Help Index
Jump to sections or pages using the links below:
Model formula | Interaction specifications | General summary |
Residual variance (Ve) | Nested regressions | Extract estimates (estim, coef, fixef, ranef) |
data frame | | Extracts variances, hyper-pars, and functions (vhest) |
linking kernels, features (linkid) | | Convergence diagnostics (conv) |
[MCMC chain] | Variance specification (V) | Trace and density plots (plot)|
workdir in/output | Reduced rank (dim, dimp) | Warnings and errors |
initial values (init) | trace and save options | |
verbose settings | linkid on rr() term | |
Model formula
Residual variance (Ve)
Data frame (data)
linking kernels, features (linkid)
MCMC chain settings
workdir in/output
initial values (init)
verbose settings
Output use
Use of summary() on the bayz output object produces a summary of parameter estimates from the fitted model including convergence diagnostics and Highest Posterior Density (HPD) regions.
The summary() method only lists a limited number of the so-called “traced” parameters - these are model-parameters for which all MCMC samples are saved in the output, allowing to compute convergence, HPD regions, and to plot traces and densities (using plot()). The traced parameters by default include: all scalar variance parameters, estimated variance-covariances up to dimension 4x4, the model mean, scalar regression coefficients, coefficient estimates from fixed effects with up to 4 levels, and nested regressions with up to 4 levels.
Coefficients estimates
Apart from using these functions, it is also quite straightforward to extract estimates directly from the bayz output object. All estimates are stored in the output in a list calledVariance components and PVEs
The vcomp() method extracts variance estimates from the output.