extract_draws.brmsfit

extract_draws

An R object typically of class <code>'brmsfit'</code>.

An optional data.frame for which to evaluate predictions. If
<code>NULL</code> (default), the original data of the model is used.

newdata

formula containing group-level effects to be considered in
the prediction. If <code>NULL</code> (default), include all group-level effects;
if <code>NA</code>, include no group-level effects.

re_formula

A flag indicating if new levels of group-level
effects are allowed (defaults to <code>FALSE</code>). Only relevant if
<code>newdata</code> is provided.

allow_new_levels

Indicates how to sample new levels for grouping
factors specified in <code>re_formula</code>. This argument is only relevant if
<code>newdata</code> is provided and <code>allow_new_levels</code> is set to
<code>TRUE</code>. If <code>"uncertainty"</code> (default), include group-level
uncertainty in the predictions based on the variation of the existing
levels. If <code>"gaussian"</code>, sample new levels from the (multivariate)
normal distribution implied by the group-level standard deviations and
correlations. This options may be useful for conducting Bayesian power
analysis. If <code>"old_levels"</code>, directly sample new levels from the
existing levels.

sample_new_levels

A flag indicating if correlation structures originally
specified via <code>autocor</code> should be included in the predictions.
Defaults to <code>TRUE</code>.

incl_autocor

Optional indices of observations for which to compute
out-of-sample rather than in-sample predictions. Only required in models
that make use of response values to make predictions, that is currently
only ARMA models.

Optional names of response variables. If specified, predictions
are performed only for the specified response variables.

resp

Positive integer indicating how many posterior samples should
be used. If <code>NULL</code> (the default) all samples are used. Ignored if
<code>subset</code> is not <code>NULL</code>.

nsamples

A numeric vector specifying the posterior samples to be used.
If <code>NULL</code> (the default), all samples are used.

subset

Small positive number for Gaussian process terms only. For
numerical reasons, the covariance matrix of a Gaussian process might not be
positive definite. Adding a very small number to the matrix's diagonal
often solves this problem. If <code>NULL</code> (the default), <code>nug</code> is
chosen internally.

Logical; If <code>TRUE</code> only draws related to the
computation of smooth terms will be extracted.

smooths_only

Logical; Indicates if offsets should be included in the
predictions. Defaults to <code>TRUE</code>.

offset

A named <code>list</code> of objects containing new data, which
cannot be passed via argument <code>newdata</code>. Required for objects passed
via <code><a rd-options="" href="/link/stanvars?package=brms&version=2.9.0" data-mini-rdoc="brms::stanvars">stanvars</a></code> and for <code><a rd-options="brms:cor_sar" href="/link/cor_sar?package=brms&version=2.9.0&to=brms%3Acor_sar" data-mini-rdoc="brms:cor_sar::cor_sar">cor_sar</a></code> and
<code><a rd-options="brms:cor_fixed" href="/link/cor_fixed?package=brms&version=2.9.0&to=brms%3Acor_fixed" data-mini-rdoc="brms:cor_fixed::cor_fixed">cor_fixed</a></code> correlation structures.

new_objects

Further arguments passed to <code><a rd-options="" href="/link/validate_newdata?package=brms&version=2.9.0" data-mini-rdoc="brms::validate_newdata">validate_newdata</a></code>.

This method helps in preparing brms models for certin post-processing
tasks most notably various forms of predictions. Unless you are a package
developer, you will rarely need to call <code>extract_draws</code> directly.

Fit Bayesian generalized (non-)linear multivariate multilevel models
using 'Stan' for full Bayesian inference. A wide range of distributions
and link functions are supported, allowing users to fit -- among others --
linear, robust linear, count data, survival, response times, ordinal,
zero-inflated, hurdle, and even self-defined mixture models all in a
multilevel context. Further modeling options include non-linear and
smooth terms, auto-correlation structures, censored data, meta-analytic
standard errors, and quite a few more. In addition, all parameters of the
response distribution can be predicted in order to perform distributional
regression. Prior specifications are flexible and explicitly encourage
users to apply prior distributions that actually reflect their beliefs.
Model fit can easily be assessed and compared with posterior predictive
checks and leave-one-out cross-validation. References: B<c3><bc>rkner (2017)
<doi:10.18637/jss.v080.i01>; Carpenter et al. (2017) <doi:10.18637/jss.v076.i01>.

PaulChristian Buerkner

brms

Bayesian Regression Models using 'Stan'

Paul-Christian B<c3><bc>rkner

extract_draws.brmsfit function

A named <code>list</code> of objects containing new data, which
cannot be passed via argument <code>newdata</code>. Required for objects passed
via <code><a rd-options='' href='stanvars'>stanvars</a></code> and for <code><a rd-options='brms:cor_sar' href='cor_sar'>cor_sar</a></code> and
<code><a rd-options='brms:cor_fixed' href='cor_fixed'>cor_fixed</a></code> correlation structures.

Further arguments passed to <code><a rd-options='' href='validate_newdata'>validate_newdata</a></code>.

extract_draws.brmsfit: Extract Data and Posterior Draws

Description

Usage

Arguments

Value