50% off: Unlimited data and AI learning.
State of Data and AI Literacy Report 2025

GLMMadaptive (version 0.1-6)

MixMod Methods: Various Methods for Standard Generics

Description

Methods for object of class "MixMod" for standard generic functions.

Usage

coef(object, …)

# S3 method for MixMod coef(object, …)

fixef(object, …)

# S3 method for MixMod fixef(object, …)

ranef(object, …)

# S3 method for MixMod ranef(object, post_vars = FALSE, …)

confint(object, parm, level = 0.95, …)

# S3 method for MixMod confint(object, parm = c("fixed-effects", "var-cov","extra"), level = 0.95, …)

anova(object, …)

# S3 method for MixMod anova(object, object2, test = TRUE, L = NULL, …)

fitted(object, …)

# S3 method for MixMod fitted(object, type = c("mean_subject", "subject_specific", "marginal"), link_fun = NULL, …)

residuals(object, …)

# S3 method for MixMod residuals(object, type = c("mean_subject", "subject_specific", "marginal"), link_fun = NULL, …)

Arguments

object, object2

objects inheriting from class "MixMod". When object2 is also provided, then the model behind object must be nested within the model behind object2.

post_vars

logical; if TRUE the posterior variances of the random effects are returned as an extra attribute of the numeric matrix produced by ranef().

parm

character string; for which type of parameters to calculate confidence intervals. Option "var-cov" corresponds to the variance-covariance matrix of the random effects. Option extra corresponds to extra (shape/dispersion) parameters in the distribution of the outcome (e.g., the θ parameter in the negative binomial family).

level

numeric scalar between 0 and 1 denoting the level of the confidence interval.

test

logical; should a p-value be calculated.

L

a numeric matrix representing a contrasts matrix. This is only used when in anova() only object is provided, and it can only be specified for the fixed effects. When L is used, a Wald test is performed.

type

character string indicating the type of fitted values / residuals to calculate. Option "mean_subject" corresponds to only using the fixed-effects part; option "subject_specific" corresponds to using both the fixed- and random-effects parts; option "marginal" is based in multiplying the fixed effects design matrix with the marginal coefficients obtained by marginal_coefs.

link_fun

the link_fun of marginal_coefs.

further arguments; currently none is used.

Value

The estimated fixed and random effects, coefficients (this is similar as in package nlme), confidence intervals fitted values (on the scase on the response) and residuals.

See Also

mixed_model, marginal_coefs

Examples

Run this code
# NOT RUN {
# simulate some data
set.seed(123L)
n <- 500
K <- 15
t.max <- 25

betas <- c(-2.13, -0.25, 0.24, -0.05)
D <- matrix(0, 2, 2)
D[1:2, 1:2] <- c(0.48, -0.08, -0.08, 0.18)

times <- c(replicate(n, c(0, sort(runif(K-1, 0, t.max)))))
group <- sample(rep(0:1, each = n/2))
DF <- data.frame(year = times, group = factor(rep(group, each = K)))
X <- model.matrix(~ group * year, data = DF)
Z <- model.matrix(~ year, data = DF)

b <- cbind(rnorm(n, sd = sqrt(D[1, 1])), rnorm(n, sd = sqrt(D[2, 2])))
id <- rep(1:n, each = K)
eta.y <- as.vector(X %*% betas + rowSums(Z * b[id, ]))
DF$y <- rbinom(n * K, 1, plogis(eta.y))
DF$id <- factor(id)

################################################

fm1 <- mixed_model(fixed = y ~ year + group, random = ~ year | id, data = DF,
                   family = binomial())

head(coef(fm1))
fixef(fm1)
head(ranef(fm1))


confint(fm1)
confint(fm1, "var-cov")

head(fitted(fm1, "subject_specific"))
head(residuals(fm1, "marginal"))

fm2 <- mixed_model(fixed = y ~ year * group, random = ~ year | id, data = DF,
                   family = binomial())

# likelihood ratio test between fm1 and fm2
anova(fm1, fm2)

# the same but with a Wald test
anova(fm2, L = rbind(c(0, 0, 0, 1)))
# }

Run the code above in your browser using DataLab