brmsformula-helpers.Rd

% Generated by roxygen2: do not edit by hand
% Please edit documentation in R/brmsformula.R
\name{brmsformula-helpers}
\alias{brmsformula-helpers}
\alias{bf-helpers}
\alias{nlf}
\alias{lf}
\alias{set_nl}
\alias{set_rescor}
\alias{nlf}
\alias{lf}
\alias{set_nl}
\alias{set_rescor}
\alias{set_mecor}
\title{Linear and Non-linear formulas in \pkg{brms}}
\usage{
nlf(formula, ..., flist = NULL, dpar = NULL, resp = NULL)

lf(..., flist = NULL, dpar = NULL, resp = NULL)

set_nl(nl = TRUE, dpar = NULL, resp = NULL)

set_rescor(rescor = TRUE)

set_mecor(mecor = TRUE)
}
\arguments{
\item{formula}{Non-linear formula for a distributional parameter.
The name of the distributional parameter can either be specified
on the left-hand side of \code{formula} or via argument \code{dpar}.}

\item{...}{Additional \code{formula} objects to specify 
predictors of non-linear and distributional parameters. 
Formulas can either be named directly or contain
names on their left-hand side. 
The following are distributional parameters of specific families
(all other parameters are treated as non-linear parameters):
\code{sigma} (residual standard deviation or scale of
the \code{gaussian}, \code{student}, \code{skew_normal}, 
\code{lognormal} \code{exgaussian}, and \code{asym_laplace} families);
\code{shape} (shape parameter of the \code{Gamma},
\code{weibull}, \code{negbinomial}, and related
zero-inflated / hurdle families); \code{nu} (degrees of freedom 
parameter of the \code{student} and \code{frechet} families);
\code{phi} (precision parameter of the \code{beta} 
and \code{zero_inflated_beta} families);
\code{kappa} (precision parameter of the \code{von_mises} family);
\code{beta} (mean parameter of the exponential component
of the \code{exgaussian} family);
\code{quantile} (quantile parameter of the \code{asym_laplace} family);
\code{zi} (zero-inflation probability); 
\code{hu} (hurdle probability);
\code{zoi} (zero-one-inflation probability);
\code{coi} (conditional one-inflation probability);
\code{disc} (discrimination) for ordinal models;
\code{bs}, \code{ndt}, and \code{bias} (boundary separation,
non-decision time, and initial bias of the \code{wiener}
diffusion model).
By default, distributional parameters are modeled 
on the log scale if they can be positive only or on the 
logit scale if the can only be within the unit interval.
See 'Details' for more explanation.}

\item{flist}{Optional list of formulas, which are treated in the 
same way as formulas passed via the \code{...} argument.}

\item{dpar}{Optional character string specifying the distributional 
parameter to which the formulas passed via \code{...} and
\code{flist} belong.}

\item{resp}{Optional character string specifying the response 
variable to which the formulas passed via \code{...} and
\code{flist} belong. Only relevant in multivariate models.}

\item{nl}{Logical; Indicates whether \code{formula} should be
treated as specifying a non-linear model. By default, \code{formula} 
is treated as an ordinary linear model formula.}

\item{rescor}{Logical; Indicates if residual correlation between
the response variables should be modeled. Currently this is only
possible in multivariate \code{gaussian} and \code{student} models.
Only relevant in multivariate models.}

\item{mecor}{Logical; Indicates if correlations between latent variables
defined by \code{\link{me}} terms should be modeled. Defaults to \code{TRUE}.}
}
\value{
For \code{lf} and \code{nlf} a \code{list} that can be 
  passed to \code{\link[brms:brmsformula]{brmsformula}} or added 
  to an existing \code{brmsformula} or \code{mvbrmsformula} object. 
  For \code{set_nl} and \code{set_rescor} a logical value that can be 
  added to an existing \code{brmsformula} or \code{mvbrmsformula} object.
}
\description{
Helper functions to specify linear and non-linear
formulas for use with \code{\link[brms:brmsformula]{brmsformula}}.
}
\examples{
# add more formulas to the model
bf(y ~ 1) + 
  nlf(sigma ~ a * exp(b * x), a ~ x) + 
  lf(b ~ z + (1|g), dpar = "sigma") +
  gaussian()

# specify 'nl' later on
bf(y ~ a * inv_logit(x * b)) +
  lf(a + b ~ z) +
  set_nl(TRUE)
  
# specify a multivariate model
bf(y1 ~ x + (1|g)) + 
  bf(y2 ~ z) +
  set_rescor(TRUE)
}
\seealso{
\code{\link{brmsformula}}, \code{\link{mvbrmsformula}}
}