Benchmark Bootstrap Methods

Compares the runtime of different tsbs() configurations by varying a single parameter (e.g., series length, number of assets, or number of bootstrap replicates).

Usage

benchmark_tsbs(
  x,
  setups,
  vary = c("num_boots", "n_boot", "n_assets"),
  values,
  times = 3,
  verbose = TRUE
)

Arguments

x

Numeric matrix of input data for bootstrapping.

setups

A named list of tsbs() argument lists. Each element should be a list of arguments to pass to tsbs(), excluding x and the varying parameter.

vary

Character string specifying which parameter to vary. One of:

"n_boot"

Vary the length of bootstrap series

"num_boots"

Vary the number of bootstrap replicates

"n_assets"

Vary the number of assets (columns)

values

Numeric vector of values for the varying parameter.

times

Integer number of times to repeat each benchmark for averaging. Default is 3.

verbose

Logical. If TRUE, print progress messages. Default is TRUE.

Value

An object of class tsbs_benchmark containing:

results

Data frame with columns: Setup, Parameter, Value, Time, Rep

summary

Data frame with mean and sd of times by Setup and Value

vary

The parameter that was varied

setups

The setup configurations used

Details

This function is useful for:

• Comparing computational cost of different bootstrap methods

• Understanding how runtime scales with data size

• Choosing appropriate settings for production use

See also

plot.tsbs_benchmark for visualization.

Examples

if (FALSE) { # \dontrun{
# Create test data
set.seed(123)
x <- matrix(rnorm(500 * 3), ncol = 3)

# Define setups to compare
setups <- list(
  "Plain" = list(bs_type = "moving", block_length = 1),
  "Moving" = list(bs_type = "moving", block_length = 5),
  "Stationary" = list(bs_type = "stationary")
)

# Benchmark varying number of replicates
bench <- benchmark_tsbs(
  x = x,
  setups = setups,
  vary = "num_boots",
  values = c(10, 25, 50, 100),
  times = 3
)

# View results
print(bench)
plot(bench)
} # }