print.compare.loo

print.loo

a list of class <code>'loo'</code> (as returned by <code><a href="/link/loo?package=loo&version=0.1.2" rd-options="" data-mini-rdoc="loo::loo">loo</a></code> or
<code><a href="/link/waic?package=loo&version=0.1.2" rd-options="" data-mini-rdoc="loo::waic">waic</a></code>) or of class <code>'compare.loo'</code> (as returned by
<code><a href="/link/compare?package=loo&version=0.1.2" rd-options="" data-mini-rdoc="loo::compare">compare</a></code>

passed to <code><a href="/link/round?package=loo&version=0.1.2&to=base" rd-options="base" data-mini-rdoc="base::round">round</a></code>.

digits

logical. If <code>TRUE</code> (the default), a warning message will be
printed if any estimates for the Pareto shape parameter $k$ are
problematic. See section PSIS-LOO in <code><a href="/link/loo-package?package=loo&version=0.1.2" rd-options="" data-mini-rdoc="loo::loo-package">loo-package</a></code>. Ignored
if <code></code>

warn

logical. If <code>TRUE</code> the estimates of $k$ are plotted.
Ignored if <code>x</code> was generated by <code><a href="/link/waic?package=loo&version=0.1.2" rd-options="" data-mini-rdoc="loo::waic">waic</a></code>.

plot_k

We efficiently approximate leave-one-out cross-validation (LOO)
  using Pareto smoothed importance sampling (PSIS), a new procedure for
  regularizing importance weights. As a byproduct of our calculations, we also
  obtain approximate standard errors for estimated predictive errors, and for
  the comparison of predictive errors between two models. We also compute the
  widely applicable information criterion (WAIC).

print.loo: Print methods

Description

Usage

Arguments

Value