stanglogitMer: stanglogitMer: Bayesian Generalized Logit Multilevels Models...

In michelescandola/stanglogitMer: Bayesian Hierarchical Generalized-Logit Model to study behavioural peripersonal space as a continuum

Description

The stanglogitMer package provides an interface to fit Bayesian Generalized Logit Multilevel Models. These models are useful when the dependent data has a sigmoid-like shape, with boundaries that are not limited within the 0-1 range, as expressed by the following function:

stanglogitMer fits the Generalized Logit Function by means of Stan.

Usage

stanglogitMer(dependent, growth.formula, shift.formula, random.formula,
  grouping.random, asymptoms.formula = NULL, covariate, data,
  cores = 1, chains = 4, warmup = 2000, iter = 4000, seed = NA,
  control = list(adapt_delta = 0.9))

Arguments

growth.formula	the formula for the fixed effects of the growth parameter.
shift.formula	the formula for the fixed effects of the shift parameter.
random.formula	the formula for the random effects (slopes).
grouping.random	the categorial factor to group the random effects (the intercept).
covariate	the x-axis covariate (distance, time, etc...) along which the dependent variable is distributed. You must not include it in the fixed or random effects.
data	the data frame
cores	Number of cores to use when executing the chains in parallel, which defaults to 1 but we recommend setting the mc.cores option to be as many processors as the hardware and RAM allow (up to the number of chains). For non-Windows OS in non-interactive R sessions, forking is used instead of PSOCK clusters.
chains	Number of Markov chains (defaults to 4).
warmup	Number of total iterations per chain (including warmup; defaults to 2000).
iter	A positive integer specifying number of warmup (aka burnin) iterations. This also specifies the number of iterations used for stepsize adaptation, so warmup samples should not be used for inference. The number of warmup should not be larger than iter and the default is 4000
seed	The seed for random number generation to make results reproducible. If NA (the default), Stan will set the seed randomly.
control	A named list of parameters to control the sampler's behavior. It defaults to NULL so all the default values are used. For a comprehensive overview see stan.

Details

a + \frac{k-a}{1+e^{-g \times (x-s)}}

By means of these functions, you can estimate the linear regression describing the growth parameter (g), and the shift parameter (s), taking into account of a multilevel structure, also known as random effects.

The necessity to split the formula in three different pieces may be confusing. In order to better explain how it works, let's say that we want to fit the following formula, with the classic lmer syntax: y ~ Condition * Group + (Condition|Participant).

In order to fit with stanglogitMer that model we will set:

• growth.formula = ~ Condition * Group

• shift.formula = ~ Condition * Group

• random.formula = ~ Condition

• grouping = ~ Participant

Details

The main function of stanglogitMer is stanglogitMer, which uses formula syntax to specify your model.

michelescandola/stanglogitMer documentation built on Nov. 4, 2019, 6:52 p.m.