thin_draws

thin

thin_draws.draws

thin_draws.rvar

Thin <code>draws</code> objects to reduce their size and autocorrelation in
the chains.

Provides useful tools for both users and developers of packages
for fitting Bayesian models or working with output from Bayesian models.
The primary goals of the package are to:
(a) Efficiently convert between many different useful formats of
draws (samples) from posterior or prior distributions.
(b) Provide consistent methods for operations commonly performed on draws,
for example, subsetting, binding, or mutating draws.
(c) Provide various summaries of draws in convenient formats.
(d) Provide lightweight implementations of state of the art posterior
inference diagnostics. References: Vehtari et al. (2021)
<doi:10.1214/20-BA1221>.

PaulChristian Buerkner

posterior

Tools for Working with Posterior Distributions

Paul-Christian Bürkner

Jonah Gabry

Matthew Kay

Aki Vehtari

Måns Magnusson

Rok Češnovar

Ben Lambert

Ozan Adıgüzel

Jacob Socolar

Noa Kallioinen

thin_draws function

<dl><dt>x</dt>
<dd>(draws) A <code>draws</code> object or another R object for which the method
is defined.</dd>
<dt>thin</dt>
<dd>(positive numeric) The period for selecting draws. Must
be between 1 and the number of iterations. If the value is not an
integer, the draws will be selected such that the number of draws
returned is equal to round(ndraws(x) / thin). Intervals between
selected draws will be either ceiling(thin) or floor(thin), such
that the average interval will be close to the thin value. If
<code>NULL</code>, it will be automatically calculated based on bulk and
tail effective sample size as suggested by Säilynoja et
al. (2022).</dd>
<dt>...</dt>
<dd>Arguments passed to individual methods (if applicable).</dd></dl>

Arguments

Thin draws objects — thin_draws

<dl>

<dt>x</dt>
<dd>(draws) A <code>draws</code> object or another R object for which the method
is defined.</dd>


<dt>thin</dt>
<dd>(positive numeric) The period for selecting draws. Must
be between 1 and the number of iterations. If the value is not an
integer, the draws will be selected such that the number of draws
returned is equal to round(ndraws(x) / thin). Intervals between
selected draws will be either ceiling(thin) or floor(thin), such
that the average interval will be close to the thin value. If
<code>NULL</code>, it will be automatically calculated based on bulk and
tail effective sample size as suggested by Säilynoja et
al. (2022).</dd>


<dt>...</dt>
<dd>Arguments passed to individual methods (if applicable).</dd>

</dl>

Thin <code>draws</code> objects

thin_draws: Thin `draws` objects

Description

Usage

Value

Arguments

References

Examples