Confidence interval for ellipsoid parameters using simulation — confint.ellipsoid • Kdirectional

Assuming normality of errors.

Usage

# S3 method for ellipsoid
confint(x, fun, nsim = 1000, probs = c(0.025, 0.975), Asolve.eps = 0.01, ...)

Arguments

x: Fitted ellipsoid
fun: Optional contrast function. See ellipse_contrast_2d for an example.
probs: The quantiles for the CI. Default is c(0.025, 0.975)
Asolve.eps: Inflate the A matrix diagonals by a factor of eps, to avoid numerical problems with (near) singular parameter sets.
...: Passed on to sample_ellipse_beta
tol: tolerance for absolute deviation in the ||beta||=1 constraint.