mr_optim_fit: Runs multiresolution optimization routine

In daktx2/MSSS: Fits multiresolution spatial models with spatially varying resolution

Description

Runs multiresolution optimization routine

Usage

mr_optim_fit(yy, locationss, knots_r1, spatial_dimension, kernel_width,
  smoothness, shrinkage, maxiter, lambdatree_seq, tau_init = tau_init,
  tau_a = tau_a, tau_b = tau_b, em_tol = 0.001, m_tol = 1e-05,
  design_mat = NULL)

Arguments

yy	response values
locationss	locations of observations, either 1d vector or matrix with 2 columns
spatial_dimension	spatial dimension of observations, either 1 or 2
kernel_width	width of the kernel
smoothness	smoothness parameter for kernel
shrinkage	vector of parameter d, the shrinkage that increases in resolution, long vector
maxiter	maximum number of iterations to run
lambdatree_seq	decreasing sequence of lambdas, the parameter that controls sparsity
tau_init	initial tau value
tau_a	a parameter for gamma prior on tau
tau_b	b parameter for gamma prior on tau
em_tol	tolerance for when tau has converged
m_tol	default is .00001, tolerance for when the coefficients have converged
design_mat	optional design matrix for fixed effects

Value

list to be processed by prediction_function

daktx2/MSSS documentation built on May 24, 2020, 4:28 a.m.