rtmvnorm(n, mean = mu, sigma = solve(K),
       lower = rep(0, p), upper = rep(Inf, p), algorithm = "gibbs",
       burn.in.samples = 100, thinning = 10)

elts_gauss_np <- get_elts(h_hp, x, setting="gaussian", domain=domain,
                centered=FALSE, profiled=FALSE, diag=dm)
res_nc_np <- get_results(elts_gauss_np, symmetric="symmetric",
               lambda1=0.35, lambda2=2, previous_res=NULL, is_refit=FALSE)
refit(res_nc_np, elts_gauss_np)

elts_gauss_p <- get_elts(h_hp, x, setting="gaussian", domain=domain,
               centered=FALSE, profiled=TRUE, diag=dm)
res_nc_p <- get_results(elts_gauss_p, symmetric="symmetric",
              lambda1=0.35, lambda2=NULL, previous_res=NULL, is_refit=FALSE)
refit(res_nc_p, elts_gauss_p)

elts_gauss_c <- get_elts(h_hp, x, setting="gaussian", domain=domain,
            centered=TRUE, diag=dm)
res_c <- get_results(elts_gauss_c, symmetric="or", lambda1=0.35,
           lambda2=NULL, previous_res=NULL, is_refit=FALSE)
refit(res_c, elts_gauss_c)

genscore documentation built on May 31, 2023, 6:28 p.m.