Compute bootstrap confidence intervals for functional statistics such as the mean function, depth values, or regression coefficients.
Usage
fdata.bootstrap.ci(
fdataobj,
statistic,
n.boot = 200,
alpha = 0.05,
method = c("percentile", "basic", "normal"),
seed = NULL
)Arguments
- fdataobj
An object of class 'fdata'.
- statistic
A function that computes the statistic of interest. Must take an fdata object and return a numeric vector.
- n.boot
Number of bootstrap replications (default 200).
- alpha
Significance level for confidence intervals (default 0.05 for 95 percent CI).
- method
CI method: "percentile" for simple percentile method, "basic" for basic bootstrap, "normal" for normal approximation (default "percentile").
- seed
Optional seed for reproducibility.
Value
A list of class 'fdata.bootstrap.ci' with components:
- estimate
The statistic computed on the original data
- ci.lower
Lower confidence bound
- ci.upper
Upper confidence bound
- boot.stats
Matrix of bootstrap statistics (n.boot x length(statistic))
- alpha
The significance level used
- method
The CI method used
Examples
# Create functional data
t <- seq(0, 1, length.out = 50)
X <- matrix(0, 20, 50)
for (i in 1:20) X[i, ] <- sin(2*pi*t) + rnorm(50, sd = 0.1)
fd <- fdata(X, argvals = t)
# Bootstrap CI for the mean function (returns numeric vector)
ci_mean <- fdata.bootstrap.ci(fd,
statistic = function(x) as.numeric(mean(x)$data),
n.boot = 100)
# Bootstrap CI for depth values
ci_depth <- fdata.bootstrap.ci(fd,
statistic = function(x) depth.FM(x),
n.boot = 100)