bootstrap_mer: Run Various Bootstrap for Mixed Models.

In marklhc/bootmlm: Bootstrap Resampling for Multilevel Models

Run Various Bootstrap for Mixed Models.

Description

Run multilevel parametric, residual, and case bootstrap with different options

Run multilevel parametric, residual, and case bootstrap with different options

Usage

bootstrap_mer(
  x,
  FUN,
  nsim = 1,
  seed = NULL,
  type = c("parametric", "residual", "residual_cgr", "residual_trans", "reb", "case"),
  corrected_trans = FALSE,
  lv1_resample = FALSE,
  reb_scale = FALSE,
  .progress = FALSE,
  verbose = FALSE,
  ...
)

bootstrap_mer(
  x,
  FUN,
  nsim = 1,
  seed = NULL,
  type = c("parametric", "residual", "residual_cgr", "residual_trans", "reb", "case"),
  corrected_trans = FALSE,
  lv1_resample = FALSE,
  reb_scale = FALSE,
  .progress = FALSE,
  verbose = FALSE,
  ...
)

Arguments

x	A fitted merMod object from lmer.
FUN	A function taking a fitted merMod object as input and returning the statistic of interest, which must be a (possibly named) numeric vector.
nsim	A positive integer telling the number of simulations, positive integer; the bootstrap R.
seed	Optional argument to set.seed.
type	A character string indicating the type of multilevel bootstrap. Currently, possible values are "parametric", "residual", "residual_cgr", "residual_trans", "reb", or "case".
corrected_trans	Logical indicating whether to use the correct variance-covariance matrix of the residuals. If FALSE, use the variance of y; if TRUE, use the variance of y - X \hat \beta. Only used for type = "residual_trans".
lv1_resample	Logical indicating whether to sample with replacement the level-1 units for each level-2 cluster. Only used for type = "case". Default is FALSE.
reb_scale	Logical indicating whether to scale the residuals for the random effect block bootstrap
.progress	Logical indicating whether to display progress bar (using txtProgressBar).
verbose	Logical indicating if progress should print output.
...	argument passed to .resid_resample.

Details

bootstrap_mer performs different bootstrapping methods to fitted model objects using the lme4 package. Currently, only models fitted using lmer is supported.

bootstrap_mer performs different bootstrapping methods to fitted model objects using the lme4 package. Currently, only models fitted using lmer is supported.

Value

An object of S3 class "boot", compatible with boot package's boot(). It contains the following components:

t0	The original statistic from FUN(x).
t	A matrix with nsim rows containing the bootstrap distribution of the statistic.
R	The value of nsim passed to the function.
data	The data used in the original analysis.

An object of S3 class "boot", compatible with boot package's boot(). It contains the following components:

t0	The original statistic from FUN(x).
t	A matrix with nsim rows containing the bootstrap distribution of the statistic.
R	The value of nsim passed to the function.
data	The data used in the original analysis.
seed	The value of .Random.seed when bootstrap_mer started to work.
statistic	The function FUN passed to bootstrap_mer.

See the documentation in for link[boot]{boot}() for the other components.

References

Carpenter, J. R., Goldstein, H., & Rasbash, J. (2003). A novel bootstrap procedure for assessing the relationship between class size and achievement. Journal of the Royal Statistical Society. Series C (Applied Statistics), 52, 431–443. https://doi.org/10.1111/1467-9876.00415

Chambers, R., & Chandra, H. (2013). A random effect block bootstrap for clustered data. Journal of Computational and Graphical Statistics, 22(2), 452–470. https://doi.org/10.1080/10618600.2012.681216

Davison, A. C. and Hinkley, D. V. (1997). Bootstrap methods and their application. Cambridge, UK: Cambridge University Press.

Morris, J. S. (2002). The BLUPs are not "best" when it comes to bootstrapping. Statistics & Probability Letters, 56(4), 425–430. https://doi.org/10.1016/S0167-7152(02)00041-X

Van der Leeden, R., Meijer, E., & Busing, F. M. T. A. (2008). Resampling multilevel models. In J. de Leeuw & E. Meijer (Eds.), Handbook of multilevel Analysis (pp. 401–433). New York, NY: Springer.

See Also

boot for single-level bootstrapping, bootMer for parametric and semi-parametric bootstrap implemented in lme4, and boot.ci for getting bootstrap confidence intervals.

• boot for single-level bootstrapping,

• bootMer for parametric and semi-parametric bootstrap implemented in lme4, and

• boot.ci for getting bootstrap confidence intervals and plot.boot for plotting the bootstrap distribution.

Examples

library(lme4)
fm01ML <- lmer(Yield ~ (1 | Batch), Dyestuff, REML = FALSE)
mySumm <- function(x) {
  c(getME(x, "beta"), sigma(x))
}
# Covariance preserving residual bootstrap
boo01 <- bootstrap_mer(fm01ML, mySumm, type = "residual", nsim = 100)
# Plot bootstrap distribution of fixed effect
library(boot)
plot(boo01, index = 1)
# Get confidence interval
boot.ci(boo01, index = 2, type = c("norm", "basic", "perc"))
# BCa using influence values computed from `empinf_`
boot.ci(boo01, index = 2, type = "bca", L = empinf_mer(fm01ML, mySumm, 2))
library(lme4)
fm01ML <- lmer(Yield ~ (1 | Batch), Dyestuff, REML = FALSE)
mySumm <- function(x) {
  c(getME(x, "beta"), sigma(x))
}
# Covariance preserving residual bootstrap
boo01 <- bootstrap_mer(fm01ML, mySumm, type = "residual", nsim = 100)
# Plot bootstrap distribution of fixed effect
library(boot)
plot(boo01, index = 1)
# Get confidence interval
boot.ci(boo01, index = 2, type = c("norm", "basic", "perc"))
# BCa using influence values computed from `empinf_mer`
boot.ci(boo01, index = 2, type = "bca", L = empinf_mer(fm01ML, mySumm, 2))

marklhc/bootmlm documentation built on May 24, 2023, 9:59 a.m.