Fitted gjrm object
A fitted joint model returned by function gjrm and of class "gjrm", "SemiParBIV", "SemiParTRIV", etc.
fit: List of values and diagnostics extracted from the output of the algorithm.
gam1: Univariate fit for equation 1. See the documentation of mgcv for full details.
gam2, gam3, ...: Univariate fit for equation 2, equation 3, etc.
coefficients: The coefficients of the fitted model.
weights: Prior weights used during model fitting.
sp: Estimated smoothing parameters of the smooth components.
iter.sp: Number of iterations performed for the smoothing parameter estimation step.
iter.if: Number of iterations performed in the initial step of the algorithm.
iter.inner: Number of iterations performed within the smoothing parameter estimation step.
theta: Estimated dependence parameter linking the two equations.
n: Sample size.
X1, X2, X3, ...: Design matrices associated with the linear predictors.
X1.d2, X2.d2, X3.d2, ...: Number of columns of X1, X2, X3, etc.
l.sp1, l.sp2, l.sp3, ...: Number of smooth components in the equations.
He: Penalized -hessian/Fisher. This is the same as HeSh for unpenalized models.
HeSh: Unpenalized -hessian/Fisher.
Vb: Inverse of He. This corresponds to the Bayesian variance-covariance matrix used for confidence/credible interval calculations.
F: This is obtained multiplying Vb by HeSh.
t.edf: Total degrees of freedom of the estimated bivariate model. It is calculated as sum(diag(F)).
edf1, edf2, edf3, ...: Degrees of freedom for the two equations of the fitted bivariate model (and for the third and fourth equations if present. They are calculated when splines are used.
bs.mgfit: List of values and diagnostics extracted from magic in mgcv.
conv.sp: If TRUE then the smoothing parameter selection algorithm stopped before reaching the maximum number of iterations allowed.
wor.c: Working model quantities.
eta1, eta2, eta3, ...: Estimated linear predictors for the two equations (as well as the third and fourth equations if present).
y1, y2: Responses of the two equations.
logLik: Value of the (unpenalized) log-likelihood evaluated at the (penalized or unpenalized) parameter estimates.
respvec: List containing response vectors.
Maintainer: Giampiero Marra giampiero.marra@ucl.ac.uk
gjrm, summary.gjrm